Olson 1957 I
Olson 1957 I
Olson 1957 I
,
05.2 (z-
I i" ;'
C,L
ACOUSTICAL
ENGINEERING
UNVERSIDADTECNOLOGICADECHiLE
B ISUOTECA
TORONTO LONDON
NEW YORK
COPYRIGHT ©1957, BY
D. VAN NOSTRAND COMPANY, INC.
IV PREFACE
dustrial expansion comes an increase in noise. Work is now under way
actively to control noise by the use of a variety of acoustic countermeasures.
The preceding brief description of the present status of acoustics shows
that it plays a very important part in our modern civilization. Furthermore,
the fundamentals and applications of the science of acoustics are so well
formulated and substantiated that a large area of the field of acoustics has
attained an engineering status. In preparing new material and in revising
existing material in the third edition, the same principles were followed as in
the first and second editions. Particular efforts have been directed towards
the development of analogies between electrical, mechanical and acoustical
systems because engineers have found that the reduction of a vibrating
system to the analogous electrical network is a valuable tool in the analysis
of vibrating systems. Each chapter has been brought up to date and ampli
fied. Two new chapters on Complete Sound Reproducing Systems and
.Yfeans for the Communication of Information have been added. As in the
first and second editions most of the illustrations contain several parts so
that a complete theme is depicted in a single illustration.
The author wishes to express his appreciation to Miss Patricia Duman
for her work in typing the manuscript and to his wife Lorene E. Olson
for assistance in compiling and correcting the manuscript.
HARRY F. OLSON
March,1957
CONTENTS
CHAPTER PAGE
1. SOUND WAVES
1.1 INTRODUCTION .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • . 1
D. Condensation .................................... 6
vi CONTE:'-JTS
CHAPTER PAGE
C. Point Source Radiating into a Solid Angle of 71" Stera
dians ............................................ 31
A. Circular Membrane .. ~ . . . . . . . . . . . . . . . . . . . . . . . . . . .. 62
CONTENTS vii
CHAPTER PAGE
3.6 LONGITUDINAL VIBRATION OF BARS 66
A. Inductance ....................................... 79
B. Mass ............................................ 79
D. Inertance ........................................ 80
V. ACOUSTICAL ELEMENTS
5.1 INTRODUCTION •..••.••.•...•.•••....••.....••••.•..•... 88
VIBRATING PISTON . . • . . • • . . . . . . . . . . . • • . . • . . . . . . . . • . . . • . 92
.,
VJ11 CONTENTS
CHAPTER PAGE
5.15 ACOUSTICAL IMPEDANCE OF A CIRCULAR ORIFICE IN A WALL
FLANGES ••.••.•..••..••••.•...••.........••••••••..... 99
CONTENTS
ix
CHAPTER PAGE
x CONTEKTS
CHAPTER PAGE
A. The Relation between the Voice Coil Mass, the Load
Chamber .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 225
VIII. MICROPHONES
8.1 INTRODUCTION 246
CONTENTS Xl
CHAPTER PAGE
3. Diaphragm Actuated Barium Titanate Micro
phone ...... . ................................ 260
..• • • . • . . • . . . • . • • • • . . . . • • • . . . . . . . . . . . . . . . . . . . .
A. Pressure Gradient Microphone ..................... 275
xu CO NT E:-'.JT S
CHAPTER PAGE
2. Line Microphone: Useful Directivity on the Line
CONTENTS Xlll
CHAPTER PAGE
b.
Ceramic Turnover Pickup ................ 362
c.
Magnetic Pickup ........................ 364
d.
Dynamic Pickup ......................... 367
e.
Frequency Modulation Pickup ............ 369
f.
Electronic Pickup ....................... 371
g.
Variable Resistance Pickup ............... 371
h.
Feedback Pickup ........................ 372
TEM 384
••....•..•...•.•••...•.•..••.•..•....•••••••...•.•
9.8 MAGNETIC TAPE CONVERSION SYSTEMS 390
.•..•...•...•.•.••
A. Frequency Conversion System ..................... 390
,
,0
XIV CONTENTS
CHAPTER PAGE
C. Headphone-Type Noise Reducer .................... 417
X. MEASUREMENTS
10.1 INTRODUCTION . • . • . . . . . • • . . • . . . . . . • . . • . . . . . . . . . . . • • • • • • 423
CONTENTS xv
CHAPTER PAGE
3. Signal Sound Level .......................... 466
••.•...•••.•..•....•••.........•....•.•••
11.2 DISPERSION OF SOUND • .. • •••••..••...••....••••.•....•• 500
XVI CONTENTS
CHAPTER PAGE
C. Functional Sound Absorbers ....................... 506
CONTENTS xvii
CHAPTER PAGE
12.12 MODULATION (VIBRATO) • • . • . • . . . • . . . . . . • . • . . • • . . • . . • . . . 574
MUSIC . • • . . . • . . . • . . . • . . . . . . . . . . . . . . • • . . . . . . . • . . . . . • . • . 579
MUSIC • • • • . . . . • . • . . . • . • . • • . • • . . . . . . . . • . . . . • . • • . . . . . • • • 601
XVlll CONTENTS
CHAPTER PAGE
B. Stereophonic Magnetic Tape Sound Reproducing Sys
tern ..... . .................. . ....... . .... . ........ 636
INFORMATION . • . . . . . • . . . • . . . . . . . . . . . . . . • . . • . . . . . • . . . • • 658
CONTENTS XIX
CHAPTER PAGE
14.30 LIST OF UNDEVELOPED SYSTEMS FOR THE TRANSMISSION OF
INFORMATION 666
XVI. ULTRASONICS
16.1 INTRODUCTION . . . • . . . . . . • . . . . . . . . • . . . . . . . . . . . . • . . . • • • . . 692
INDEX 705
r'"
I
SOUND WAVES
craft and obstacles over great distances, in depth sounders, and in other
acoustic applications in undersea communication.
The industrial applications of ultrasonics have unfolded a new field in
acoustics. Some of the important ultrasonic developments include the
cleaning of machined parts, drilling, and flaw detection.
The science of architectural acoustics has advanced to the point where
auditoriums, studios, and rooms can be designed to obtain excellent acoustics
under severe artistic conditions.
With ever-increasing industrial expansion comes an increase in noise.
Work is now under way actively to control noise by the use of a variety of
acoustic countermeasures.
The preceding brief introduction to the present status of acoustics shows
that it plays a very important part in our modern civilization. Further
more, the fundamentals and applications of the science of acoustics are so
well formulated and substantiated that a large area of the field of acoustics
has attained an engineering status.
In this book the author has attempted to outline the essentials of modern
acoustics from the standpoint of the engineer or applied scientist. It has
been the aim and purpose to make the book as complete as possible by
covering all the major aspects of modern acoustics as outlined in the
preceding text of the introduction. In order to cover a wide range of
readers, the book has been written and illustrated so that the derivations
may be taken for granted. The concepts of mechanical and acoustical
impedance have been introduced and applied so that anyone who is familiar
with electrical circuits will be able to analyze the action of vibrating systems.
1.2. Sound Waves.-Sound is an alteration in pressure, particle dis
placement or particle velocity propagated in an elastic material or the
superposition of such propagated alterations.
Sound is also the sensation produced through the ear by the alterations
described above.
Sound is produced when air is set into vibration by any means whatso
ever, but sound is usually produced by some vibrating object which is in
contact with the air. If a string, such as one used in a banjo or similar in
strument, is stretched between two solid supports and plucked, sound is
produced which dies down in a fairly short time. When the string is
plucked it tends to spring back into its rest position, but due to its weight
(mass) and speed (velocity) it goes beyond its normal position of rest.
Then, in returning it again goes beyond its normal position of rest. The
excursions become smaller and smaller and finally the string comes to rest.
As the string moves forward it pushes air before it and compresses it, while
air rushes in to fill the space left behind the moving string. In this way
air is set in motion. Since air is an elastic medium, the disturbed portion
transmits its motion to the surrounding air so that the disturbance is propa
gated in all directions from the source of disturbance.
If the string is connected in some way to a diaphragm such as a stretched
drumhead of a banjo, the motion is transmitted to the drum. The drum,
r
i ~ SOUND WAVES
___________________________________________________________
having a large area exposed to the air, sets a greater volume of air in motion
and a much louder sound is produced.
3
PISTON
BAFFLE BOARD
the cycle under consideration, is the frequency. The unit is the cycle per
second.
Fundamental Frequency.-A fundamental frequency is the lowest com
ponent frequency of a periodic wave or quantity.
Harmonic.-A harmonic is a component of a periodic wave or quantity
having a frequency which is an integral multiple of the fundamental fre
quency. For example, a component, the frequency of which is twice the
fundamental frequency, is called the second harmonic.
Subharmonic.-A subharmonic is a component of a complex wave having
a frequency which is an integral submultiple of the basic frequency.
Wavelength.-The wavelength of a periodic wave in an isotropic medium
is the perpendicular distance between two wave fronts in which the dis
placements have a phase difference of one complete cycle.
Octave.-An octave is the interval between two frequencies having a
ratio of two to one.
Transducer.-A transducer is a device by means of which energy may
flow from one or more transmission systems to one or more other trans
mission systems. The energy transmitted by these systems may be of
any form (for example, it may be electrical, mechanical, or acoustical) and
it may be the same form or different forms in the various input and out
put systems.
The example of Fig. 1.1 has shown graphically some of the properties
of wave motion. It is the purpose of the next section to derive the funda
mental wave equation. It is not necessary that the reader digest all
the assumptions and processes involved in order to obtain valuable infor
mation concerning the properties of a sound wave.
1.3. Acoustical Wave Equation.-The general case of sound propaga
tion involves three dimensions. The general relation for sound propaga
tion of small amplitudes in three dimensions will be derived and then these
relations will be applied to special problems.
A. Equation of Continuity.-The fundamental equation of hydrokinetics
is the equation of continuity. This equation is merely a mathematical
statement of an otherwise obvious fact that matter is neither created nor
destroyed in the interior of the medium. That is, the amount of matter
which enters the boundaries of a.small volume equals the increase of matter
inside. Consider the influx and efflux through each pair of faces of the
cube of dimensions ~x, ~y, and ~z, the difference between the latter and
the former for the whole cube is
( po' - opo'
ox 2
DoX) Doy Doz and ( Po' + opo'
ox 2
DoX)DoY Doz
Po' =
Po
(i)Y
p
= (1 + s)y = 1 + ys 1.7
~ + : + :; + ~~ = 0 1.10
The type of motion to be considered is irrotational, that is Curl V uvw =
o. That is a necessary and sufficient condition for the existence of a
scalar velocity potential 7> which is defined as
07>
u = -, v = -,
07> W = -07> 1.11
ox oy oz
or
V uvw = Grad 7>
SOUND WAVES 7
1.13
where P = excess pressure.
From equations 1.9, 1.10, 1.11, and 1.13
f) 2e/> YPo(f)2e/> 82e/> 82e/»_
8t2 - P 8x2 + 8y2 + 8z2 - 0 1.14
SECOND, AND THE SPECIFIC ACOUSTICAL RESISTANCE pC, IN GRAMS PER SECOND PER
SQUARE CENTIMETER
METALS
Substance Q a 0 C pC
8 ACOUSTICAL ENGINEERING
METALS (continued)
Substance Q
I
u p c pC
ALLOYS
CERAMICS, ROCKS
Substance Q u p c pC
Cellulose Acetate,
Cellulose Acetate,
Cellulose Acetate,
Methyl Metha-Crylate
Methyl Metha-Crylate
Phenol-Formaldehyde
Phenol-Formaldehyde
Phenol-Formaldehyde
Phenol-Formaldehyde
LIQUIDS
GASES
10 ACOUSTICAL ENGINEERING
~; = - c2s 1.19
s = -
8e
8x 1.25
Using equations 1.6, 1.9, and 1.13, equation 1.33 may be written,
which is the same as equation 1.28. The solution of equation 1.35 for
diverging waves is
r/> = A- €1k(ct-r) 1.36
r
From equations 1.19 and 1.36 the condensation is given by
80 -- "-or--......
<J)
!oJ
t.J
cr 60
"o
t.J
z 50
t.J
i3 40
70
'" f\
\
\
\ I
z
«
t.J
<J)
30
\
«
I
Q. 20 "
10
o
" I'-... ......
r-
.01 .02 .04 .08.1 .2 .8 I 2 4 8 10 20 40 80100
kr
FIG. 1.2. Phase angle between the pressure and particle velocity in a spherical sound
wave in t erms of kr. where k = 21T/11, 11 = wavelength and r = distance from the source.
P= _jkcAp €1k(ct-r)
1.39
r
--
SOUND WAVES 13
Retaining the real part of equation 1.39 the pressure is
p = p kcA sin k(ct - r)
1.40
r
B. Particle Velocity in a Spherical Sound Wave.-The particle velocity
from equations 1.11 and 1.36 is
40
1\ 32
30 \\ 3o
28
20
\ 26
U
Q.J
x
\ 1\ 24
22
0 2o
Q 19
~
IX 8
'\ I8
7 \.
6 \. I
5 I4
4 \ I2
3 \ Io
2 1\ 8
6
'\ 4
5
2
i"... 2
r--
1••001 0
.02 .04 .08.1 .2 .4 .81 2 4 8 10
kr
FIG. 1.3. Ratio of the absolute magnitude of the particle velocity to the
pressure in a spherical sound wave in terms of kr, where k = 2,,/)..,
).. = wavelength and r = distance from the source.
14 ACOUSTICAL ENGINEERING
Comparing equation 1.43 with equation 1.40 for the pressure it will be
seen that the phase angle between the pressure and velocity in a spherical
wave is given by
8 = tan-1 ~ 1.44
kr
For very large values of kr, that is, plane waves, the pressure and particle
velocity are in phase. The phase angle as a function of kr is depicted in
Fig. 1.2.
D. Ratio of the Absolute Magnitudes of the Particle Velocity and the Pres
sure in a Spherical Sound Wave.-From equations 1.40 and 1.43 the ratio
of the absolute value of the particle velocity to the absolute value of the
pressure is given by
·
R a t 10 vI + k 2r2
= --;.,- 1.45
pckr
The ratio in equation 1.45, as a function of kr, is depicted in Fig. 1.3.
1.6. Stationary Sound Waves.-Stationary waves are the wave system
resulting from the interference of waves of the same frequencies and are
characterized by the existence of nodes or partial nodes.
Consider two plane waves of equal amplitude traveling in opposite direc
tions; the velocity potential may be expressed as
g, = A [cos k(ct - x) + cos k(ct + x)] 1.46
The pressure in this wave system from equations 1.19 and 1.21 is
Equations 1.48 and 1.51 show that the maxima of the particle velocity
and pressure are separated by a quarter wavelength. The maxima of p
and u differ by 90° in time phase.
A stationary wave system is produced by the reflection of a plane wave
by an infinite wall normal to the direction of propagation. This is the
simplest type of standing wave system. Complex stationary wave sys
tems are produced when a sound source operates in a room due to the re
flections from the walls, ceiling, and floor.
SOUND WAVES 15
1.7. Sound Energy Density.-Sound energy density is the sound energy
per unit volume. The unit is the erg per cubic centimeter.
The sound energy density in a plane wave is
E = p2 1.52
pc 2
p = (y + I)E 1.53
where E = energy density of the incident wave train in ergs per cubic
centimeter, and
y = ratio of specific heats, 1.4 for air.
Instruments for measuring the radiation pressure have been built, con
sisting of a light piston mounted in a large wall with means for measuring
the force on the piston. Since the radiation pressure is very small these
instruments must be quite delicate.
When the conditions are such that ratios of currents or ratios of voltages
(or the analogous quantities such as pressures, volume currents, forces, and
particle velocities) are the square roots of the corresponding power ratios,
the number of decibels by which the corresponding powers differ is ex
pressed by the following formulas:
1 0 1 0
2 3.0 2 6.0
3 4.S 3 9.5
4 6.0 4 12.0
5 7.0 5 14.0
6 7.8 6 15.6
7 8.5 7 16 .9
8 9.0 8 lS.1
9 9.5 9 19.1
10 10 10 20
100 20 100 40
1000 30 1000 60
10,000 40 10,000 SO
100,000 SO 100,000 100
1,000,000 60 1,000,000 120
SOUND WAVES 17
1.10. Doppler Effect.1-The change in pitch of a sound due to the rela
tive motion of the source and observer is termed the Doppler Effect.
When the source and observer are approaching each other the pitch ob
served by the listener is higher than the actual frequency of the sound
source. If the source and observer are receding from each other the pitch
is lower.
The frequency at the observation point is
v - vo
fo = --fs
v -
Vs
1.58
Equation 1.59 shows that the wind does not produce any change in pitch
unless there is some relative motion of the sound source and the observer.
oo 1--------~
0 0 ~
~
1--'~WAVE
--.,. '
~
OJC 0 DISKS
~lo----l
SERPENTINE PLATES
C] (J ~~
llJb II 1_~
~
6.15 and 8.6B. Practical systems have been developed based upon obstacle
arrays and path length devices.
Obstacle arrays increase the effective density of the medium and thus
produce a reduced propagation velocity of sound waves passing through
the array. Three different obstacle arrays are shown in Fig. 1.5.
SOUND WAVES 19
The index of refraction n, of a spherical obstacle array as shown in
Fig. 1.5, is given by
n2 = 1 +
i7Ta3N 1.60
where a = radius of the sphere, and
N = number of spheres per unit volume.
The index of refraction n, of a disk obstacle array as shown in Fig. 1.5,
is given by
+
n 2 = 1 le 3N 1.61
where c = radius of the disk, and
N = number of disks per unit volume.
The index of refraction n, of a series of strips as shown in Fig. 1.5, is
given by
1.62
where b = half breadth of the strip normal to the direction of propagation
of the wave and
N = number of strips per unit area viewed endwise.
Path length devices increase the time of travel of the waves through the
path over that in free space. Three different path length devices are shown
in Fig. 1.5.
The index of refraction n, of parallel plates as shown in Fig. 1.5, is unity.
The index of refraction n, of the serpentine plates as shown in Fig. 1.5,
is given by
1
n=- 1.63
10
where 1 = path length through the plates, and
10= path in the absence of the plates.
The index of refraction n, of slant plates as shown in Fig. 1.5, is given by
1 1
n=-=-- 1.64
10 cos ()
where () = angle between the direction of propagation of the wave and
the plane of the plates.
In the examples in this book the serpentine system will be used.
An acoustic lens which converges the impinging sound wave is shown
in Fig. 1.6A. The sheet metal is arranged so that the path length through
the lens is the greatest at the center of the lens. The action of the lens is
depicted by the ray and wavefront diagram of Fig. 1.6A. The path lengths
of all the pencils of the incident sound wave are all the same at focus. An
acoustic lens which diverges the incident sound wave is shown in Fig. 1.6B.
The action of the lens is depicted by the ray diagram and wavefront diagram
of Fig. 1.6B.
20 ACOUSTICAL ENGINEERING
INCIDENT SOUND
~
SHEET _ _ ~
METAL ---....-- _
i ~
:f-------~----------7FOCUS
~ ~
~
~
-----
A
INCIDENT SOUND ~
SHEET~
METAL .
./"
An acoustic prism is shown in Fig. 1.7. The acoustic prism changes the
direction of the impinging sound wave. The action of the prism is depicted
by the ray and wavefront diagrams of Fig. 1.7.
Diffraction is the change in direction of propagation of sound due to
the passage of sound around an obstacle. It is well known that sound
will travel around an obstacle. The larger the ratio of the wavelength
to the dimensions of the obstacle the greater the diffraction. The dif
fraction around the head is important in both speaking and listening.
The diffraction of sound by microphones and loudspeakers is important
in the performance of these instruments. The diffraction4 of sound by a
sphere, a cube, and a cylinder as a function of the dimensions is shown in
Fig. 1.8. These data may be used to predict the diffraction of sound by
objects of these general shapes. As, for example, the sphere may be used
to predict the diffraction of sound by the human head.
4 Muller, Black, and Dunn, Jour. Acous. Soc. Amer., Vol. 10, No. 1, p. 6, 1938.
SOUND WAVES 21
~~~~r~
0~~
FIG. 1.7.
-
Acoustic prism.
10~~~
o
I"
CYLINDER
\1} \1 \
~::or 3d'
""'"""I"'" I 90. 60·
'T'
180"
~
10
SPHERE ,=0·
30
5 60
o l's"d
-5
()
/::J.ill DIRECTION or
PLANE WAVE I"-" '\
120
-IS
.02 .04 .06 0.1 0.2
I
0.4 0.6 1.0 2.0 4.0 6.0
D/>
FIG. 1.8. The diffraction of a sound wave by a cylinder, cube,
and sphere. (After Muller, Black and Dunn.)
22 ACOUSTICAL ENGINEERING
There are other shapes 5 besides the cylinder, cube, and sphere that are
used for microphone and loudspeaker enclosures. In order to provide
additional information, the diffraction of sound by the shapes shown in
Fig. 1.9 were obtained experimentally. The dimensions of the ten different
SPHERE HEMISPHERE
CYLINDER CYLINDER
A
_ERE 0 B HEMISPHERE 0
.
_iHffiff8
ffi
5
o
0
~'
•
...... ......
5
0 600 10vu 2000 4
FREQUENCY FREQUENCY
c
CYLINDER [J D
CYLINOfR
i~glrut IPJ+I~I±
100 200 300 800 1000 2000 4000 100 ZOO 300 800 1000 2000 4000
0
FREQUENCY FREQUENCY
I¥JOOFt·~ j:jHWhffi
100 200 300 600 1000 2000 4000 loa 2OO!OO 600 1000 2000 4000
III10"]HHH. iEHHJEt[
See Sec. 1.13. For example, if the linear dimensions of the new enclosure
are two times that of Fig. 1.9, the frequency scale of Fig. 1.10 should be
multiplied by one-half.
Another example of diffraction of sound is illustrated by the zone plate
shown in Fig. 1.11. The path lengths of the sound from the source to the
focus vary by an integral wavelength. As a consequence, all the pencils
of sound are in phase at the focus with the result that the sound pressure
is considerably greater at this point than any other position behind the
zone plate.
~
m
121
FRONT VIEW CROSS - SECTIONAL VIEW
FIG. 1.11. Zone plate. The source 5 and the focus F are equidistant from the zone
plate.
1"
-vgradp' 1,
o--vgradp"
0= 0 1.66
p p
where the surface integral is taken over the boundaries of the volume.
In the simple case in which there are only two pressures, as illustrated
in the free field acoustical system of Fig. 1.12, equation 1.71 becomes
p'v" = P"v' 1.72
where p', P" and v', v" are the pressures and particle velocities depicted
in the free field acoustical system of Fig. 1.12.
1??2?? uun anna???? ?222222UU22222222222222?222Zd
ox + og = ox + :! ox 1.75
F=QS~
oX 1.77
F + SF = QS ~+~
ox Sg) Sx
ax (QS ox 1.78
c=J~
1.83
The velocity of sound, Young's modulus and the density for vanous
solids are given in Table 1.1.
1.15. Torsional Waves in a Rod.-A rod may be twisted about an
axis of the rod in such a manner that each transverse section remains in its
own plane. If the section is not circular there will be motion parallel to
the axis of the bar. For a circular cross section and a homogeneous bar the
equations of motion are analogous to those of longitudinal waves in the rod.
The velocity of propagation, in centimeters per second, of torsional waves
in a rod, is
c-J
- 2p (0'
Q+ 1) 1.84
12 Morse, "Vibration and Sound," McGraw-Hill Book Company, New York, N.Y.,
1948.
SOUND WAVES 29
The particle velocity u, in centimeters per second, under the same con
ditions is given by
U = 7TaU o JZ
cr
ff.i k (r-ct)-i7T/4 1.87
It will be seen that the pressure and particle velocity decrease inversely
as the square root of the distance from the cylinder.
The product of the pressure and the particle velocity gives the flow of
energy per square centimeter as follows,
1.88
It will be seen that the intensity falls off inversely as the distance.
2
The total average power in ergs per second emitted by the source is
PT = 217pck2A2 2.6
where p = density of the medium, in grams per cubic centimeter,
c = velocity of sound, in centimeters per second,
k = 217/A,
A= wavelength, in centimeters, and
A is defined by equation 2.1.
B. Point Source Radiating into a Semi-Infinite Medium. Solid Angle
of 21T Steradians.-The above example considered a point source operating
in an infinite medium. The next problem of interest is that of a point
source operating in a semi-infinite medium, for example, a point source
near an infinite wall.
In this case we can employ the principle of images as shown in Fig. 2.1.
The pressure, assuming the same distance from the source, is two times
that of the infinite medium. The particle velocity is also two times that
of the infinite medium. The average power transmitted through a unit
area is four times that of the infinite medium. The average power out
put of the source, however, is two times that of a simple source operating
in an infinite medium.
C. Point Source Radiating into a Solid Angle of 17 Steradians.-Em
ploying the method of images Fig. 2.1 the pressure is four times, the par
ticle velocity is four times, and the average power transmitted through a
unit area is sixteen times that of an infinite medium for the same distance.
The average power output of the source is four times that of a simple
source operating in an infinite medium.
D. Point Source Radiating into a Solid Angle of 17/2 Steradians.-Em
ploying the method of images, Fig. 2.1, the pressure is eight times, the
particle velocity eight times, and the average power transmitted through a
unit area is sixty-four times that of the same source operating in an infinite
medium at the same distance. The average power output is eight times
that of the same simple source operating in an infinite medium.
E. Application of the Simple Source.-The above data may be applied
to acoustic radiators in which the dimensions are small compared to the
wavelength and located close to the boundaries indicated above. For
example, A would correspond to a loudspeaker, which acts as a simple
source, suspended in space at a large distance from any walls or boundaries.
B would correspond to a loudspeaker placed on the floor in the center of
the room. C would correspond to a loudspeaker placed on the floor along
a wall, and D would correspond to a loudspeaker placed in the comer of
32 ACOUSTICAL ENGINEERING
the room. Of course, as pointed out above, these examples hold only when
the dimensions of the radiator and the distance from the wall are small
compared to the wavelength.
SOLID ANGLE PRESSURE POWER ENERGY
OF SOUND AT A OUTPUT DENSITY
EMISSION DISTANCE r DISTANCEr
•
SOURCE
41T P W
,.~+~, 2Tf 2p 2W 41
dF-
IMAGES
1T 4p 4W 161
SOURCE
1! 8p 8W 641
2
IMAGES
FIG. 2.1. The sound pressure, total power output, and energy
density delivered by a point source operating in solid angles of
47T, 27T, 7T, and 7T/2 steradians.
(!+ jk)A
cfo = r €jk(ct-r) cos a 2.7
r
1 Lamb, "Dynamical Theory of Sound," E. Arnold, London, 1931.
2 Davis, "Modern Acoustics," The Macmillan Co., New York, N.Y., 1934.
P= _ p 8e/>
8t
= _ j pckA
r
(!r + jk) €jk(ct-r) cos a 2.8
The particle velocity has two components, the radial ~t and the trans
verse !r 88e/>·a The radial component of the particle velocity from equation
2.7 is,
u= ~= - [(~ +~~) +jk (h +~)] A €1k(ct-r) cos a 2.12
U = -
18
j = - (_;1+_ _J.k) A€1k(ct-r) sin a 2.16
r 8a r2
Retaining the real parts of equation 2.16
Fig. 2.2 shows the velocity components and the pressure for various
points around a doublet source. A common example of a doublet source
is a direct radiator loudspeaker mounted in a small baffle. (Dimensions
of the baffle are small compared to the
~----
, ~,
/
wavelength.) If the response of such a
loudspeaker is measured with a pressure
I
.I , "" microphone for various angles at a con
~\lIi"/PARTICLE stant distance, the result will be a cosine
characteristic. If the response is measured
"' /
I .,/// '
• ./""'\ VELOCITY
2.20
FIG. 2.2. The sound pressure and
particle velocity at a constant dis where p = pressure, in dynes per square
tance from a doublet source. The centimeter,
PT = 27T1'2 i'"
o
pCk4A2
- - - cos 2 e;; sin e;; de;;
2r2
2.21
PT = i1Tpck4A2 2.22
where p = density, in grams per cubic centimeter,
k = 217/1..,
FIG. 2.3. Directional characteristics of two separated equal small sources vibrating in
phase as a function of the distance between the sources and the wavelength. The polar
graph depicts the sound pressure, at a fixed distance, as a function of the angle. The
sound pressure for the angle 0° is arbitrarily chosen as unity. The direction correspond
ing to the angle 0° is perpendicular to the line joining the two sources. The directional
characteristics in three dimensions are surfaces of revolution about the line joining the
two sources as an axis.
. (nwd. )
sm TsmlX
Ra = 2.23
. (wd.
nsm TsmlX )
~Wolff, I., and Malter, L., Jour. Acous. Soc. Amer., Vol. 2, No.2., p. 201, 1930.
8 Stenzel, H., Elek. Nach. Tech., Vol. 4, No.6, p. 239, 1927.
7 Stenzel, H., Elek. Nach. Tech., Vol. 6, No.5, p. 165, 1929.
36 ACOUSTICAL ENGINEERING
the limiting case IS the line source. If this is carried out, equation 2.23
becomes
. (TTl.
sm Xsma )
Ra = ----;.-----=
TTl .
2.24
1" sma
iII 1i • 3.'.0
LENG1tt.fA30
z .•
LENGTH-A
z
~0
.6
30
LENGT~_'~A
• 30
LENGTH-3A
30
a
60
. 60
00 80
60 eo 60 00
90 . 90
90 vee 90 90 90 90 90
•
oLENGTH.8A30 eoLENGTI;l=4\0
LENGT'~_'tA LENGTH. 2> .LENGT~-
4>
~ we 30 .e 30 3 L1.I 30 3 30
Z ,6 zo
•
60 4 60
60 80 60 Z eo 60 00
'm'O'0 .'.00
to
,0LENGTH-lf~
t09 90 90 1,0 90 90
LENGTH-~ LENGTH- 8~ LENGTH .2~" LENGTH -6>
z',:'O ~O
~~: '0 . z•
:o 30
~Q
• ~
•
~A
eo
60 eo 60 60 eo 60
eo ~ ~ eo 90 to 90 90 90 90
FIG. 2.4. Directional characteristics of a line source as a function of the length and the
wavelength. The polar graph depicts the sound pressure, at a large fixed distance, as a
function of the angle. The sound pressure for the angle 0° is arbitrarily chosen as unity.
The direction corresponding to the angle 0 0 is perpendicular to the line. The directional
characteristics in three dimensions are surfaces of revolution about the line as an axis.
SOUNO
SOURCES
DELAY
UNITS
INPUT
sin 2 (~ sin a)
2.27
Ra = (l
;;>.. sin a
)2
where Ra = ratio of the pressure for an angle a to the pressure for an
angle a = O. The direction a = 0 is normal to the line,
1= total length of the line in centimeters, and
,\ = wavelength, in centimeters.
The directional characteristics of a tapered line source are shown in
Fig. 2.6. Comparing the directional characteristics of Fig. 2.6 with those
of the uniform line of Fig. 2.4, it will be seen that the main lobe is broader
and the secondary lobes are reduced in amplitude.
8 Menges, Karl, Akus Zeit., Vol. 6, No. 2, p . 90,1941.
38
.~ • "'~~'i'
ACOUSTICAL ENGINEERING
",o,,~., •• If
LENG H: >- to.oL]~~~~~lIJ_,C;~~~:;t:l>-.O ''rENGTH:2X'° 90LENGTH:4X'"
0
8
I~ 3~
to,ll:1~~~~~:IJ .~
.0
30_
'LENGTH:6>
'0
.0 901 '0 10 90
FIG. 2.6. Directional characteristics of a tapered line source as a function of the length
and the wavelength. The volume current output along the line varies linearly from a
maximum at the center to zero at the two ends. The polar graph depicts the sound
pressure, at a fixed distance, as a function of the angle. The sound pressure for the
angle 0° is arbitrarily chosen as unity. The direction corresponding to the angle 0° is
perpendicular to the line. The directional characteristics in three dimensions are sur
faces of revolution about the line as an axis.
2.28
'80
.80
LENGTH' t LENGTH. i LENGTH .).
'.0
.10
LENGTH '2). LENGTH' 4).
10 Pritchard. R. L.. Jour. Acous. Soc. Amer., Vol. 25, No.5, p. 879, 1953.
40 ACOUSTICAL ENGINEERING
is shown in Fig. 2.8. It will be seen that alternate elements are oppositely
phased. Comparing the directional pattern of the super directivity source
with the directivity pattern of the simple line source of Fig. 2.4, it will be
seen that approximately the same
30
directivity is obtained with a line of
20
one-third the length of the simple line
II)
source. This added directivity is
o
obtained at the expense of some
...o~ 10 other factors. The reversed phase
..
::>
::; 0
0.
excitation results in a loss in
:<
efficiency. Close tolerances must be
"-10 maintained upon the strength of the
elements and the operating wave
~2~ELATIVE AMPLITUDES length, otherwise the directivity pat
tern will not be maintained. Thus
FIG. 2.8. Directional characteristics of a
super directivity source consisting of five a super directional system is sensitive
sources spaced 1 wavelength apart. The to frequency changes and is, there
relative amplitudes and the phases of the fore, not suitable for broad band
sources are shown in the diagrams above.
The polar graph depicts the sound pressure operation. Therefore, the applica
at a large fixed distance from the array. tions for a super directional source
The sound pressure for the angle 0° is are where a narrow frequency band
arbitrarily chosen as unity. The direc
tion corresponding to the angle 0° is width and low efficiency can be
perpendicular to the line joining the tolerated in exchange for smaller
sources. The directional characteristics
in three dimensions are surfaces of revolu space requirements.
tion about the line as an axis. 2.11. Curved-Line Source (Arc
of a Circle).-A curved-line source
may be made up of a large number of point sources vibrating in phase on the
arc of a circle separated by very small distances. The directional charac
teristics of such a)ine in the plane of the arc are,
A= wavelength, in centimeters,
k = variable,
RADIUS. ; RADIUS- ).
to 1.0
Ll.LL.L:-::::£.~!f:3:::::::LL...l.JUoo 90LLL::£~~I:::::L--L-.J90
RADIUS" 2), RADIUS- 4), RADIUS - 8), RADIUS -18).
FIG. 2.9. Directional characteristics of a 60° arc as a function of the radius and the
wavelength. The polar graph depicts the sound pressure, at a large fixed distance, as a
function of the angle in the plane of the arc. The sound pressure for the angle 0° is
arbitrarily chosen as unity.
The directional characteristics for an arc of 60°, 90°, and 120° are shown
in Figs. 2.9, 2.10, and 2.11. The interesting feature of the directional
11 Wolff, I., and Malter, L., Jour. Acous. Soc. Amer., Vol. 2, No.2, p. 201, 1930.
42 ACOUSTICAL ENGINEERING
RADIUS
1.0
·i RADIUS-i
1.0
RADIUS->"
•.0
FIG. 2.10. Directional characteristics of a 90° arc as a function of the radius and the
wavelength. The polar graph depicts the sound pressure at a large fixed distance, as a
function of the angle in the plane of the arc. The sound pressure for the angle 0° is
arbitrarily chosen as unity.
'0
RADIUS=~
o
60
'II
60
:"
•
RADIUS=~
go
60
RADIUS-'>'
•
FIG. 2.11. Directional characteristics of a 120° arc as a function of the radius and the
wavelength. The polar graph depicts the sound pressure, at a large fixed distance, as a
function of the angle in the plane of the arc. The sound pressure for the angle 0° is
arbitrarily chosen as unity.
ACOUSTICAL RADIATING SYSTEMS 43
2.12. Circular-Rin~ Source.-The directional characteristics l2 ,13 of a
circular-ring source of uniform strength and the same phase at all points on
the ring is
2.32
-t DIAME1~R -ll>.
•
DIAMET.~R
•
5
R -3>.
JlDIAMET
30 30
z .• ..
..
:
60 ., eo
a
DIAMET~R_l
DIAMET~R_12~
• •
3 w.s 30 3 30 3 0
•
.. A ..,
Z
0.
.0 00 60 10 eo eo
gO gO 90 '10 90 '1090 90
parts of the surface of the piston vibrating with the same strength and phase
are
2lI (2;R sin IX)
Ra = 271'R. 2.33
-A- SIll IX
3$
00 .:~.o 'li1.0 30
80 60
LJ~~~~;;r:Jw
B\iY
00 00
•
00 ~4
00 00 ZA 00 .:W:
90
&0
.'.~ ]I
90
80
eo
•
en.4
go
·00
; 30 31i'.030
981AMETER
60
-~>.
60
·~IAMETER.6~
60 80
90 liO 90 ao 90 IKt 80 90
180· 180" I O·
DIAMETER· -} DIAMETER. -} DIAMETER. >.
FIG. 2.14. Directional characteristics of a circular-piston source located in the
end of an infinite pipe as a function of the diameter and wavelength. The
polar graph depicts the sound pressure, at a large fixed distance, as a function
of the angle. The sound pressure for the angle 0° is arbitrarily chosen as
unity. The direction corresponding to the angle 0° is the axis. The axis is
the center line perpendicular to the plane of the piston. The directional
characteristics in three dimensions are surfaces of revolution about the axis.
same as that of a uniform line source having a length equal to one side of
the square (equation 2.24).
The directional characteristics of a plane square-surface source, with all
parts of the surface vibrating with the same strength and phase, in a normal
plane containing the diagonal is the same as that of the tapered line source
having a length equal to the diagonal (equation 2.28).
180·
DIAMETER' t 180"
DIAMETER·
A
"2
180·
DIAMETER' A
H+8Jl8+f--+ 0'
SIll T a S111
. (7Tl . a)
Ra = ---:---..:.. 2.34
7Tla .
T SIlla
characteristics are broader than those obtained from a piston the size of the
mouth. From another point of view, the diameter of the piston which will
yield the same directional characteristic is smaller than the mouth. These
results also show that the directional characteristics vary very slowly with
frequency at these smaller wavelengths. Referring to Fig. 2.17 it will be
seen that for any particular high frequency, 4000, 7000, or 10,000 cycles per
second, the directional characteristics become progressively sharper as the
rate of flare decreases .
'. ",0 ,. 30 . .
:0.IJ"
oe
.
45
Ii'. z
°8 I. . ,
0 .
I.
.e
• .. ° .5
IS °
FIG. 2.16. The directional characteristics of a group of exponential horns, with a con
stant flare and throat diameter of t inch as a function of the mouth diameter. The
number at the right of each polar diagram indicates the diameter of a circular piston
which will yield the same directional characteristic. The polar graph depicts the sound
pressure, at a fixed distance. as a function of the angle. The sound pressure for the
angle 0° is arbitrarily chosen as unity. The direction corresponding to 0° is the axis
of the horn. The directional characteristics in three dimensions are surfaces of revolu
tion about the horn axis.
B. Conical Horns.-In the case of the circular conical horn the direc
tional pattern should be the same as that of a circular, spherical surface
source. The radius of the spherical surface is the distance along the side
of the horn from the apex to the mouth. The directional characteristics of
two conical horns are shown in Fig. 2.18. At the lower frequencies the
directional pattern is approximately the same as that of a piston of the same
size as the mouth. The directional pattern becomes sharper with an increase
of the frequency. However, at the higher frequencies where the diameter
of the mouth is several wavelengths, the pattern becomes broader as would
be expected from a spherical surface source. The directional characteristics
of a conical horn as depicted in Fig. 2.18 are practically the same as those of
a spherical surface source.
C. Parabolic Horns.-In the parabolic horn the sectional area is pro
portional to the distance from the apex. This horn may be constructed
ACOUSTICAL RADIATING SYSTEMS 49
~?1" "~• •
~3~\.5r~~a~t}\~~JO
:.=.=
Sll" _
:~. ~. ~. ~.
8
8
,,"
FIG. 2.18. The directional characteristics of two conical horns with mouth
diameters of 12 inches and throat diameters of %inch and lengths of 12 inches and
24 inches. The polar graph depicts the sound pressure, at a fixed distance, as a
function of the angle. The sound pressure for the angle 0° is arbitrarily chosen as
unity. The direction corresponding to 0° is the axis of the horn. The directional
characteristics in three dimensions are surfaces of revolution about the horn axis.
as shown in Fig. 2.19 in which two opposite horn walls are parallel and the
other two are inclined at an angle with respect to each other. The direc
tional characteristics of a 90° parabolic horn are shown in Fig. 2.19. The
50 ACOUSTICAL ENGINEERING
23 Wente, E. C., and Thuras, A. L., Jour. A. I. E. E., Vol. 53, No. !, p . 17, 1934.
FRONT
VICW
0 / ~ ..
~: ~
.....
:0:"
a~
° ••
B~ foil- .
PLA.N
VIEW
SIDE
VIEW
Di§~~3::~90
4400"\1
eo
8800'\.1
gO ,ot::.~~~~j
132001'\J 17600ru
R;4)" R ;8>.. R :.12>. R= 18>..
FIG. 2.19. The directional characteristics of a parabolic horn of the shape and the
dimensions shown in the sketches on the left. The patterns were obtained in the plane
midway between and parallel to the two parallel sides. The polar graph depicts the
sound pressure, at a fixed distance, as a function of the angle. The sound pressure for
the angle 0° is arbitrarily chosen as unity. The direction corresponding to 0° is spaced
midway between the two nonparallel sides of the horn. R = 12 inches. The ratio of
RIA is also given for comparison with Fig. 2.10.
TOP VIEW
A,
e,
~----x
z B
FIG. 2.21. A. A spherical radiating surface consisting of 15 individual
exponential horns. B. Geometry for predicting the directional characteristics
of a cluster of small horns.
52 ACOUSTICAL ENGINEERING
Assume that the point of observation is located on the OY axis, Fig. 2.21B,
at a distance several times the length of the horn. The amplitude of the
vector contributed by an individual horn for the angle cp can be deter
mined from its individual directional characteristic. In this illustration,
the plane XOZ is chosen as reference plane for the phase of the vector.
The phase angle of the vector associated with an individual horn is
8= ~ 360° 2.35
where d = the distance between the center of the mouth of the horn and
the reference plane X'O'Z', in centimeters, and
,.\ = wavelength, in centimeters.
The vectors, having amplitudes AI, A 2 , A3, A4, etc., determined from the
directional characteristics and having phase angles 81 , 82 , 83 , 84 , etc., de
termined from equation 2.35, are added vectorially as shown in Fig. 2.21B.
This method of predicting the directional characteristics assumes that
there is no interaction between individual horns which changes the com
plexion of the velocities at the mouth from that which obtains when
operating an individual horn. Obviously, this condition is not absolutely
satisfied. Apparently, the discrepancy has no practical significance because
it has been found that this method of analysis agrees quite well with experi
mental results.
The directional characteristics of the cellular horn of Fig. 2.21A are
shown in Figs. 2.22 and 2.23. Above 2000 cycles the dimensions of the
total mouth surface are several wavelengths and the directional character
istics are fairly uniform and defined by the total angular spread. Where
the dimensions are comparable to the wavelength the directional charac
teristics become very sharp, as shown by the polar curves for 500 and 1000
cycles. Then, as the dimensions of the surface become smaller than the
wavelength, 250 cycles, the angular spread broadens, as is illustrated by
the larger spread for the smaller vertical dimension when compared to the
smaller spread for the larger horizontal dimension.
The directional characteristics of a cellular horn show a striking resem
blance to those of an arc of the same angular spread. For example, the
angular spread of the horn of Fig. 2.21 in the plane containing the line
AA' and the axis is 87!0. This may be compared to the arc of Fig. 2.10.
In this case ,.\/4, "12, "\, 2"\, 4"\, and 8"\ will correspond to 145, 290, 580,
1160, 2320, and 4640 cycles. The angular spread in the plane containing
the line BB' and the axis is 52!0. This may be compared to the 60° arc
of Fig. 2.9 with the same relation between the wavelengths and frequencies,
as noted above. It will be seen that there is a marked resemblance between
corresponding frequencies. Of course, there is some variation due to the
fact that the frequencies do not correspond exactly. Further, there is some
difference in the angular spread. For most spherical surfaces of this type
the directional characteristics in various planes correspond very closely to
the directional characteristics of the corresponding arc.
ACOUSTICAL RADIATING SYSTEMS 53
250N
10
"...
•
60
"B
~
500N
~4
00
eo
e
60
1000N
"O
tt~
00
~
30
•
20001V 40001V 8000N
" ~ .
•0 eo
~ -
shown in Fig. 2.21A in a plane containing the line B-B' and the
axis of the center horn. The polar graph depicts the sound
pressure, at a fixed distance, as a function of the angle. The
sound pressure for the angle 0° is arbitrarily chosen.
E 'ml.
2SON SOON loo01V
o 1.0
~.
'" '111.0
z. 30
O
~.
'"
eo 4 w ~4 ~
60 60 60 60
90 90 90 90 90 90
fraction of the wavelength of sound in the paper. The output of the cone
at any angle is the vector sum of the vectors A o, AI, A2 ... An where the
A's are the amplitudes of the individual rings.
The phase angle of the amplitude of the first ring is
80 = 0 2.36
The phase angle of the amplitude of the second ring is
fh
27T(dl - DI) cos IX
= 2.37
AA Ap
The phase angle of the amplitude of the third ring is
82 = 27T(dl + d2 _
AA
DI +
Ap
D2) cos IX 2.38
Ao
FIG. 2.24. Geometry for obtaining the directional pattern of a cone-type radiator.
where dl, d2 , ••• =axial distances shown in Fig. 2.24 in centimeters, and
Dl, D2, ... =
distances along the cone shown in Fig. 2.24 in centi
meters,
AA = wavelength of sound in air, in centimeters,
Ap = wavelength of the sound in the paper cone, in centi
meters, and
IX = angle between the axis of the cone and the line joining
the observation point and the center of the first ring.
where T" =
An = 27T1'nDnJo e:
The relative amplitude of the vector A n is given by
R _
2: A
K=O
K cos (} K - j
K-O
2: A K sin (} K
a - K-n
2.41
2: A
K=O
K
where Ra = ratio of the pressure for an angle a to the pressure for an angle
a = o.
A consideration of equation 2.41 shows that the directional pattern is
a function of the frequency and becomes sharper as the frequency increases.
For a particular frequency, cone angle, and material the directional patterns
are practically similar for the same ratio of cone diameter to wavelength.
For a particular frequency and the same cone material the directional
pattern becomes broader as the cone angle is made larger. For a particu
lar frequency and cone angle the directional pattern becomes broader as
the velocity of propagation in the material decreases (see Sec. 6.2).
3
/= !..J!
2l m
3.1
-....::::::::::---
·
FIRST OVERTONE SECOND HARMONIC
..............:---~- •
·
SECOND OVERTONE THIRD HARMONIC
::""""'N L N
~
THIRD ~
OVERTONE FOURTH HARMONIC
L L L
N~~~~N
• FOURTH OVERTONE fiFTH HARMONIC
•
N~NL~LN~.NL N
~~~ ~
FI frH OVERTONE SIXTH HARMONIC
plane and, as in the case of the string, only the transverse vibrations will
be considered.
A. Bar Clamped at One End.-Consider a bar clamped in a rigid support
at one end with the other end free (Fig. 3.2A). The fundamental frequency
is given by
/I = .S596JQK 2 3.2
l2 p
riRST
A
--
OVERTONE
"""'"
fUNDAMENTAL
OVERTONE
D E F
----------
......... L
~ ~ ~
' .............................,
fUNDAMENTAL fUNDAMENTAL
fUNDAMENTAL
~
~ ~
fiRST
SECOND
OVERTONE
OVERTONE
fiRST
}""'="'"-~~
OVERTONE
SECOND OVERTONE
~~
FIRST
SECOND
-----------
OVERTONE
.......... ~-
OVERTONE
~
FIG. 3.2. Modes of transverse vibrations of bars. A. A bar clamped at one end and
free at the other. B. A bar clamped at one end and supported at the other. C. A bar
supported at one end and free at the other. D. A bar free at both ends. E. A bar sup
ported at both ends. F. A bar clamped at both ends.
K=~
v'll
where a = thickness of the bar, in centimeters, in the direction of vibration.
For a circular cross section,
1 0 11
2 1 .2165 , 6.26711
3 2 .1321, .4999 17.5511
4 3 .0944, .3558 .. 6439 34.3911
It will be seen that the overtones are not harmonics. The first overtone
of a bar or reed has a higher frequency than the sixth harmonic of a string.
The tuning fork is the most common example of a bar clamped at one end,
because it can be considered to be two vibrating bars clamped at the lower
ends. The overtone or the high-frequency sound of a tuning fork is quickly
damped out leaving almost a pure sound.
B. Bar Free at Both Ends.-Consider a perfectly free bar (Fig. 3.2D).
The fundamental frequency is given by
where 1 = length of the bar, in centimeters. All the other quantities are
the same as in equation 3.2.
The modes of vibration of a perfectly free bar are shown in Fig. 3.2D.
The table which follows gives the position of the nodes· and the frequencies
of the overtones.
1 2 .2242, .7758 11
2 3 .1321. .50, .8679 2.75611
3 4 .0944, .3558 •. 6442.. 9056 5.40411
4 5 .0734.. 277•. 05, .723, .9266 9.93311
60 ACOUSTICAL ENGINEERING
/1 = 2.45jQK2 3.5
12 P
The overtones are
/2 = 3.25/1
fa = 6.75/1
/4 = 11.5/1,
and
Is = 17.7/1
The modes of vibration are shown in Fig. 3.2B.
F. Bar Supported at One End and Free at the Other.-Consider a bar
supported at one end and free at the other (Fig. 3.2C) . The fundamental
frequency is zero. The first overtone is given by
/2 = 2.45jQK2 3.6
12 P
The overtones are
/1 = 0
fa = 3.25/2
/4 = 6.75/2
Is = 11.5/2,
and
/6 = 17.7/2
The modes of vibration are shown in Fig. 3.2C.
G. Tapered Cantilever Bars.-In the preceding, considerations have
been concerned with bars of uniform cross section. It is the purpose of
MECHANICAL VIBRATING SYSTEMS 61
this section to give the formulas for the resonant frequencies of tapered
cantilever bars.
The resonant frequency of a wedge-shaped bar vibrating normal to the
two parallel sides of the wedge, Fig. 3.3A, is
f= li!4ji~; 3.7
where b = thickness of the bar in the direction of vibration, in centimeters.
f= '~25ji~; 3.8
The resonant frequency of a conical bar, Fig. 3.3C, is
f= 1.39jQa 2 3.9
l2 4p
where a = radius of the cone at the base, in centimeters.
3.4 Stretched Membranes. 6 ,7,8,9,lO-The ideal membrane is assumed to
be flexible and very thin in cross section, and stretched in all directions by a
6 Lamb, " Dynamical Theory of Sound," E. Arnold, London, 1931.
8 Morse, "Vibration and Sound," McGraw-Hill Book Company, New York, N.Y.,
1936.
9Wood, " A Text Book of Sound," Bell and Sons, London, 1930.
10Crandall, " Theory of Vibrating Systems and Sound," D. Van Nostrand Company,
Princeton, N.J., 1926.
62 ACOUSTICAL ENGINEERING
The frequency of one nodal circle and two nodal diameters, Fig. 3.4H, is
/22 = 3.50/
The frequency of two nodal circles and one nodal diameter, Fig. 3.41, is
/13 = 4.22/01
The stretched circular membrane is used in the condenser microphone
(see Sec. 8.2B). The fundamental resonance frequency is placed at the
upper limit of the frequency range. A resistive load is coupled to the
diaphragm for damping the response in the neighborhood of the funda
mental resonance frequency. This resistance is incorporated in the back
plate which serves as the stationary electrode.
A stretched circular membrane is also used in all types of drums. In
this case the air enclosure as well as the characteristics of the membrane
controls the modes of vibration.
B. Square Membrane.-The fundamental frequency of a square stretched
membrane is given by
/=.705J7: 3.11
a m
where m =mass, in grams per square centimeter of area,
a =length of a side, in centimeters, and
T = tension, in dynes per centimeter.
C. Rectangular Membrane.-The fundamental frequency of a rectangu
lar stretched membrane with the sides in the ratio of 1 to 2 is given by
-
/ - vab
.7~J!
m 3.12
the purpose of this section to consider the vibration of circular plates for
the various support means of Fig. 3.5.
A. Circular Clamped Plate.-Consider a circular clamped plate as shown
in Fig. 3.5A. The fundamental frequency is given by
.467tJ Q 3.13
]01 = R2 p(1 _ a2)
where t =thickness of the plate, in centimeters,
R =radius of the plate up to the clamping boundary, in centimeters,
p= density, in grams per cubic centimeters (see Table 1.1),
a =Poisson's ratio (see Table 1.1), and
Q= Young's modulus, in dynes per square centimeter (see Table
1.1).
The fundamental frequency is with the circumference as a node and a
maximum displacement at the center (Fig. 3.6A).
The frequency of the next two overtones with nodal circles, Fig. 3.6B
and 3.6C, are,
]02 = 3.91]01
]03 = 8.75]01
The frequencies of the first, second, and third overtones with nodal diame
ters are
]11 = 2.09]01
121 = 3.43]01
!al = 4.95]01
These nodes are shown in Figs. 3.6D, 3.6E, and 3.6F.
Following these simpler forms of vibration are combinations of nodal
circles and nodal diameters. The frequency of one nodal circle and one
nodal diameter, Fig. 3.6G, is
!I2 = 5.98]01
The frequency of one nodal circle and two nodal diameters, Fig. 3.6H, is
122 = 8.74]01
The frequency of two nodal circles and one nodal diameter, Fig. 3.61, is
!I3 = 11.9]01
The clamped plate is used in electromagnetic telephone receivers in which
the steel diaphragm serves as the armature (see Sec. 9.2A). It is used in
carbon microphones (see Sec. 8.2A) . It is used in the subaqueous condenser
microphone (Sec. 15.4) and the magnetic subaqueous loud speaker (sec. 15.6).
Clamped plate diaphragms have been used in miniature condenser micro
phones. The disadvantage of a plate is the difficulty of mounting a thin
plate to give a small mass per unit area for high sensitivity and still have
sufficient stiffness to yield a high fundamental frequency.
MECHANICAL VIBRATING SYSTEMS 65
!(Yo
~ 8
I I
CLAMPED EDGE. SUPPORTED [DGE
C
0 I
D
0
§Sj
The air or water load on the diaphragm can be determined by assuming the
effective radius of the equivalent piston to be .55 times the radius of the
diaphragm (see Sec. 5.8).
B. Circular Free Plate.-Consider a circular plate under no tension,
uniform in cross section and perfectly free (Fig. 3.5D). For a vibration
with nodal circle, as depicted in Fig. 3.4B, the frequency is
.412t) Q
f = R2 p(1 _ a 2) 3.14
where t = thickness of the plate, in centimeters,
R = radius of the plate, in centimeters,
p = density, in grams per cubic centimeter (see Table 1.1),
a = Poisson's ratio (see Table 1.1), and
Q= Young's modulus, in dynes per square centimeter (see Table
1.1).
For a vibration with two nodal diameters, as depicted in Fig. 3.4E, the
frequency is
.193t) Q 3.15
f = R2 (pl - a2)
C. Circular Plate Supported at the Center.-Consider a circular plate under
no tension, uniform in cross section, edges perfectly free and supported at
the center (Fig. 3.5C). The frequency, for the umbrella mode, is
.1ntJ Q 3.16
f = R2 p(1 _ a2)
D. Circular Plate Supported at the Outside.-Consider a plate under no
tension, uniform in cross section, edges simply supported at the periphery
(Fig. 3.SB). The fundamental frequency is
.233tJ Q 3.17
f = R2 p(1 _ a 2)
3.18
- -
--
----
- --
-)
fUNDAMeNTAL F' I RST HARMONIC
L N L N L
(1 --
FIRST
::::
OVERTONE
+
---
~ ~
--
SECOND
:;
HARMONIC
=~)
0=
N
-
=
L
:::
N
. . --
-::
-
~
L
--
N
- - .- :=: ~)
L
The fundamental resonance frequency occurs when the length of the rod
is one-half wavelength. This fact provides a means of computing the
velocity of sound when the density, Young's modulus, and the frequency
are known, or the frequency of sound when the velocity, density, and
Young's modulus are known.
Rods in which the longitudinal waves are excited by striking the ends
are used as standards of high-frequency sounds, 5000 cycles and above,
where a tuning fork is not very satisfactory.
Longitudinal waves in a rod may be set up by electromagnetic, electro
static, or magnetostriction means. In the first case, if the rod is of mag
netic material and is held near an electromagnet in which an alternating
current is flowing a longitudinal force will be set up in the rod. If the
frequency of the driving current is continuously variable, the rod will be
set into violent vibrations at the fundamental resonant frequency. If the
plane end of a rod is placed near a metallic disk, the two plane surfaces
may be used to serve as plates of a condenser. An alternating current sent
through the condenser will cause an alternating force to be exerted upon the
end of the rod. The rod will be sent into violent vibrations when the
frequency of the impressed alternating current corresponds to the funda
mental frequency or one of the overtones. Magnetization of magnetic
materials produces small changes in the dimensions of these materials.
68 ACOUSTICAL ENGINEERING
(]'J~tt'tttt ...
F"UNDAMENTAL F"IRST HARMONIC
L N L N L
L N L N L N L
there will be motion parallel to the axis of the bar. Consider an entirely
free rod of homogeneous material and circular cross section. The simplest
or fundamental mode of torsional vibration occurs when there is a node in
the middle and a loop at each end, that is, when the length of the rod is one
half wavelength. The fundamental resonant frequency, Fig. 3.8, may be
obtained from equation 1.79, as follows
C
11 = X= 2t = 2i
C IJ 2p(a
Q
+ 1) 3.19
where 1 = length of the rod, in centimeters,
p = density, in grams per cubic centimeter (see Table 1.1),
Q = Young's modulus, in dynes per square centimeter (see Table
1.1),
a = Poisson's ratio (see Table 1.1),
C = velocity of propagation of torsional waves, m centimeters per
second, see equation 1.79, and
A = wavelength of the torsional wave, in centimeters.
20 Pierce, G. W., Proc. Am. Acad. Arts and Sci., Vol. 63, p. 1, 1928.
21 Wood, " A Text Book of Sound," Bell and Sons, London, 1930.
22 Rayleigh, "Theory of Sound," Macmillan and Company, London, 1926.
MECHANICAL VIBRATING SYSTEMS 69
The overtones, as in the case of longitudinal vibrations, are harmonics of
the fundamental. That is, h = 2/1, fa = 3/1, /4 = 4/1, etc. The nodes
and antinodes for the various harmonics are formed as in the case of longi
tudinal vibrations.
Torsional vibrations may be set up in bars by any means which applies
tangential forces to the free end. From a comparison of the longitudinal
and torsional vibrations in the same bar, Poisson's ratio may be determined.
3.8. Open and Closed Pipes.-The vibrations of a column of gas or
fluid in a cylindrical tube are analogous to the longitudinal vibrations in
a solid bar. For the open pipe there must be a loop of displacement at
the open ends.
The fundamental resonant frequency of a pipe, open at both ends,
Fig. 3.9, is
c c
1=X=U 3.20
L N L L
L N L N
L N L N L N L b
><::
L N L
FIG. 3.9. Modes of vibration of the air column in a pipe open at both ends and
in a pipe closed at one end and open at the other end. The velocity nodes and
loops are indicated by Nand L.
The overtones of the pipe closed at one end are the odd harmonics.
That is h = 311, fa = 5!I, etc.
In the above examples the end connection has been omitted. Rayleigh 23
23 Rayleigh, "Theory of Sound," Macmillan and Company, London, 1926.
70 ACOUSTICAL ENGINEERING
shows the added length at the open end to be .82R where R is the radius
of the pipe. If the pipe is terminated in a large flange the end connection
will be that given in Sec. 5.12.
Organ pipes and whistles have been built to cover the range from 16
cycles to 30,000 cycles. The frequency of open and closed pipes may be
computed from the above equations. The sound vibrations in the pipe
are set up by the stream of air which is controlled by the vibration in the
pipe. It is an oscillatory system fed by a direct current of air or gas.
r
DYNAMICAL ANALOGIES
1 The use of the terms" circuit" and" network" in the literature is not established.
The term" circuit" is often used to designate a network with several branches.
2 The term" single dimension" implies that the movement or variation occurs along
a path. In a field problem there is a variation in two or three dimensions.
3 For further considerations of analogies see Olson, " Dynamical Analogies," D . Van
Nostrand Company, Princeton, N.J., 1943.
DYNAMICAL ANALOGIES 73
4.2. Definitions.-A few of the terms used in dynamical analogies will
be defined in this section. Terms not listed below will be defined in sub
sequent sections.
Abvolt-An abvolt is the unit of electromotive force.
Instantaneous Electromotive Force-The instantaneous electromotive
force between two points is the total instantaneous electromotive force.
The unit is the abvolt.
Effective Electromotive Force-The effective electromotive force is the
root mean square of the instantaneous electromotive force over a complete
cycle between two points. The unit is the abvolt.
Maximum Electromotive Force-The maximum electromotive force for
any given cycle is the maximum absolute value of the instantaneous electro
motive force during that cycle. The unit is the abvolt.
Peak Electromotive Force-the peak electromotive force for any specified
time interval is the maximum absolute value of the instantaneous electro
motive force during that interval. The unit is the abvolt.
Dyne-A dyne is the unit of force or mechanomotive force.
Instantaneous Force (Instantaneous Mechanomotive Force)-The in
stantaneous force at a point is the total instantaneous force. The unit is
the dyne.
Effective Force (Effective Mechanomotive Force)-The effective force
is the root mean square of the instantaneous force over a complete cycle.
The unit is the dyne.
Maximum Force (Maximum Mechanomotive Force)-The maximum
force for any given cycle is the maximum absolute value of the instanta
neous force during that cycle. The unit is the dyne.
Peak Force (Peak Mechanomotive Force)-The peak force for any
specified interval is the maximum absolute value of the instantaneous
force during that interval. The unit is the dyne.
Dyne Centimeter-A dyne centimeter is the unit of torque or rotato
motive force.
Instantaneous Torque (Instantaneous Rotatomotive Force)-The in
stantaneous torque at a point is the total instantaneous torque. The unit
is the dyne centimeter.
Effective Torque (Effective Rotatomotive Force)-The effective torque
is the root mean square of the instantaneous torque over a complete cycle.
The unit is the dyne centimeter.
Maximum Torque (Maximum Rotatomotive Force)-The maximum
torque for any given cycle is the maximum absolute value of the instanta
neous torque during that cycle. The unit is the dyne centimeter.
Peak Torque (Peak Rotatomotive Force)-The peak torque for a speci
fied interval is the maximum absolute value of the instantaneous torque
during that interval. The unit is the dyne centimeter.
Dyne per Square Centimeter-A dyne per square centimeter is the unit
of sound pressure.
Static Pressure-The static pressure is the pressure that would exist
74 ACOUSTICAL ENGINEERING
in a medium with no sound waves present. The unit is the dyne per square
centimeter.
Instantaneous Sound Pressure (Instantaneous Acoustomotive Force)
The instantaneous sound pressure at a point is the total instantaneous
pressure at the point minus the static pressure. The unit is the dyne per
square centimeter.
Effective Sound Pressure (Effective Acoustomotive Force)-The effective
sound pressure at a point is the root mean square value of the instantaneous
sound pressure over a complete cycle at the point. The unit is the dyne
per square centimeter.
Maximum Sound Pressure (Maximum Acoustomotive Force)-The
maximum sound pressure for any given cycle is the maximum absolute value
of the instantaneous sound pressure during that cycle. The unit is the
dyne per square centimeter.
Peak Sound Pressure (Maximum Acoustomotive Force)-The peak sound
pressure for any specified time interval is the maximum absolute value of
the instantaneous sound pressure in that interval. The unit is the dyne per
square centimeter.
Abampere-An abampere is the unit of current.
Instantaneous Current-The instantaneous current at a point is the total
instantaneous current at that point. The unit is the abampere.
Effective Current-The effective current at a point is the root mean
square value of the instantaneous current over a complete cycle at that
point. The unit is the abampere.
Maximum Current-The maximum current for any given cycle is the
maximum absolute value of the instantaneous current during that cycle.
The unit is the abampere.
Peak Current-The peak current for any specified time interval is the
maximum absolute value of the instantaneous current in that interval.
The unit is the abampere.
Centimeter per Second-A centimeter per second is the unit of velocity.
Instantaneous Velocity-The instantaneous velocity at a point is the total
instantaneous velocity at that point. The unit is the centimeter per second.
Effective Velocity-The effective velocity at a point is the root mean
square value of the instantaneous velocity over a complete cycle at that
point. The unit is the centimeter per second.
Maximum Velocity-The maximum velocity for any given cycle is the
maximum absolute value of the instantaneous velocity during that cycle.
The unit is the centimeter per second.
Peak Velocity-The peak velocity for any specified time interval is the
maximum absolute value of the instantaneous velocity in that interval.
The unit is the centimeter per second.
Radian per Second-A radian per second is the unit of angular velocity.
Instantaneous Angular Velocity-The instantaneous angular velocity
at a point is the total instantaneous angular velocity at that point. The
unit is the radian per second.
DYNAMICAL ANALOGIES 75
Effective Angular Velocity-The effective angular velocity at a point
is the root mean square value of the instantaneous angular velocity over
a complete cycle at that point. The unit is the radian per second.
Maximum Angular Velocity-The maximum angular velocity for any
given cycle is the maximum absolute value of the instantaneous angular
velocity during that cycle. The unit is the radian per second.
Peak Angular Velocity-The peak angular velocity for any specified
time interval is the maximum absolute value of the instantaneous angular
velocity in that interval. The unit is the radian per second.
Cubic Centimeter per Second-A cubic centimeter per second is the unit
of volume current.
Instantaneous Volume Current-The instantaneous volume current
at a point is the total instantaneous volume current at that point. The
unit is the cubic centimeter per second.
Effective Volume Current-The effective volume current at a point is
the root mean square value of the instantaneous volume current over a
complete cycle at that point. The unit is the cubic centimeter per second.
Maximum Volume Current-The maximum volume current for any
given cycle is the maximum absolute value of the instantaneous volume
current during that cycle. The unit is the cubic centimeter per second.
Peak Volume Current-The peak volume current for any specified time
interval is the maximum absolute value of the instantaneous volume current
in that interval. The unit is the cubic centimeter per second.
Electrical Impedance-Electrical impedance is the complex quotient
of the alternating electromotive force applied to the system by the resulting
current. The unit is the abohm.
Electrical Resistance-Electrical resistance is the real part of the elec
trical impedance. This is the part responsible for the dissipation of energy.
The unit is the abohm.
Electrical Reactance-Electrical reactance is the imaginary part of the
electrical impedance. The unit is the abohm.
Inductance-Inductance in an electrical system is that coefficient which,
when multiplied by 277 times the frequency, gives the positive imaginary
part of the electrical impedance. The unit is the abhenry.
Electrical Capacitance-Electrical capacitance in an electrical system
is that coefficient which, when multiplied by 277 times the frequency, is
the reciprocal of the negative imaginary part of the electrical impedance.
The unit is the abfarad.
Mechanical Rectilineal Impedance 4 (Mechanical Impedance)-Mechan
ical rectilineal impedance is the complex quotient of the alternating force
4 The word" mechanical" is ordinarily used as a modifier to designate a mechanical
system with rectilineal displacements and the word .. rotational" is ordinarily used as
a modifier to designate a mechanical system with rotational displacements. To avoid
ambiguity in this book, where both systems are considered concurrently, the words
" mechanical rectilineal " are used as modifiers to designate a mechanical system with
rectilineal displacements and the words .. mechanical rotational " are used as modifiers
to designate a mechanical system with rotational displacements.
76 ACOUSTICAL ENGINEERING
applied to the system by the resulting linear velocity in the direction of the
force at its point of application. The unit is the mechanical ohm.
Mechanical Rectilineal Resistance (Mechanical Resistance)-Mechanical
rectilineal resistance is the real part of the mechanical rectilineal impedance.
This is the part responsible for the dissipation of energy. The unit is the
mechanical ohm.
Mechanical Rectilineal Reactance (Mechanical Reactance)-Mechanical
rectilineal reactance is the imaginary part of the mechanical rectilineal
impedance. The unit is the mechanical ohm.
Mass-Mass in a mechanical system is that coefficient which, when
multiplied by 21T times the frequency, gives the positive imaginary part of
the mechanical rectilineal impedance. The unit is the gram.
Compliance-Compliance in a mechanical system is that coefficient which,
when multiplied by 21T times the frequency, is the reciprocal of the negative
imaginary part of the mechanical rectilineal impedance. The unit is the
centimeter per dyne.
Mechanical Rotational Impedance (Rotational Impedance)-Mechanical
rotational impedance is the complex quotient of the alternating torque
applied to the system by the resulting angular velocity in the direction of
the torque at its point of application. The unit is the rotational ohm.
Mechanical Rotational Resistance (Rotational Resistance)-Mechanical
rotational resistance is the real part of the mechanical rotational impedance.
This is the part responsible for the dissipation of energy. The unit is the
rotational ohm.
Mechanical Rotational Reactance (Rotational Reactance)-Mechanical
rotational reactance is the imaginary part of the mechanical rotational
impedance. The unit is the rotational ohm.
Moment of Inertia-Moment of inertia in a mechanical rotational system
is that coefficient which, when multiplied by 21T times the frequency, gives
the positive imaginary part of the mechanical rotational impedance. The
unit is the gram centimeter to the second power.
Rotational Compliance-Rotational compliance in a mechanical rota
tional system is that coefficient which, when multiplied by 21T times the
frequency, is the reciprocal of the negative imaginary part of the
mechanical rotational impedance. The unit is the radian per centimeter
per dyne.
Acoustical Impedance-Acoustical impedance is the complex quotient
of the alternating pressure applied to the system by the resulting volume
current. The unit is the acoustical ohm.
Acoustical Resistance-Acoustical resistance is the real part of the
acoustical impedance. This is the part responsible for the dissipation of
energy. The unit is the acoustical ohm.
Acoustical Reactance-Acoustical reactance is the imaginary part of the
acoustical impedance. The unit is the acoustical ohm.
Inertance-Inertance in an acoustical system is that coefficient which,
when multiplied by 21T times the frequency, gives the positive imaginary
DYNAMICAL ANALOGIES 77
part of the acoustical impedance. The unit is the gram per centimeter
to the fourth power.
Acoustical Capacitance-Acoustical capacitance in an acoustical system
is that coefficient which, when multiplied by 2rr times the frequency, is the
reciprocal negative imaginary part of the acoustical impedance. The unit
is the centimeter to the fifth power per dyne.
Element-An element or circuit parameter in an electrical system defines
a distinct activity in its part of the circuit. In the same way, an element
in a mechanical rectilineal, mechanical rotational, or acoustical system
defines a distinct activity in its part of the system. The elements in an
electrical circuit are electrical resistance, inductance, and electrical capaci
tance. The elements in a mechanical rectilineal system are mechanical
rectilineal resistance, mass, and compliance. The elements in a mechanical
rotational system are mechanical rotational resistance, moment of inertia,
and rotational compliance. The elements in an acoustical system are
acoustical resistance, inertance, and acoustical capacitance.
Electrical System-An electrical system is a system adapted for the
transmission of electrical currents consisting of one or all of the electrical
elements: electrical resistance, inductance, and electrical capacitance.
Mechanical Rectilineal System-A mechanical rectilineal system is a
system adapted for the transmission of vibrations consisting of one or all
of the following mechanical rectilineal elements: mechanical rectilineal
resistance, mass, and compliance.
Mechanical Rotational System-A mechanical rotational system is a
system adapted for the transmission of rotational vibrations consisting of
one or all of the following mechanical rotational elements: mechanical
rotational resistance, moment of inertia, and rotational compliance.
Acoustical System-An acoustical system is a system adapted for the
transmission of sound consisting of one or all of the following acoustical
elements: acoustical resistance, inertance, and acoustical capacitance.
Electrical Abohm-An electrical resistance, reactance, or impedance is
said to have a magnitude of one abohm when an electromotive force of one
abvolt produces a current of one abampere.
Mechanical Ohm-A mechanical rectilineal resistance, reactance, or
impedance is said to have a magnitude of one mechanical ohm when a force
of one dyne produces a velocity of one centimeter per second.
Rotational Ohm-A mechanical rotational resistance, reactance, or
impedance is said to have a magnitude of one rotational ohm when a torque
of one dyne centimeter produces an angular velocity of one radian per second.
Acoustical Ohm-An acoustical resistance, reactance, or impedance is
said to have a magnitude of one acoustical ohm when a pressure of one
dyne per square centimeter produces a volume current of one cubic centi
meter per second.
4.3. Elements. 5-An element or circuit parameter in an electrical system
5 The symbols used in this book conform with American Standards Association,
"Letter Symbols for Acoustics," Yl0.11-1953.
78 ACOUSTICAL ENGINEERING
4.1
4.2
rR=O
JR 4.3
rA = 1.u 4.4
!R = I~~ 4.7
where dO/dt angular acceleration, in radians per second per second, and
=
JR =
torque, in dyne centimeters.
Equation 4.7 states that the driving torque applied to the moment of
inertia is proportional to the moment of inertia and the rate of change of
angular velocity.
D. Inertance.-Acoustical inertial energy is associated with inertance
in the acoustical system. Acoustical energy increases as the volume current
of an inertance increases. It decreases when the volume current decreases.
It remains constant when the volume current of the inertance is a constant.
Inertance is the acoustical element that opposes a change in volume current.
Inertance M, in grams per (centimeter)4, is defined as
p = Mdd~ 4.8
e = C1E f'd t
1 = q
CE 4.12
4.13
/R=i...
C
4.14
R
rE rA TR
fN
~
IZZZZZI
-.NvVV nIffff!r
IZiZz::II
L M m I
~
IZZZZZI
~ D
•
cE cA CN CR
~f-
ELECTRICAL
:J
ACOUSTICAL
=-=
RECTILINfAL
MECHANICAL
ROTATIONAL
TABLE 4.2
TABLE 4.3
Electromo- Volts x 10 8
I e
Ml/2L3/2
T-2 Force Dynes JM MLT-2
tive Force
--
Charge or Coulombs x 10-1 q Mlj2Vj2 Linear Dis- Centimeters x L
Quantity placement
----
i Mlj2Vj2 Linear Centimeters i or v
Current Amperes X 10-1 T-l LT-l
Velocity per Second
Electrical
Capaci- Farads X 10- 9 CE L-IT2 Compliance Centimeters CM M-IT2
per Dyne
tance
Second PM ML2T-3
DYNAMICAL ANALOGIES 87
TABLE 4.3-Continued.
Sym-
Unit Quantity Unit
Quantity bol - Dimension
Isym bol Dimension
--- - - - - -
Dynes per
Dyne ML2T-2 Pressure Square p ML-IT-2
Torque Centimeter JR
Centimeter
ACOUSTICAL ELEMENTS
5.1
5.2
w = 27Tj, and
1 Crandall, " Vibrating Systems and Sound," D. Van Nostrand Company, Princeton,
N.J., 1926.
2 Lamb, " Dynamical Theory of Sound," E. Arnold, London, 1931.
3 Rayleigh, " Theory of Sound," Macmillan and Company, London, 1926.
4 Crandall, " Vibrating Systems and Sound," D. Van Nostrand Company, Princeton,
N.J., 1926.
5 Lamb, " Dynamical Theory of Sound," E. Arnold, London, 1931
6 Rayleigh, " Theory of Sound," Macmillan and Company, London, 1926.
90 ACOUSTICAL ENGINEERING
so~-+--~--+-~--~--~--~-+--~~
«
'"«~45~-+--~--+-~--~--~--~-+--~~
i40~-1--~--+---~-+--~--+--1--~~
=>
~35r--+---r--+-~---r--~-.~-+---r--1
'"
Q.
",30~-+--~--+-~--~--~--~-+---r~
u
z
«25r--+---r--+-~--~--~--~-+--~~
lii
~20r--+---r--+-~---r--~--~-+---r--1
~
~ISr--+---r--+-~---r--~--~-+---r--1
u
~IOr--+---r--+-~---r--~--~-+---r--4
o=>
u
« 5r--±---r--+-~---r--~--r--+---r--i
oL-~--~2~~3--~4--~5--~6--~7~~8--~9~~IO·
LAYERS OF SILK CLOTH
and the number of layers of the cloth. The acoustical resistance of sheer
silk cloth as a function of the number of layers of the material is shown in
Fig. 5.1. As in the case of the small tube and narrow slit, the ratio of
acoustical resistance to inertance is governed by the size of the holes (see
equations 5.1 and 5.2).
Silk cloth has been used as an acoustical resistance element in micro
phones, telephone receivers, and loudspeakers for many years. The struc
tural simplicity and the high ratio of acoustical resistance to inertance
make it particularly desirable as an acoustical resistance.
ACOUSTICAL ELEMENTS 91
5.6. Inertance.-Inertance is defined, in Sec. 4.5D, as
M = mass 5.3
S2
where S = area, in square centimeters, over which the driving pressure
acts to drive the mass, in grams.
The acoustical impedance of various types of systems will be considered
in Secs. 5.8, 5.9, 5.10, 5.11, 5.12, 5.13 and 5.14. The imaginary part of these
expressions is due to the inertance of the systems.
For closed systems the acoustical resistance term of Secs. 5.8, 5.9, 5.10,
5.11,5.12, 5.13 and 5.14 should be omitted because there is no radiation.
In this case the entire acoustical impedance is positive acoustical reactance.
The acoustical reactance term of equations 5.1 and 5.2 is due to inertance.
5.7. Acoustical Capacitance.-The most common type of acoustical
capacitance used in acoustical systems consists of a cavity or volume with
rigid boundaries. The linear dimensions of the enclosure are assumed to
be small compared to the wavelength.
For equation 1.21 the sound pressure is
p = pc 2s 5.4
where p = density of air, in grams per cubic centimeter,
c = velocity of sound, in centimeters per second, and
s = condensation.
The condensation, from Sec. 1.3D, is
dV
s=V 5.5
w = 2'TTJ, and
ZM
=
'TTR2 pc [1 _h(2kR)]
kR + J.'TTWPK
2k 3 (2kR) 1 5.10
where R = radius of piston, in centimeters,
p = density, in grams per cubic centimeter,
c = velocity of sound, in centimeters per second,
k = 2'TT/A,
A = wavelength, in centimeters,
W = 2'TTJ, and
f = frequency, in cycles per second.
hand Kl may be found in treatises10 ,l1 on Bessel functions. They
are also defined by the series,
h(2kR) k2R2 k4R4 k 6R6
1 - ---rR = 2 +
- 22.3 22.32.4'"
K (2kR)
1
= 3[(2kR)3 _ (2kR)5
'TT 3 32.5
+ 32.5
(2kR)7
2.7' . .
]
5.11
The acoustical impedance, in acoustical ohms, of the air load upon one
side of a vibrating piston in an infinite baffle is
ZA
= ~~
'TTR2
[1 - h(2kR)]
kR
+ 2'TTR4k3
jwp K (2kR)
1
5 12
.
I
t
.8
~
.'
1//
........ "
",
.4
~ .' -'" j',
,,;V .., \" '.
.2
..'
,,~ .....' -' II/ 1/'~
I·
.1
« .08
"'a:«
.'
.04
2 I 3 . \fl
. 11
" /
x.
" .... /2 J,
.."'
a:
.0
XI 00 XI :3
w
<J
Z .01
",y" ....... ,<!, : r,
«
a .008
..
w
:f
"
.
.004
..J
« !
~
I-'
<Il
:J , II
o
<J I
« .001
.0008
:
.0004
I /
/ II
.000
/'/ .02 .04 .08 . 1 .2 .4 .8 I 2 4 8 10
kR=~
FIG. 5.2. The acoustical resistance, 1'1, and the acoustical reactance,
Xl, load per unit area divided by pc, as a function of kR for the
following radiators: 1. a vibrating piston of radius R set in an
infinite baffle ; 2. a pulsating sphere of radius R; 3. an oscillating
sphere of radius R. Note: The ordinate scale of the characteristics
labeled 3 must be multiplied by one-third. (See Sec. 5.10.)
motion of the sphere itself, take place only in radial directions and will have
the same velocity in all directions, but will depend upon the distance from
the center of the sphere.
The mechanical impedance, in mechanical ohms, of a pulsating sphere is
ZM -
_
47TR pc
2
1 ++
[(kR)2 jkR]
(kR)2 . 5.14
The resistive and reactive impedance components of the air load per unit
area of a pulsating cylinder are shown in Fig. 5.3.
12 Rueter and Bolt, "Sonics," John Wiley and Sons, New York, N.Y., 1955.
96 ACOUSTICAL ENGINEERING
I
.8
A 1--' I-"'"
//
" 1'-'''',
u / ~/ "- "
a. .2
,,'V
-I-
...
\oJ
.I
.08
~/
k'r'i
lI, ~/ J ""
...
It:
l
i' .04 /1 If1 "
::>
/ XI r. "'
'"
\oJ
CL .02
/ 2.
\oJ
./
U / r,
~ .01
~ .008
Q.
:I!
.004 L J
.....J
U
.'
:;; . 002 /
::>
o
u
.... 001 /
.0008
.0004
.0002
.0001
.01 .02 .04 0.8 J .2 .4 .8 I 4 e 10 20
kR= 2~R , kD = 2~O
FIG. 5.3. The acoustical resistance, YI, and acoustical reactance, Xl,
load per unit area divided by pc as a function of kR or kD, for two
radiators as follows: 1. A pulsating cylinder of infinite length and of
radius R. 2. A vibrating strip of infinite length and of width 2D, set in
an infinite baffle. (1. After Rueter and Bolt.)
13Rueter and Bolt, "Sonics," John Wiley and Sons, New York, N.Y., 1955.
14The expressions given in equations 5.23, 5.24, and 5.25 are not the result of an
analytical derivation, but are approximations obtained from numerical integration.
ACOUSTICAL ELEMENTS 97
where 2D = width of the vibrating strip, in centimeters,
p = density, air grams per cubic centimeter,
c = velocity of sound, in centimeters per second,
k = 27r/).., and
).. = wavelength, in centimeters.
The acoustical impedance, in acoustical ohms, of the air load per unit
length, upon a vibrating strip set in an infinite baffle is
_ pc [(2kD)3/2 +(2kD)2/3]
ZA - 2-D 1 + (2kD)3/2 5.24
The acoustical impedance per unit area of the strip is
_ [(2kD)3/2 + (2kD)2/3]
Zl - pc 1 + (2kD)3/2 5.25
The resistive and reactive acoustical inpedance components of the air load
per unit area of one side of an infinitely long vibrating strip set in an infinite
baffle is shown in Fig. 5.3.
5.13. Mechanical and Acoustical Impedance upon a Vibrating
Piston in the End of an Infinite Tube. 15 ,16-The resistive and reactive
acoustical impedance components of the air load per unit area on the free
space side of a vibrating piston set in the end of an infinite tube is shown in
Fig. 5.4.
The mechanical impedance, ZM, in mechanical ohms, of the air load on the
free space side of a vibrating piston set in the end of an infinite tube is
given by
ZM = 7rR2Z1 = 7rR2(rl Xl) + 5.26
where Zl = acoustical impedance per unit area,
rl = acoustical resistance per unit area, Fig. 5.4,
Xl = acoustical reactance per unit area, Fig. 5.4, and
98 ACOUSTICAL ENGINEERING
Referring to Figs. 5.2 and 5.4, it will be seen that in the region below k = 1
the radiation resistance for a vibrating piston in an infinite bailie is two times
the radiation resistance of a vibrating piston located in the end of an infinite
tube. This agrees with the conclusions of Sec. 2.2 for a point source radiating
2
I
-.-.
.8
.,
.4 I \
/ I
I
.2
I .~ \ I
u
Q.. I
X, ~ .'x, "
I "'~
.~ .08
\oJ
.,
a: ,. '1
>- .04
Z
:>
/ ~
ffi .02
.' Vi 2
/~ I.
Q.
\oJ
~ .0 I
r,
~ .008
Q.
~
~ .004 I
u
;:: I
(/)
..g
u
.002
"
.00 I
I
.0008
.0004
I
I
.0002 I
.000 I
.01
/
.02 .04 .08 .I .2
I
I
.4 .8 I 2 4 8 10 20
kR= 2~R
FIG. 5.4. The acollstical resistance, rl, and the acoustical reactance, Xl,
load per unit area divided by pC'as a function of kR for the following radia
tors: 1. a vibrating piston of radius R set in the end of an infinite pipe.
2. a vibrating piston of radius R in free space.
into 27T and 47T solid angles. This is to be expected because in the region
below kR = 1 the piston is essentially a point source.
Computing the end correction17 for a flanged and an unflanged pipe from
the mass reactance in the region below kR = 1 from Figs. 5.2 and 5.4, the
end corrections are .82R and .61R, respectively.
17 Levine and Schwinger, Phys. Rev., Vol. 73, No.4, p. 383, 1948.
ACOUSTICAL ELEMENTS 99
5.14. Mechanical and Acoustical Impedance upon a Vibratin~
Piston in Free Space,18,19-The resistive and reactive acoustical impedance
components of the air load on one side of a vibrating piston in free space is
shown in Fig, 5.4.
The mechanical impedance, ZM, in mechanical ohms, of the air load on
one side of a vibrating piston in free space is given by
ZM = 7TR2zl = 7TR2(rl + Xl) 5.28
where Zl = acoustical impedance per unit area,
rl = acoustical resistance per unit area, Fig. 5.4,
Xl = acoustical reactance per unit area, Fig. 5.4, and
R = radius of the piston, in centimeters.
The acoustical impedance, ZA, in acoustical ohms, of the air load on one
side of a vibrating piston in free space is given by
Zl (rl + Xl)
5.29
ZA = 7TR2 = 7TR2
w = 27r/, and
18 Wiener, F. M., Jour. Acous. Soc. Amer., Vol. 23, No.6, p. 697,1951.
22 Goldsmith and Minton, Proc. Inst. Rad. Eng., Vol. 12, No.4, p. 423,1924.
23 Slepian and Hanna, Jour. A mer. Inst. Elec. Eng., Vol. 43, p. 393, 1924.
24 Ballantine, G., Jour. Frank. Inst., Vol. 203, No. 1. p. 85, 1927.
25 Crandall, " Vibrating Systems and Sound," D. Van Nostrand Company, Princeton,
N.J., 1926.
26 Stewart and Lindsay, " Acoustics," D. Van Nostrand Company, Princeton, N.J.,
1930.
27 Olson and Massa, .. Applied Acoustics," P. Blakiston's Son and Company, Phila
delphia, Pa., 1934.
28 Mawardi, Osman K., Jour. Acous. Soc. Amer., Vol. 21, No.4, p. 323, 1949.
29 Lambert, Robert F., Jour. Acous. Soc. Amer., Vol. 26, No.4, p. 1024, 1954.
30 Jensen and Lambert, Jour. Acous. Soc. Amer., Vol. 26, No.4, p. 1029, 1954.
31 Scibor-Marchoki, R. I., Jour. Acous. Soc. Amer., Vol. 27, No.5, p. 939, 1955.
ACOUSTICAL ELEMENTS 101
The principle of continuity was expressed in Sec. 1.3. Applying the
principle, the difference between the influx and efflux of the fluid into the
element of volume must be equal to the time rate of growth of mass.
~'S~x = - ~xo(Sp'u) 5
ot ox .34
or
sop' + o(Sp'u) = 0 5.35
ot ox
From equations 1.19 and 1.6
- Pi> = c2.p' 5.36
From equation 1.11
oq,
u= - 5.37
ox
Substituting equations 5.36 and 5.37 in 5.35 the result may be written as
.. oq, 0 o2q,
q, - c 2 ox ox(log 5) - c2 ox 2 0 5.38
Equation 5.38 is the wave equation for the axial motion in a tube of
varying section.
5.19. Infinite Cylindrical Horn (Infinite Pipe).-The equation express
ing the cross-sectional area as a function of the distance along the axis is
5 = 51 5.39
where 51 = cross section of the pipe, in square centimeters.
The general horn equation for the infinite pipe from equations 5.38 and
5.39 is
.. o2q,
q, - C2 0X2 = 0 5.40
5.44
5.45
102 ACOUSTICAL ENGINEERING
p= - jwpE-(m/2)x [A C1--2-X]
V4k'-m'
Ejwt
5.63
XA =
pc(m2k -
Si ) 1 - 4k2
m2) 5.67
104 ACOUSTICAL ENGINEERING
The expressions for the velocity potential, pressure, and volume current
are quite complex and will not be considered.
The real and imaginary components of the acoustical impedance, in
acoustical ohms, at the throat are
rA
pc
=51
J1-!
1 _ T2
f-L
5.69
1--
f-L2
T
pc f-L
XA = s;: 1 1 - T2 5.70
--py:
where f-L = kxo = /1/0,
k = 2'TTI>',
/0 = cutoff frequency, and
/ = frequency under consideration.
Below the cutoff frequency, f-L = 1, the acoustical impedance is entirely
reactive and
I -J~f-L2 -1
pc f-L
XA = s;, 1 T2 5.71
1--
f-L2
ZAI =
PI
U1 5.72
ZAZ =
PZ
Uz 5.73
From equations 5.52 and 5.53 the expressions for the pressures and
volume currents at the throat and mouth are given by
5, 52 At x = 0, PI = kcpA,}kct 5.74
Ul = SlkA€1kct 5.75
At x = l, P2 = kcpA€1k(ct-l) 5.76
CYLINDRICAL U 2 = SlkA€1k(ct-l) 5.77
L==£ CONICAL
0,
ZAl, at the throat in terms of the length and
cross-sectional area of the horn and the acousti
cal impedance, ZA2, at the mouth is
~2 ZAI = ~ (~IZA2 co~ (kl)
51 JSIZA2 SIn (kl
+
+
jpc sin (kl))
pC cos (kl)
5.78
ZA'~A2
where p = density of the medium, in grams per
~l EXPONENTIAL
'I cubic centimeter,
k = 2rrjA,
FIG. 5.6. Finite cylindrical,
conical, and exponential
A = wavelength, in centimeters,
horns. Z..tl = input acousti c = velocity of sound, in centimeters per
cal impedance at the throat. second,
Sl = cross-sectional area at 51 = cross-sectional area of the pipe, in
the throat, in square centi
meters. X..t2 = terminating square centimeters,
acoustical impedance at the l = length of the pipe, in centimeters, and
throat. S2 = cross-sectional ZA2 = acoustical impedance at the mouth,
area at the mouth, in square
centimeters. 1 = length, in
in acoustical ohms.
centimeters. The throat acoustical impedance character
Z"-S~~CM. ZA.
f-----Z-S-C-M-.-~""'i
ZO
18
~14
;; 12
Ii..
x 10
t: 8
z
;§ 6
w "
I 1
~ 4
l
----
.' I [)
..J Z
0(
!! 0
/, "- "r1 ~'-:
;- ....
-:~
~ -2 /"
80( -4
l I
I XA,
-6
-8
II
-,0,02
"
] 5 6 7 8 'loJ 2 J"" 5 & 7 8 • 1041
fREQUENCY IN CYCLES PER SECOND
FIG. 5.7. The throat acoustical resistance and acoustical reactance fre
quency characteristics of a finite cylindrical horn. 1'.tIl = acoustical
resistance. XA1 = acoustical reactance. Note: The characteristics shown
are the throat acoustical resistance and acoustical reactance multiplied by
S1 and divided by pc.
A = wavelength, in centimeters,
1.4
v
."
'1.2
vi
1.0
'"z V
rAJr ~
u
V"
.. .8
.
o
'"
:I .6
,11
..
..J
,tV ,
, - , X..,
u " , \ '"\
-....
~ ,
~.
::>
o
~
r ~V .....
.....
.2 ..... .....
o
102
"
_...----- /
.
FREQUENCY
.5 6 7. 101
IN CYCLES PER
1
SECOND
" 5 I 1 • 10•
pc 1m J
~
pc
+ j S2
ZA2(1 - m1)
2 2
ZAI = ~ jZA/!!! + ~(l + m1) 5.81
2 S2 2 I
1.4
2
II
f\' I\
x
bJ
U
1.0
11II \ VV '-'
~
a
.8 ,
bJ \
"
::; .6 f
I
I
I.'
I
II \
\
, , I, ,
I
,
,..XAI
\
~,/ / " I \
-'
,~
/ '-' , \~,
' \
\., \, ...
010 2 2 • 4• 6 7 ' 10' 2 '4 S 6 7 8 104
FREQUENCY IN CYCLES PER SECOND
A c
..I~. 5
~
8
12 « III
u )( rAJ
7 ,\ .t\
i= wI.
"
6
5
I
I
III U
:>z
~
o «
r;].5
JnVJ\\ 1/ \.. ~X
"
I
I
rAI 11. J ,J
~,
'-; \..,., , f--
4
,: ~
o
f--
3 ,.
'.
100
2
,, . I
"
I
: \
1\
11!\ J
'-;J.\t
o •
I) VI b?i l\ " I
D
79'_
I
,/ .,
• I
"" ..I
1.25
Uo
-,
~A'
~~
4
I « ~.25 ...
100 200 400 1000
~
~
--
FREQUENCY zoo 400 1000
5. ___ -L fREQUENCY
..L~ 20"
B 1;;r~T
3
u
~)(
i=w
lIlu
~
..I Iff2
:>z I,
0..:
I:
: ' ,r~ r,,'A. no'
/\ VJ' yJ".
"'",
, I
• I
A-
..I~.OO , ,,
~ " •75
t- w
IIlU
5 ~.50
E
/'-,
, )l
.'x ~ / \ 1'\ 1'.. . ...
uo ,
« ~o •• Al
u8 I ,~
X~I
~
I
"
« 11. .25
~ -- -
100 zoo 400 1000 200 400 1000
FREQUENCY fREQUENCY
throat size with the mouth and flare held constant is of interest in determin
ing the optimum length and a suitable matching impedance for the driving
mechanism. The acoustical impedance characteristics of four horns hav
ing a cutoff of 100 cycles, mouth diameter of 20 inches, and throat diameter
of 1, 2, 4, and 8 inches are shown in Fig. 5.11. A consideration of these
112 ACOUSTICAL ENGINEERING
characteristics shows that the throat size has no appreciable effect upon
the amplitude of the variations in the acoustical impedance characteristics.
However, the separation in frequency between successive maxima is in
creased, as the throat becomes larger, due to the decreased length of the
horn. The frequency at which the first maximum in the acoustical resis
tance characteristic occurs becomes progressively higher as the length is
decreased.
The characteristics in Figs. 5.10 and 5.11 cover the range from 100 to
1000 cycles, the lower value being the flare cutoff frequency. The finite
u
. 4
c
~
~ "'-z A
~ 1\
<.1 )( rA. rA •
" /11\
j:w I
"'0
5~
00
« ~o
:I,
VlV\ V\
1\
~"
,,
'XAl. 1\
/\
~,lr. ,
11;:".. /\I'
.'
• , •
~
~
IA
\ "
i'"
::::Ii iI v'
-I I
100 zoo 400 1000 100 zoo 400 1000
fREQUENCY fREQUENCY
.....L~T
8# 20#
B D TI____ .L
1+ 19 5".1
FIG. 5.11. The throat acoustical resistance and acoustical reactance fre
quency characteristics of a group of exponential horns, with a flare cutoff
of 100 cycles and a mouth diameter of 20 inches, as a function of the throat
mouth diameter. Sl = the .throat diameter, in square centimeters.
r.H = acoustical resistance. XA1 = acoustical reactance. Note: The
characteristics shown are the throat acoustical resistance or acoustical
reactance multiplied by Sl and divided by pc.
ZAI .......
-~---
~16.7CM~
r
,...
,
/\ lA, fI
v
, .f.
..... ,, \\ X
•, , V I
"
--
2
~'
- - 100 1000
[1"\
"'
\
10000 '00
\Y '\rl"
1000 10000
FREQUENCY FREQUENCY
FIG. 5.12. The throat acoustical resistance and acoustical reactance frequency charac
teristics of two exponential connectors with a flare cutoff of 100 cycles. The mouth of
the horn is connected to an infinite pipe. r A1 = acousticalresistance. XA1 = acoustical
reactance. Note: The characteristics shown are the acoustical resistance or acoustical
reactance multiplied by 51 and divided by pc.
pcrl + j~ - ~l
ZAI = ~ I + l~ + j l~_ 5.89
114 ACOUSTICAL ENGINEERING
,~~~
,:Jtr~:~1
1.2r-,--,-rrTTrr-'--'-""-'---'--'''''TTr--'-''''-'--'--'rTl"TT---'
1.I1-++_++t+f-t---+-H-+-Ht+I--+_-+-+--Hm%----j
l.ol-+_+--i---t-1f-ttt---+_-+-++-+++++---+---t-t-+hVH++\-\v-T-I 1
uilg,gl--t-++++++t---t--t-++++++I----+-+-+-++-f-ttt------1
x .8t---~+-r+++H_--+-~-+~++t----+--+-+~-H+r-~
~ ~1-++_++t+f-t---+-H-+-Ht+I--+_-+~/__H-H~----j
z ~I~
8·61--1--=:~:~::::~====:==~~~:~::::~===:~~~~+L-~r-~~~+~~+-_-_-_~~
Q..5
~ ,
~t__"mr+Hr+++H---+-l--
v~+-+~+I------t--+-r,~,~H++--~
/
~.3~,fl+~,I--H~~~-~~~-I--+-+~+I------tX~~,~-I--~H+-~
~.2 ,! \:
6 . j I,
1
I
~,
, I
;: \ l\._, .. _ .-,'
,," ,
",,'
,-... ,
'J
" ,,
~ 0 \:
-1 20 100 1000 10000
FREQUENCY IN CYCLES PER SECOND
The ratio of the pressure at the closed end of the tube to the free space
pressure is useful in predicting the performance of pipes and cavities.
The ratio of the pressure at the closed end to that in free space is
.2
~, .2 .3 .4 .S.6 .8 1.0 2 3 4 .S
kR=~
The characteristics of Fig. 5.14 depict the ratio of the pressure at the
closed end of a cylindrical cavity to the pressure in free space as a function
of the dimensions of the cavity and the wavelength of the sound.
5.32. Sound Transmission in Tubes. 38 ,39,4o,41-The effect of viscosity
upon the characteristics of small holes and slits was considered in Sees. 5.3
38 Crandall, .. Vibrating Systems and Sound," D. Van Nostrand Company, Princeton,
N.J., 1926.
39 Lamb, .. Theory of Sound," E. Arnold, London, 1931
40 Rayleigh, .. Theory of Sound," Macmillan and Company, London, 1926.
41 Mason, W. P., Phys. Rev., Vol. 31, No.2. p. 283, 1928.
ACOUSTICAL ELEMENTS 117
and 5.4. The transmission loss in tubes of circular section is of interest in
problems in acoustics involving the use of tubes. The equation 41 expressing
the sound transmission in a tube is
A = Aoc ax 5.93
where A = amplitude (pressure or volume current) at a distance x centi
meters from the amplitude A o,
a = ;~J~:,
R = radius of the tube, in centimeters,
w = 27TJ,
- --
· ~
2
•
V - 2$" V
.-
-
•
· .-
..2
2
··
I
~
~
f.- - ~
~
..
L.-
-- I
'-
0
I--"'" i--'
2
l.--
V- f-"
.0I. 3
• $ e 7 a IJI
10 2
2 3 4 5 & 7 a II
/0 3
2 3 .... 5 e 7ee 10'"
rREQUENCY IN CYCLES PER SECOND
rAl = -
pc PI
=- 5.96
51 Ul
where p = density of the medium, in grams per cubic
centimeter,
FIG. 5.16. Two con c = velocity of sound in the medium, in centi
nected pipes of cross meters per second, and
sectional areas S 1 and
52. 51 = cross-sectional area of the first pipe, in
square centimeters.
rA2=-=-
pc P2' 5.97
52 U2
where 52 = cross-sectional area of the second pipe, in square centimeters.
Expressing equation 5.91 in terms of pressure
h51 - h'51 = P25 2
or
PL -P'1 -_P25
51
2
5.98
U2=~ 5.100
1 + 51
52
ACOUSTICAL ELEMENTS 119
Equations 5.99 and 5.100 show that the pressure and volume current
of the transmitted wave in pipe 52 is always in phase with the pressure and
volume current of the incident wave in pipe 51.
The reflected pressure in terms of the incident pressure is
Equations 5.101 and 5.102 show that if 51 < 52 the reflected pressure or
volume current are in phase with the incident pressure or volume current. If
51> 52, the reflected pressure or volume current are opposite in phase with the
incident pressure or volume current. If 51 = 52, there is no reflected wave.
The ratio of the transmitted power to the incident power is
5.103
k = 21T/A,
A = wavelength, in centimeters.
At the boundary B the conditions for the volume current may be written
U2clkl - U2'Elkl = Ua 5.107
where U 2 =transmitted volume current in 52, in cubic centimeters per
second,
U 2 ' = reflected volume current in 52, in cubic centimeters per second,
and
Ua = transmitted volume current in 5 a, in cubic centimeters per
second.
From equations 5.104,5.105,5.106, and 5.107,
k = 27T/A,
47 Sivian, L. J., Jour. Acous. Soc. Amer., Vol. 9, No.2, p. 135, 1937.
48 Molloy, C. T., Jour. Acous. Soc. Amer., Vol. 16, No.1, p. 31, 1944.
L CE To
~=:Ji
ELECTRICAL CIRC UIT
VALUES
OF
Qr
1.0
.8 1 - '-f-
.2
l--' ./
V V / \\ '\r--... I'-.I'-. ~ 10
......
I---t
I
.08
,,,
0 - 1-
~ 7
w .04
f/)
z
o
~
7
/"
V
/ ,I
I \~
\.. " r-.....
~
.........
II.
f/) .02 ""- so ......
~ V V II \ \ I'-. r-...
w
II: ..........
.0 I
.008
./ " ~ 00
'-'" K
.004
~ / '" r-.
./ r-...
ff
.002 I-f-
V
.00.I 5
l--' .6 .7 .8 .9 1.0 1.1 1.2 1.3 IA 1.5 1.6 1.71.81.9 2.0
1---1--
f+fr
124
A B c
<! r! 10'
r
I~
I/ /
u
z
C3
·fu • \ x.... <'i x, 10" <:
o
,~ u
'" I / '\ / ~ 10'
,
Q.
Q. :;::,
:!!
" , / ..L "
..J
<:
II
7 " // ..J
<:
u "
/
X..a 'XV \ , X MS
I~
~Io'
2
c/
<: I
J:
II Xw:' ,
'/ ~
J:
u u /
'"" , / '
,
XMS
!oJ
1,1 /17<
I v
/ // , V
" / / X"" /~A ,
10
2 4 air! 2 "" 8103 2 "" Buf Z
o. 4
J •
•10' 2 4 •
"
2 10 2 4
r/ .,0' 2 4 "10' 2
102 02' •
fREQUENCY FREQUENCY fREQUENCY
,
}J
~ I.b lu.
,. 10 0
II', , 1;
0
,
!:1 zw
";;:'
..'"u I I
I '\
U
...;;: I
I I\,
'"
I
'0. 2
I 4
I
"10> 2. 1()3 2
fREQUENCY
• 4
•10' • 0.I,
"10> 2 .•
fREQUENCY
10' • 4
•10' 2
FIG. 6.2. The mechanical impedance frequency characteristics of three direct radiator
loudspeakers having I-inch, 4-inch, and I6-inch diameter cones. X.lfC = the mechanical
reactance due to the cone and coil. XMS = the mechanical reactance due to the suspen
sion system. XMA = the mechanical reactance due to the air load. 'YMA = the
mechanical resistance due to the air load. The efficiency characteristics shown are for
the constants as shown in the table and the graphs of the mechanical impedances. In
the efficiency characteristics, /-'1 = the efficiency for XMS equal to zero. /-'2 = the
efficiency for XM8 as indicated by the graph.
128 ACOUSTICAL ENGINEERING
In terms of the resistivity and density of the voice coil, equation 6.6
becomes
6.7
TABLE 6.1. DENSITY p. in GRAMS PER CUBIC CENTIMETER; RESISTIVITY K" IN MICROHMS
PER CENTIMETER CUBE AND DENSITY-RESISTIVITY PRODUCT pKT OF VARIOUS ELEMENTS;
TEMPERATURE, 20° C.
Element p KT pKT
The relation between the efficiency and the ratio of the mass of the voice
coil to the mass of the cone and the air load may be obtained from equation
6.7 and is depicted in Fig. 6.3. The maximum efficiency occurs when the
mass of the voice coil is equal to the mass of the cone and air load.
DIRECT RADIATOR LOUDSPEAKERS 129
~o
w
In
Vi-' -.......
li
w
a
I
V
/
"
~
-2 / "\
III
'-I"
o
V \
-3 / '\
\
/ 1\
/ .2 .4 .8 I 2 4 8 10
m,
mo+m"
FIG. 6.3. The efficiency loss in a direct radiator loudspeaker as a
function of the ratio mr , where mr = the mass of the voice
mn + mA
coil, mn = the mass of the diaphragm, mA = the mass of the air
load. The maximum efficiency is arbitrarily depicted as db, °
In general, in commercial loudspeakers 3 ,4 it is not practical to make the
cone mass equal to the voice-coil mass. As a matter of fact, the cone mass
is usually several times the voice-coil mass. A consideration of equation
6.7 shows that the efficiency can be increased by the use of a light-weight
9
----- ----
06
~
t
:::J
Q.
t-3
r-
:::J
o
o
LO 1.5 2.0 2.5 3.0
RELATIVE CONE MASS
4 Olson, Preston, and May, Jour. Aud. Eng. Soc., Vol. 2, No.4, p. 219, 1954.
There is a limit to the extent to which the reduction in mass of the cone can
be carried because, as the cone mass is reduced, the strength of the cone is
reduced and as a consequence the nonlinear distortion is increased due to
overload of the material of the cone. High sensitivity and low distortion
are not compatible. In order to obtain low nonlinear distortion, a relatively
heavy cone must be used. The subject of nonlinear distortion and cone
weight will be discussed in Sec. 6.26.
The mechanical impedance and corresponding efficiency characteristics
assuming the mechanical reactance due to the suspension to be zero are
shown in Fig. 6.2. The air load mechanical resistance and mechanical
reactance are assumed to be the same as those on two sides of a vibrating
piston with the diameter equal to the cone diameter (see Sec. 5.8). The
weights of the cones and voice coils are typical of loudspeakers in actual
use today. It will be seen that the efficiencies of all three systems are
practically the same. Of course, the power-handling capacity of the smaller
cones is very small at the lower frequencies.
In the preceding considerations the mechanical reactance due to the
suspension system was assumed to be zero. The efficiency in which all
the elements of the vibrating system are included may be obtained from
equation 6.5. The mechanical, resistance rMC, due to the suspension system
is also a factor in the efficiency in the region of resonance. Typical values
of rMC for 16-, 4-, and I-inch cones are shown in Fig. 6.2. The efficiency
characteristics under these conditions are shown in Fig. 6.2. It will be
noted that the efficiency is high at the resonant frequency. However,
when coupled to a vacuum tube driving system the motional electrical
impedance is also increased which reduces the power input to the voice
coil. For this reason the response is not accentuated to the degree depicted
by the peak in the efficiency characteristic. It will be seen that the
efficiency decreases very rapidly below the resonant frequency. There
fore, in a direct radiator l~)Udspeaker the limit at the low-frequency end
of the frequency range is determined by the resonant frequency of the
system.
The motional electrical impedance of a dynamic loudspeaker is given
by equation 6.3. The normal electrical impedance, in abohms, of voice
coil is given by
ZEN = ZEM + ZED 6.8
where ZEM = motional electrical impedance, in abohms, and
ZED = electrical impedance of the voice coil in the absence of motion,
that is blocked, in abohms.
The components of the motional electrical impedance are shown in
Fig. 6.5. At the resonant frequency the motional electrical impedance is
large because the mechanical impedance is small. The current in the voice
coil circuit may be determined from the voice coil electrical circuit, the
driving voltage and the electrical resistance of the generator.
DIRECT RADIATOR LOUDSPEAKERS 131
The mechamotive force or driving force,5 in dynes, applied to the mechan
ical system is
1M = Eli 6.9
where B = flux density in the air gap, in gausses,
l = length of the conductor, in centimeters, and
i = current in the voice coil circuit, in abamperes.
This is the driving force,1M, applied to the mechanical system as shown in
Fig. 6.1.
100 140
80 120
Ul
:::;;
a::
1\ ZE... JrEM +jX~ ... Ul
:::;; ZEN 1= ZED+tEM
0 60
5 100
I"E...
40 80
f\
!!:
.., 20
J \ ~
60
\ ZEN
~/
\ w
u
z
« 0
'-.... «
u
Z
40 \
..,
0
~-20 -
XE, ~ 0
w
Q.
~
20
V \
a 102
1I 2
./ ZED
'"
4 8 102 2 4 8 1()3
FREQUENCY FREQUENCY
FIG. 6.5. The electrical impedance characteristics of the voice coil in a direct
radiator loudspeaker. ZEN = the normal electrical impedance. ZED = the
damped electrical impedance. ZEM = the motional electrical impedance.
rEM = the resistive component of the motional electrical impedance. XEM =
the reactive component of the motional electrical impedance.
40
y 1/
;:'~T
~
/ 10T"
/ V
/'" ~
..
.J
<)
~ 10
I-
~ I
~
1--"---
r------ ~
f--'-"'
..,.J
. ..
o 1 ••
j
2 3 . oS 6
rREQUENCY IN CYCLES
7 8 •
10'
2
PER SECOND
3 4 !1e7S'
10'
2
FIG. 6.6. The electrical impedance characteristics of 1!-inch diameter voice coils of 140,
70, a nd 18 turns and all having 10 ohms d-c resistance.
.5 45
45
in Fig. 6.2, are shown in Figs. 6.7 and 6.8. Employing the mechanical
circuit and the electrical circuit of Fig. 6.1 and the data of Fig. 6.2, the total
output of the loudspeaker may be determined as outlined in this section.
It is quite obvious that the response on the axis will be accentuated at the
high frequencies due to the sharpening of the directional pattern.
134 ACOUSTICAL ENGINEERING
6.12
where P = total power, in watts,
p = density of the medium, in grams per cubic centimeter,
c= velocity of sound in the medium, in centimeters per second,
p = root mean square sound pressure over the element of area dS,
in dynes per square centimeter, and
dS = element of area on the spherical surface, in square centimeters.
In the case under consideration the power output, P, as a function of
the frequency may be determined from equation 6.5 and the electrical
input. The directional patterns for the cones having diameters of 4 and
16 inches are shown in Figs. 6.7 and 6.8. From these data, the pressure
on the axis may be determined from equation 6.12. The computed re
sponse frequency characteristics of the loudspeakers of Fig. 6.2 are shown
in Fig. 6.9. These characteristics are quite similar to the actual response
frequency characteristics.
Another factor of interest in a direct radiator is the power handling
capacity. The sound power output, in watts, is given by
6.13
where rMA = mechanical resistance, in mechanical ohms, obtained from
Sec. 5.8, and
x= root-mean-square velocity of the piston, in centimeters per
second.
Equation 6.13 may be used to compute the power output of a direct
radiator loudspeaker in the region were all parts of the cone move in phase.
In general, the output is limited by the permissible amplitude. The greatest
amplitude occurs at the low frequencies where the action is essentially that
of a piston. Therefore, piston action may be assumed.
The peak amplitude characteristics of a 16-inch, a 4-inch, and a I-inch
piston mounted in an infinite baffle for 1 watt of sound output are shown
in Fig. 6.10. These characteristics show that for practical amplitudes
a relatively large piston is required to deliver adequate power at the low
frequencies.
The directional pattern of a vibrating paper cone depends on three
principal factors: the cone diameter, the cone angle, and the frequency.
DIRECT RADIATOR LOUDSPEAKERS 135
Other factors, such as the paper pulp, the processing, the corrugations,
the voice coil diameter, and the suspension also influence the directional
pattern, but in a lesser degree. The directional patterns for various fre
quencies of 110° cones having diameters of 4 and 16 inches are shown in
A B c
0 50 50
.,40 10 40
o 10 40
h o
l.r.. o
~ 30
.....+-. ~30
l-
~30 1'\ ......
1"\ /
z
o
~20
z
~20 i \ z
o
:};20
1\ /
<II
"''" 10 "' 10
a: '"a: 10
7 V
°2 ... 810 Z 2 A 8'012 4 1'042 °2 ... 81022 4 8'0,2 '" "0... 2 o. • 810 " • "10" • "104'
~REQUENCY ~REQUENCY ,REQUENCY
FIG. 6.9. Pressure response frequency characteristics of the loudspeakers of Fig. 6.2
having cone diameters of 1 inch, 4 inches, and 16 inches.
..• \.
\.
\.
2
1\ \
1\ 10"
1\:" I"
r\
8
•..
\. \. \.
2
\ \ 1\
1'\ \
\
~.o8I
... ,
.. I"\, "\.
'\.
2
f\ 1\.
~
'\.
r"-
0.001 2 3 " 5.78.102 2 3" 5.789101 2 3 .. 5 . 7 •• 10" 2
Figs. 6.7 and 6.8. It will be see that the directional pattern becomes
sharper with increase in frequency. However, the pattern is broader than
that of a vibrating piston of the same diameter due to the relatively low
velocity of propagation of sound in the paper cone. The directional patterns
136 ACOUSTICAL ENGINEERING
of 130° and 100° cones 4 inches in diameter are shown in Fig. 6.11. It
will be seen that the directional pattern at the high frequencies becomes
broader as the cone angle is increased. This is to be expected because the
velocity of propagation of sound in the paper cone is about two times
the velocity of sound in air. Under these conditions the delay between the
sound emitted from the outside and the center of the cone will increase as
the angle of the cone is increased. As a result the directional pattern will
be broadest for the cone with the widest angle. The preceding observations
..
.. ..
FIG. 6.11. Directional characteristics of dynamic direct radiator loudspeakers with
cones 4 inches in diameter for two different cone angles. Row A. 130 0 cone. Row B.
100 0 cone.
L--------'B
L--------~D
L---------'c
FIG. 6.12. Multiple single-cone, single-coil, direct radiator, dynamic, loud
speaker systems. A, C, and D. Large low-frequency unit, small high
frequency unit, and filter system. B. Seven small units connected in parallel.
diameter light cone and small voice coil for the high-frequency range and
a filter system for allocating the power in the high- and low-frequency ranges
to the respective low- and high-frequency units. The filter system consists
of an inductance in series with the low-frequency unit and a condenser in
series with the high-frequency unit. Due to the large inductance of the
large voice coil, as shown in Fig. 6.6, it has been found that for most applica
tions the inductance in series with the low-frequency unit may be omitted.
On the other hand, if a more elaborate filter system is required, the circuit
of Fig. 7.16 may be used.
138 ACOUSTICAL ENGINEERING
o 2f DIAMETER CONE
region the two cones vibrate together as a single cone. In this way phase
and diffraction effects are eliminated.
In a two-unit loudspeaker, employing a large cone for the reproduction of
the low-frequency range and a small cone for the reproduction of the high
frequency range, a uniform directivity pattern can be obtained over the
entire audio-frequency range. This has been described in connection with
Figs. 6.6 and 6.7. This is illustrated further in Fig. 6.13 in which the
directivity patterns of IS-inch and 2t-inch cone loudspeakers are compared
for a six to one ratio of frequency, that is, for a constant ratio of diameter
to wavelength. Fig. 6.13 shows that the directivity pattern of a IS-inch
loudspeaker at 200 to 1000 cycles corresponds to that of a 2t-inch
7 Olson and Preston. RCA RBview. Vol. 7. No.2. p. 155. 1946.
DIRECT RADIATOR LOUDSPEAKERS 139
loudspeaker at 1200 and 6000 cycles. These relationships were used in
designing the two units of the system shown in Fig. 6.12D.
In the loudspeaker8 ,9 shown in Fig. 6.12D, small cones may be attached
to the large cone to reduce the velocity of wave propagation in the large
cone. Fig. 6.14. This broadens the directivity pattern of the low-frequency
cone. In the high-frequency range, the conical domes attached to the
surface of the low-frequency cone improve the performance in three ways:
by decreasing the angle into which the high-frequency cone feeds, thereby
increasing the output of the high
frequency cone; by diffusely reflecting
some of the sound emitted by the high
frequency cone, thereby eliminating
discrete reflections; and by diffracting
some of the sound emitted by the high
frequency cone, thereby broadening the
directivity pattern.
The angles into which the high
frequency cone feeds, without and with
the conical domes applied to the low
frequency cone of Fig. 6.14, are desig
nated as 7>1 and 7>2 in Fig. 6.15A and
Fig. 6.15B. Since 7>2 is smaller than 7>1,
the acoustic radiation load upon the
cone is greater with the conical domes
than without them. When the acoustic
radiation load upon a direct radiator
loudspeaker is increased, the sound
power output is increased. Thus, it will
be seen that the conical domes increase
the high-frequency sound radiated by
the high-frequency cone. In other FIG. 6.14. A perspective view of a
duo-cone loudspeaker with domes
words, the high-frequency efficiency is attached to the low-frequency cone.
improved. (After Olson, Preston, and May.)
Some of the sound emitted by the
high-frequency cone is diffusely reflected by the conical domes, as shown in
Fig. 6.16. Without the domes, there would be many similar reflections
which would lead to reinforcements and cancellations with the direct radia
tion. The result would be corresponding peaks and dips in the response of
the high-frequency cone. With the domes, the symmetry of the low
frequency cone is upset and there are many reflections in different directions
and of different path lengths. The reflections, therefore, cancel out and the
net result is a smooth response-frequency characteristic.
Some of the sound emitted by the high-frequency cone is diffracted by
the conical domes as shown in Fig. 6.17. By diffraction is meant the bending
8 Olson, H. F., Radio and Television News, Vol. 51, No.2, p. 69, 1954.
9 Olson, Preston, and May, Jour. Aud. Eng. Soc., Vol. 2, No.4, p. 219, 1954.
I-'~ANGLE
HIGH I
FREQUENCY
CONE
SECTIONAL VIEW
A
CONICAL DOME
HIGH
FREQUENCY HIGH
CONE FREQUENCY
CONE
SECTIONAL VIEW
B
FIG. 6.15. A. Duo-cone loudspeaker with a
plain low-frequency cone. B. Duo-cone loud
speaker with domes attached to the low-fre
quency cone.
Referring to Figs. 6.2 and 6.9 it will be seen that uniform response may
be obtained over a wide frequency range by means of a light cone driven
by a light coil and resonant at the lower limit of the frequency "range.
Of course, the power handling capacity of a single unit of this type is in
adequate and a multiple set of units must be employed. The number of
units required may be determined from the required power output and the
allowable excursion together with equation 6.13 and Fig. 6.10. An arrange
ment of seven small loudspeaker units mounted in a flat baffle with the voice
coils connected in parallel is shown in Fig. 6.12B. The voice coils of the
loudspeakers may, of course, be connected in parallel, series, or series-parallel.
In order to obtain better high-frequency spatial distribution the units may
be inclined at various angles, for example, the units may be mounted so that
the resulting vibrating surface approximates a spherical surface (see
Sec. 2.20).
DIRECT RADIATOR LOUDSPEAKERS 141
CONICAL
DOME
HIGH
FREQUENCY
CONE
REFLECTED
PENCILS
OF SOUND
LOW
FREQUENCY
CONE
REFLECTED
PENCILS CONICAL
OF SOUND DOME
2
HIGH
FREQUENCY LOW
CONE FREQUENCY
CONE
SECTIONAL VIEW
DIRECT PENCILS
OF SOUND
HIGH
FREQUENCY
CONE
DIFfRACTED
PENCILS
Of SOUND
LOW
FREQUENCY
CONE
PENCILS
OF SOUND
CONICAL
DOME
SECTIONAL VIEW
the voice coil at the high frequencies. The two cones are separated by a
compliance. At low frequencies the mechanical reactance of the compliance,
eM, is large compared to the mechanical impedance, ZMl, of the large cone
and consequently the entire system moves as a whole. At high frequencies
the mechanical reactance of the compliance, eM, is small compared to the
mechanical impedance, ZMl, of the large cone, and the small cone, ZM2 moves
while the large cone, ZMl, remains stationary. By means of this reduction
in cone mechanical impedance the range may be extended almost a full
octave, depending upon the mass and electrical impedance characteristics
of the voice coil. The response characteristics of a single-coil, single-cone
DIRECT RADIATOR LOUDSPEAKERS 143
loudspeaker is shown in Fig. 6.18A. The voice coil and large cone of Fig.
6.l8B is the same as that of Fig. 6.18A. The high-frequency range has been
extended about one-half octave without any sacrifice of power handling
capacity by the addition of the small cone.
6.5. Double-Coil, Single-Cone Loudspeaker.ll-The double-coil,
single-cone loudspeaker, Fig. 6.18C, consists of a voice coil, divided into
two parts separated by a compliance, coupled to a single corrugated cone.
The inductance and electrical resistance of the larger portion of the voice
coil, LI, rEI, is shunted by an electrical capacitance, CEo At low frequencies
the electrical reactance of the electrical capacitance is large compared to
the electrical impedance of the larger portion of the voice coil, LI, rEI,
and the mechanical reactance of the compliance, CM, separating the two
portions of the voice coil is large compared to the mechanical mass react
ance of ml, and the mechanical impedance ZMI. Therefore, in the low
frequency range the action is the same as that of a single-coil loudspeaker.
At high frequencies the reactance of the electrical capacitance, CE, is small
compared to the electrical impedance of LI, rEI or L 2, rE2; and the mechani
cal reactance of the compliance, CM, is small compared to the mechanical
reactance of large coil, mI. The cone is driven by the lighter portion, m2,
of the voice coil and the heavy coil, ml, remains stationary. In the mid
range there is a phase difference between the currents in the two portions
of the voice coil. A corresponding phase shift occurs in the mechanical
system. As a consequence, a smooth overlap is obtained in going from two
coil operation at the low frequencies to a single-coil operation at the high
frequencies. Above the frequency of ultimate resistance the radiation
resistance is a constant. In order to obtain uniform output in this range
the mechanical impedance of the system must be independent of the fre
quency. This may be accomplished by embossing suitable corrugations
in the cone which reduce the effective mass reactance. The double-coil
system reduces the effective mass reactance of the voice coil as compared
to a single coil, as well as the electrical impedance at the higher frequencies.
A typical response characteristic of this loudspeaker is shown in Fig. 6.18C.
6.6. Double-Coil, Double-Cone Loudspeaker.12-The double-coil,
double-cone loudspeaker, Fig. 6.18D, consists of a light coil coupled to a
small cone, connected by a compliance to a heavy coil and large cone. In
this system an increase in range is obtained by reducing the impedance of
both the coil and the diaphragm at the higher frequencies. At low fre
quencies the electrical reactance of the capacitance, CE , is large compared to
the electrical impedance of the large portion of the voice coil, LI, rEI, and
the same current flows in both coils. The mechanical reactance of the
compliance, CM, separating the two portions of the coil is large compared
to the mechanical impedance of ml, plus ZMI. Therefore, at low frequencies
the system behaves as a single-coil, single-cone loudspeaker. Both parts
11 Olson, H. F .. Proc. Inst. Rad. Eng., Vol. 22, No. 1, p. 33, 1034.
12 Olson, H. F., Jour. Acous. Soc. Amer., Vol. 10, No.4, p. 305, 1939.
8L~
SECTIONAL VIEW MECHANICAL CIRCUIT VOICE COIL B -..,
~r~ CIRCUIT SECTIONAL VIEW
3 ) 30,
h
~f dl
) 0 20
'0" 20 !-"" r-' ~
/\ )
w w
%'" z
'"
~ ~
"'I )
w gJ I' )
II! 0:
- '-'---
ZMI
m, m. ~rE2L2r.1 ~
Zw, ~ rEG ZM2
~ C • LL_~f
f",~r...2~ C L,
D g"r.
VOICE
C m, eM'..
ECOIL m, J M..
SECTIONAL VIEW MECHANICAL NETWORK NETWORK SECTIONAL VIEW
30, I I I I II 30
dl dl
0 20) ..... ""'-"" O 2 ,)
1""0...
w '" w
~
'"zo '"~
"- "
jl ) III l
a:
J, ...a:
, ._-
go 100 1000 10000 40 100
fR£QUENCY
FIG. 6.18. A. Cross-sectional view of a single-cone, single-coil loudspeaker with the voice c
circuit of the mechanical system. In the voice coil electrical circuit, e = the internal voltag
electrical resistance of the generator. rEi and L = the electrical resistance and the inductan
circuit. m = the mass of the voice coil. ZMl = the mechanical impedance of the cone at the
in the voice coil. The graph shows the pressure response frequency. B. Cross-sectional vi
speaker with the voice coil electrical circuit and mechanical network of the mechanical syste
e = the internal voltage of the generator. 'YEG = the internal electrical resistance of the g
resistance and inductance of the voice coil. In the mechanical network, m = the mass of
mechanical impedance of the large and small cones. CM and 'I'M = the compliance and mech
the large cone. 1M = the force-generated in the voice coil. The graph shows the pressu
C. Cross-sectional view of a single-cone, double-coil loudspeaker with the voice coil networ
mechanical system. In the voice coil electrical network, e = the internal voltage of the ge
resistance of the generator. rEI and LI = the electrical resistance and the inductance of the l
resistance and inductance of the small coil. CE = the electrical capacitance shunting the la
mi = the mass of the large coil. m2 = the mass of the small coil. ZMI = the mechanical i
CM and 'I'M = the compliance and mechanical resistance of the corrugation separating the l
force generated in the large coil. 1M2 = the force generated in the small coil. The graph s
characteristic. D. Cross-sectional view of a double-cone, double-coil loudspeaker with t
mechanical network of the mechanical system. In the voice coil electrical network, e = t
rEG = the internal electrical resistance of the generator. rEI and LI = the electrical resist
coil. r/i:2 and L2 = the electrical resistance and the inductance of the small coil. C E = th
In the mechanical network, mi = the mass of the large coil. m2 = the mass of the small co
of the large cone. ZM2 = the mechanical impedance of the small cone. CM and 'I'M = the
of the corrugation separating the large cone and coil and the small cone and coil. 1MI =
1M2 = force generated in the small coil. The graph shows the pressure response frequency c
~ P> 0-. ~ g.~ ~ (ti @ ~
g. qg. 8 8" ~ @ 8 1=?' tr ~ ~ ~ @ § ~ ~'.o ~ S @ 'B 'E..
citl S '0 .-+ ~ ~ P ::1. ~.-+
~ ~ @ ('I)
~ >9 0:: .-+ <: 'B ~ S Po. ~ p.. ~ ':" ~ ~ @ Pl'
.-+ ('I) _('I) 0 Pl. ('I) <: C" ~ 0 p" ('I) Po. .-+ .....
p" Pl "1 ('I) Pl Pl 0. ~ p" .-+ Po. ~
g- (") .-+ 0 S. S ~ If (=). ~ ~2 ~ 8 p.. it ~ ~ ~ g. ~ S 5';, ~ <:'~. § § _~
.-+ 0 p" ~ ~ "0 5' (") ('I) 0 P 0 ~ m Po. "1 Pl ~ ('I)..... ('I) Po. El m ;l ..... (").-+
~ ('I) ~ ~. ~ ~ (fl~' 8" [ ~ _('I) 9" ~ ~ _ ;. C" ;. ~ ('I) 0 ~ g.
S Po. ('I) V" <: .~J ('I) ...... -
..... --. P ...... 0 .... .0 Pl Pl ('I) '< ('I) ('I) 0 ...
o (") ..... ~ Pl.o @ ('I) m (") Pl p:; ...... Po. P .-+ "1 m C)' ...... ;. -
(2.:g 8..... aq'... §.:_ a ::r
Po. Po. ~. P .0 (ti P" .-+ ('I) .-+ (") ~ o·~ S "1 ('I)~.
§ S g g. g"
a g El :;r it.-+ S'
~ ('I» m?'Po.Po.El~"O SPlito Oo('l)Pl~ PlitPl-('I)"O~
Pl_.-+.-+@ Pl ....... m 0 (") El ('I)!::! _.-+t::1 a~ e:..ti)~
El <:
(") ::::Pl ('I)'" t::1 m
.-+ v ('I) m 00"0 P '< "1" Pl .... • ;:::> - n ~ .• 0 '< - .-+ Po. ;:::> S
('I) ('I) "1 ;'''0 S tj ~ C" ~. C)'..g ~ ~ S .-+ '< _~ ~ .-+ ~ ~ (")
is:: p" Pl Pl ~ Pl
aq' 8
.... (") 0 ('I) g~. po;' ~ ;J ~. Po.
If & (") Pl "1 ~.-+
8. it _('I) ~8.::r aq'
~ 15.. p" ::::
v p " ...... mm~ ('1)(") ;:::>('I)_~oPl m('l) 0 - - P"('I) Pl
~ @ .-+ Pl ('I) m >-l "1 >-l
g~ § Po. e; (ti f!J.
a~ 5' >-l .-+ >-l '" -
5.: 8 "'::r .-+~. 8
('I)~ p" S'-+0 0 ('I) (") p" (") '-+0 (1Q (") .... "' toO
8: .... ('I) p"
<e. o~ 8: ~ ~ ~ ~ 0 p" D> S
~1-ooI~(t) ~rnrn ... I (D(t) (t) I cn ..... (t)('t>r.nC't> en ............ ('I)I ....... ('t)i--ol
m
~Zwc
~cC B c .. rw. , ~m
SECTIONAL VIEW
~
MECHANICAL CIRCUIT seCTIONAL VIEW
; 0
~.0 Ii!Z3 0
~ o
0
:?20 1 ... 2 0
..'"
a: "' a:
0 l 0
0 .............. "
'40 100 1000 10000 "40 100
rREQUENCV
~~ Cwfo...
D c.....r.. 04Zwc
i=--m, c",rw~ h m,
SECTIONAL VIEW MECHANICAL NETWORK CIol:~M~TlONAL Vlrv:
'0 ,0
........... I-
'0 5·
m r-.... ID
- i"---
°4' °40
0
... ...
~310 ~310
f? o
0
:30: .0
..:320
a:
10 I
0
40 ,~
100 1000 Innnn
'" ~ 100
FREQUENCY
CONEEr CONEfJI
30
A r- FT.1
4
30
B !-4FT'1
~24
-
ID 24
o
=
'"'"z
012
"
16
/
".. ... 16
:r'"
z
12
V ""'
...
ell
0:
v "''0:"
6 6
V ,/
o ./ o /'
1000
20 100 1000 20 100
FREQUENCY FREQUENCY
",24
o
",24
o
v-,
... 16
ell
Z
~ 12
::Ja: V
'" 6
6
/ ' '--r ",IB
In
z
~I 2
''""
0:
6
'j
V
V
0
30 100 500 0 100 5 00 go 100 500
FREQUENCY FREQUENCY FREQUENCY
,"",~l[
.i.
Qo-:
CONEI:lI CONEeJ~
1---8 n-j
~ +12FT ~
3 3a 30
.
",I 8
<II
~I 2
/'
...... I
'" 18
'"z
~I 2
)/
".
" 12~
<II
'"
'"
<II
'"0: V ...'" V
6/ 0: '"
a
30 100 500 100 50 a ~O 100 500
I'REQUENCY I'REQUENCY FREQUENCY
FIG. 6.22. Pressure response frequency characteristics of mass-controlled, direct
radiator, dynamic loudspeaker mechanisms, with to-inch diameter cones, mounted
in square baffles of 8, 4, and 2 feet on a side.
'O"t~l~
Ii
:fJ1
-j 2fT.
no-:
CONt~~
B 1 c T
SO CYCLES 100 CYCLES 200 CYCLES
3 30 3a
'"0: 6
'"'"0:
6
/
a
30 100 500 100 500 a 100
/ 50 a
FREQUENCY FREQUENCY FREQUENCY
14 Olson. H. F., Audio Eng.. Vol. 35, No. 11, p. 34, 1951.
15 Olson, H. F., Radio and Tel. News, Vol. 45, No.5, p. 53, 1951.
16 Meeker, Slaymaker, and Merrill, Jour. Acous. Soc. Amer., Vol. 22, No.2, p. 206,
1950.
DIRECT RADIATOR LOUDSPEAKERS 153
acoustic path length is sufficiently large to insure good reproduction of low
frequencies. One of the most troublesome acoustical factors in conventional
cabinets is the resonance in the enclosure back of the cone. This resonance
is termed cabinet resonance. The system may be considered from the
standpoint of lumped or distributed constants. In the case of most systems,
the latter viewpoint seems to yield better agreement with experiment. The
cabinet enclosing the back of the cone may be considered to be a pipe with
distributed constants.
A cross-sectional view of a direct radiator loudspeaker mounted in an
open-back cabinet and the mechanical circuit of the mechanical system is
shown in Fig. 6.24. The input acoustical impedance of a finite cylindrical
pipe has been considered in Sec. 5.25. In this chapter it has been more
convenient to use mechanical impedance instead of acoustical impedance.
The mechanical impedance due to the cabinet in terms of the acoustical
impedance is
ZMl = ZAlSc 2 6.14
where ZMl = mechanical input impedance of the cabinet, in mechanical
ohms.
ZAI = acoustical impedance of the cabinet, in acoustical ohms, and
Sc = area of the cone in square centimeters.
The power output of the system may be determined from the mechanical
and electrical circuits of Fig. 6.24 and the constants of the system.
It is the purpose of the sections which follow to consider the performance
3f various types of cabinets and loudspeaker systems.
A. Low Resonant Frequency, Different Cabinet Sizes.-The pressure
response frequency characteristics of a direct radiator loudspeaker mechan
[sm, having a resonant frequency of 20 cycles mounted in cabinets of
various sizes, is shown in Fig. 6.25. The resonant frequencies at 80, 150,
0 30 3
Q Q
"wl8
I
"- I /"' '- 1 V~
wl8 wi 8
V
oil
Z
o0.1 V '"z <II
z
.
oil
0:
2
I
~J2
<II
OJ
0:
6
/'
~I 2
'"
OJ
0:
6 ~
o30 V
100 50o ~o 100 500 ~o 100 500
FREQUENCY FR~QUENCY ,REQUENCY
and 250 cycles for the 4-, 2-, and I-foot cabinets, respectively, is quite
evident. In this region the output is somewhat exaggerated in spite of
the fact that the cabinets are relatively shallow. Below the resonant fre
quency the system behaves as a doublet. Therefore, with a mass-controlled
mechanism the response falls off 6 db per octave.
r-2FT--1..1... -j 1fT tt
B LRJ'.o
100CYC~
C W'.,
200 CYC~S
0 30 3
.,24 10 24 102 4
r-...
.........
o
",18
i"......
o
",18
........ 1-
o
:;:1 8
/ I'r
<II
z
5.' 1
1
<II
z
~12
z
o0.1
2
V
'"a: 211
<II 0/1 <II
6 '"
a: (, '"a: 6
w~ B
LR..Jl
P
B r-24"....jT c 1-24"-1
30 30 30
.,24 ",24
(\
<0 24
o
l/""'
o
'" IS
A o
!oJ 18
\
'" 18
<II
z
<II
Z z
o
'" /1;'
~ 12 ~ 12 a. 12
en
V
<II
'"a: 6 V !oJ '"cr
w
cr 6 6
a 030 100
30 100 500 500 030 100 500
nU:QUENCY FREQUENCY fREQUENCY
CD 24
c / -\ CD 24
c
/" \
'" 18
If)
"- ...... ",18
If)
'
z z
~ 12 ~ 12
If) If)
w
II: 6 '"
II:
6
o I
30 100 500 100 500
FREQUENCY FREQUENCY
W1
1--24'~
c
30 CYCLES
D
CRJl
30 CYCLES
30 30
CD 2
,,-... c
i;'
",I 8
If)
z
'"
oQ. 12/
/ If)
'"
II:
6
o 0
30 100 500 30 100 500
FREQUENCY FREQUENCY
FIG. 6.30. Front and sectional views of a compound direct radiator loudspeaker.
(After Olson, Preston, and May.)
~::r~4~~
f"2 f '"
MECHANICAL NETWORK
ELECTRICAL NETWORK
CROSS-SECTIONAL VIEW
FIG. 6.31. Cross-sectional view and the electrical and acoustical networks of a
compound direct radiator loudspeaker. In the electrical network, e = internal
voltage of the generator. rEG = internal electrical resistance of the generator.
YEI and Ll = electrical resistance and inductance of the voice coil of the radiating
loudspeaker. ZEMI = electrical motional impedance of the radiating loudspeaker.
YE2 and L2 = electrical resistance and inductance of the secondary loudspeaker.
ZE'M2 = electrical motional impedance of secondary loudspeaker. CE"2 = elec
trical capacitance. In the acoustical network, Y.>fA and m.ll = mechanical
resistance and mass of the air load upon the cone of the radiating loudspeaker.
11'11 = mass of the cone and coil of the radiating loudspeaker. YMI and C Ml =
mechanical resistance and acoustical resistance and compliance of the suspensions
of the radiating loudspeaker. fill = mechanomotive force in the voice coil of the
radiating loudspeaker. C,IIY = compliance of the cabinet volume. rM2 and
C M 2 = mechanical resistance and compliance of the suspensions of the secondary
loudspeaker. 11'12 and mA2 = masses of the cone and air load of the secondary
londspeaker. C MVl = compliance of the volume between the two loudspeakers.
fW2 ~ mechanomotive force in the voice coil of the secondary loudspeaker.
DIRECT RADIATOR LOUDSPEAKERS 159
of the cabinet can be increased by making the diameter of the driving loud
speaker small. By this expedient the resonant frequency of the driving
loudspeaker and cabinet can be made lower than the radiating loudspeaker
and the cabinet. By this two-stage system the response in the low-fre
quency range can be extended. A IS-inch loudspeaker is the radiating
loudspeaker in the system of Fig. 6.30. The response-frequency charac
teristic of the duo-cone loudspeaker mechanism in a back-enclosed cabinet
of two cubic feet is shown in Fig. 6.32. The response frequency charac
teristic of the compound direct radiator loudspeaker housed in the same
30
25
cn20 , ---...
l'
....:
o
!!:
ljll5
lJV', ,
/
z
oQ.
...
<II ,/A
0::10
cabinet is also shown in Fig. 6.32. It will be seen that a substantial gain in
low-frequency response is obtained by the use of the compound direct
radiator loudspeaker system.
6.12. Acoustical Phase Inverter Loudspeaker.-The acoustical phase
inverter loudspeaker18 system consists of a direct radiator loudspeaker
mechanism mounted in a completely closed cabinet save for a port coupling
the cabinet volume to the air, Fig. 6.33. The phase of the velocities on
the two sides of the cone differs by 180°. Referring to the mechanical
network of Fig. 6.33, it will be seen that the velocities in the branches 1
and 2 may differ by as much as 180° for positive mechanical reactances and
no mechanical resistances in branches 1 and 2 and a pure compliance in
18 Dickey, Caulton, and Perry, Radio Engineering, Vol. 8, No.2, p. 104, 1936.
160 ACOUSTICAL ENGINEERING
., 24
A
30
]
f+-18',-!
0
C
30
of+- 18'-+1
",24
o V\ o
V r- .......
'" 24
0
iYO ~
'" 18 ~ 18 w 18
<II
'"
z z
o
z
0 2J1
~ 12 "- 12 "-
<II
12
/
'"0:'" 6 "''"
0:
6
w
cr
6
3
V
o o
~o 100 500 30 100 500 30 100 500
FREQUENCY FREQUENCY FREQUENCY
CROSS-SECTiONAL VIEW
FIG. 6.35. Cross-sectional view, electrical circuit and mechanical network of a phase
inverter cabinet equipped with a drone cone. In the electrical circuit, e = internal
voltage of the electrical generator. rEO = internal electrical resistance of the generator.
Land rEe = inductance and electrical resistance of the voice coil. ZBM = motional
electrical impedance. In the mechanical circuit, me and mA = mass of the cone and voice
coil and air load. eMS and rMS = compliance and mechanical resistance of the suspen
sion system. rMA = mechanical resistance of the air load. CMF = compliance of the
cabinet. m'A = mass of the air load upon the drone cone. r'M A = mechanical resistance
load of the air load upon the drone cone. r'M8 = mechanical resistance of the suspension
of the drone cone. m'e = mass of drone cone. C'MS = compliance of the suspension of
the drone cone. (After Olson, Preston, and May.)
area of the drone cone can be made the same as the active cone. The phase
and amplitude of the particle velocity are the same over the entire area of
the drone cone. Furthermore, the particle velocity is relatively low because
the area of the drone cone is large compared to a port. As a result, the losses
are low in the drone cone phase inverter.
A typical response-frequency characteristic of the drone cone phase
inverter is shown in Fig. 6.36. Two response-frequency characteristics of
the same loudspeaker in the same cabinet but with two different ports are
also shown in Fig. 6.36. In one case the frequency range of the port is the
same as the drone cone, but the output which is obtained with the port is
30
25
.--<
D\:E CONE
32 SQ. IN. PORT
~--
:;<
m20
o ,-::--..:.
~
10 SQ. IN. FORT
::lz 15
~
!/l
'" 10
a:
lower. In the other case the output of the port is the same as the drone
cone, but the frequency range which is obtained is less. To summarize,
these characteristics show that a wider frequency range with greater output
can be obtained with the drone cone type of phase inverter, as compared
to the port type, the reason being that the losses in the drone cone are less
than in the port.
6.14. Acoustical Labyrinth Loudspeaker.20-The acoustical labyrinth
loudspeaker consists of an absorbent walled conduit with one end tightly
coupled to the back of the cone of a direct loudspeaker mechanism and the
other end opening in front or at the bottom of the cabinet within which it
is folded (Fig. 6.37). The labyrinth is a piston driven tube with absorbing
walls. At the first half wavelength resonance, the velocity at the open
end is in phase with that of the front of the cone. The radiation, then,
20 Olney, Benj., Jour. Acous. Soc. Amer., Vol. 8, No.2, p. 104, 1936.
DIRECT RADIATOR LOUDSPEAKERS 163
from both sources is additive and the response is increased. An increase
in response can be obtained over about an octave. The rising absorption
of the tube lining with increase in frequency damps out the higher reso
nances. The transmission through the tube is very low above 150 cycles.
An anti-resonance occurs when the tube is one-quarter wavelength long.
The deleterious effect of the fundamental resonance of the cone with its
suspension system upon the response may be eliminated by choosing the
constants so that fundamental resonance of the loudspeaker coincides
with the quarter wavelength anti-resonance of the tube. The pressure
response characteristic of a direct radiator loudspeaker with and without a
labyrinth is shown in Fig. 6.37. These characteristics show that the
accentuated response due to cabinet resonance has been eliminated and that
the low-frequency range has been extended.
A
5
:" \A
.,20
C
)
J \. ,''v " ~
..,1 5
'"
Z ~ "
210
..,'"
,,
cr 5
A' ..
o
~20
B
~ ~rHn---~~-H~+-~~-+++~~H
mtWJ
Q.
~ IO~~H---~r1-H~+-~~-+++ITH--H
a:
FIG. 6.38. Sectional views of the combination horn and double-voice coil, direct
radiator loudspeaker. In the voice coil circuit, e = the internal voltage of the generator.
rEG = the internal electrical resistance of the generator. rEI and Ll = the electrical
resistance and inductance of the large coil. rE2 and L2 = the electrical resistance and
inductance of the small coil. C E = electrical capacitance. In the mechanical network,
ml = the mass of the large coil. m2 = the mass of the small coil. C M2 and rMa = the
compliance and mechanical resistance of the corrugation separating the large and small
coils. ma = the mass of the cone. rMS and CMS = the mechanical resistance and com
pliance of the suspension system. m4 and YM2 = the mass and mechanical resistance of
the air load on the front of the cone. CMl = the compliance of the chamber behind the
cone. ZMl = the mechanical impedance at the throat of the horn. hill and JM2 = the
mechanomotive forces generated in the large and small voice coil sections. The graph
shows the pressure response frequency characteristic of the combination horn and direct
radiatorloudspeaker. The overlap between the horn and direct radiator action is shown
by the dotted and dashed characteristics. (After Olson and Hackley.)
LOUDSPEAKER
SECTION B- a'
LOUDSPEAKER
SOUND
PATH
INPUT INPUT
A'
rRONT VIEW A SECTION A-fI< rRONT VIEW B SECTION A-fI<
FIG. 6.41. Combination horn and direct radiator loudspeakers. A. A direct radiator
loudspeaker is used for the reproduction of the low-frequency range; and a horn loud
speaker, in which the pole and low-frequency cone form the horn, is used for the repro
duction of the high-frequency range. B. A direct radiator loudspeaker is used for the
reproduction .of the low-frequency range and a cellular horn loudspeaker is used for the
reproduction of the high-frequency range.
DIRECT RADIATOR LOUDSPEAKERS 167
A modification of the system of Fig. 6.41A, shown in Fig. 6.42, includes
a diverging acoustic lens 24 ,25 at the mouth of the high-frequency horn unit.
Referring to Sec. 2.19, it will be seen
that the high-frequency radiation from
a simple horn is emitted in a relatively
narrow beam. It has been shown in
Sec. 1.11 that a diverging acoustic lens
will spread out the impinging wave
front. An acoustic lens placed in the
mouth of the high-frequency horn will
increase the spread of the high-frequency
radiation and thus provide improved
directivity.
6.16. Loudspeaker Mechanisms
for Small Space Requirements. 26
There are many applications in the field
of sound reproduction where space is
limited and the volume occupied by
the loudspeaker must be kept to a
mmlmum. This is particularly the
case in pocket type personal radio
receivers where the loudspeaker is the
largest single component. For these
FIG. 6.42. Combination horn and
applications, an inverted magnetic direct radiator loudspeaker equipped
field structure has developed as shown with a lens in the mouth of the high
in Fig. 6.43. The field structure con frequency horn.
sists of three parts, namely, the top and
bottom plates and the magnet. The cone is located between the top and
bottom plates. The top and bottom plates are perforated for efficient
transmission of sound. The performance of the loudspeaker is the same
OPENINGS OPENINGS
as that of the conventional design with a cone of the same diameter and
weight and a magnet of the same weight. An examination of Fig. 6.43
reveals that the space occupied by the inverted-magnet field structure is
considerably less than in the case of the conventional design of Fig. 6.1.
6.17. Feedback Applied to a Loudspeaker.-Feedback in a transmis
sion system or a section thereof is the returning of a fraction of the output
to the input. Negative feedback is feedback which results in decreasing the
amplification. Among the sources of nonlinear distortion and nonuniform
response in a reproducing system may be the power amplifier and loud
speaker. It is possible to reduce distortion and improve the res.ponse as a
function of the frequency of an amplifier by making the amplification
deliberately higher than necessary and then feeding the output back in
such a way as to throwaway excess gain. In the same way this system
may be made to include the loudspeaker. It is not an easy proposition
..
0
5
I V\ 11
V ~I\
30
m
Oz5
3
'"
"'ZO
z
o
"
, - --- -~
:::1 5
0::
0
V
5
y'
A B
0
'10 100 1000 10000
FREQUENCY IN CYCLES PER SECOND
FIG. 6.44. Loudspeaker and amplifier feedback systems. A. The output of the pickup
coil is fed into the input side of the amplifier. B. The output of the microphone is fed
into the input side of the amplifier. The graph shows the pressure response frequency
characteristic of System A: 1. Without feedback. 2. With feedback. 3. With feed
back and compensation.
to employ feedback in this way because of the very special control required
of phase shifts in the amplifier and loudspeaker system. Unless certain
phase relations 27 ,28 are maintain~d, oscillation will occur. Fig. 6.44 shows
feedback systems applied to an amplifier and loudspeaker. In Fig. 6.44A
a pickup coil is attached to the cone. The output from the pickup coil is
fed into the input of the amplifier out of phase with the signal input. The
response characteristic of the amplifier loudspeaker without feedback is
shown in Fig. 6.44. The same loudspeaker and amplifier with 15 db nega
tive feedback from the pickup coil are also shown in Fig. 6.44. It will be
seen that the response at the high frequencies is improved. This system
tends to drive the cone at constant velocity for constant signal voltage
27 Nyquist. R .• Bell Syst. Tech. Jour .• Vol. 11. No. 1, p. 126. 1932.
28 Black, R. C.• Bell Syst. Tech. Jour., Vol. 13. No. 1, p. 1, 1934.
31 Olson. Preston, and May, Jour. Aud. Eng. Soc ., Vol. 2. No.4. p. 219. 1954.
32 Olson. H. F .• Radio and Television News. Vol. 45. No. 5. p. 53. 1951.
33 Olson. Preston, and May. Jour . Aud. Eng. Soc .• Vol. 2, No.4, p. 219. 1954.
30
25
CD 20
r-... ..,......r,:: r-..,,t:. ~
, ~ ~r-
\\ / '~
o I I
~
\ IA I I
llll5
z
\I / II
oQ. V ~
"
'"'"
a: 10
o
1000 2000 3000 4000 6000 8000 10000 15000
FREQUENCY IN CYCLES PER SECOND
I 1 I :
~'~1
1)4
/~.\
~
./1
L~I
~tS3 / '\:~I \ LS4'~
I I \ I
I / \ ~
I~~__ \ 1
TOP VIEW
r; C"1
I
f
I'"
I )J
t LS3
/;:1:
I,
~
LSI
LS4
'1'1I
l~
1 I
I LS2 I
~o~_ o
-'
1
TABLE MODEL
, I
: I
l LS4 \ LssA
r'", I \ r( I
t).! / \ . . '~
: / \ I
IL~Ls6 ~~{LS7 F~{~J
r.t~.c""_":>':\f.~~
TOP VIEW
SIDE VIEW
I
I
:I :
I
) I I
\
I \ I
/ \ I
1 \ I
/ \ I
I \ I
/ \ I
/ \ LS3 I
1 LSI I
1 \
\ L~2YA I
ItI-------- ~\.. L2:/ '~
TOP VIEW
I
I t~I,~-~~
I O~-\
I '\_-_/
===1
I
I
I
I
I
I
I
;{/ 0
o, - ' ~~,
LSI
\==:1
-I
'...,..,."
I L53 J
L _______________
and LS5 and the high frequency loudspeaker mechanisms may be placed
in either one or both of the locations LS6 and LS7.
The two most common console type phonograph cabinets and the loud
speaker locations in these cabinets are shown in Figs. 6.52 and 6.53. Seven
loudspeaker locations are shown in Fig. 6.52. The locations LSI, LS2,
LS3, LS4, and LS5 are employed for mechanisms with full frequency range.
A single loudspeaker at location LSI is the most common arrangement.
Other arrangements are as follows: Two loudspeaker mechanisms LS2 and
LS3 are placed in the corner and angled to increase the coverage. The two
loudspeaker mechanisms LS4 and LS5 are placed in the two sides of the
TOP VIEW
r -l
i "'----------:::l,
~=~-~.;~:-.:~=-~~----4
t,', /'" I
I ' l ' ,-~ I
';J - / I
" \ )lLS!5 I
~~--~'::-.::::-
L 0 0
-----~J
CHASSIS
SIDE VIEW
I: ._--.... iI i
~
1,
'" 'II
/TURNTABLE\II I
I'
1
, I N I I
1 n 0 1/1 n
L\ I
i/
I
).
LS4
\
,
" /
1/
(/; ' LS5~ 1
'i
ILS2 ~... LSI ......... __ -- r-*S3 I
1.,./-'"-", LS6 "';::\..... LS7 / , - .... I
L ---~Y- ~~------~
TOP VIEW
r
I
1
I
I
TOP VIEW
1-;;;:;'- - ----1
I III '1
1==;=:0=·:::----~1=-..::I1 \~Y
1 L.3~~<?..._J IILS2
I
TURNTABLE I
_C...::.:-:::rr==-:l__ I
r =
=. -=--...::.::.::: =:.."" I
1_ _
FRONT VIEW
mechanism can be placed at a greater distance from the floor. It has been
established that the most natural sound reproduction is obtained when the
loudspeaker is located at a distance from the floor corresponding to the ear
level of the listener. In the cabinet of Fig. 6.53 the most common
arrangement is that of a single loudspeaker mechanism covering the entire
frequency range and placed near the top of the cabinet. In another
arrangement a high-frequency loudspeaker mechanism LSI is used to cover
the low-frequency range. Another loudspeaker location is LS3. This loud
speaker may be used in various combinations with LSI and LS2. For
example, in one arrangement, each of the three loudspeakers covers a section
of the frequency range. In another combination LS2 may be omitted and
LSI and LS3 may be operated in parallel.
6.22. Loudspeaker Locations in Radio Receivers.-Radio receivers
may be classed as follows: console, table, portable, personal, and automobile.
Each of these types requires a different arrangement and location for the
loudspeaker mechanism.
In general, the loudspeaker mechanism locations in the console type
radio receiver are the same as the console type phonograph. See Sec. 6.21
and Figs. 6.52 and 6.53.
The six most common locations for loudspeaker mechanisms in table
model radio receivers are shown in Fig. 6.54. The most common arrange
ment of the loudspeaker mechanism in a table model radio receiver is the
use of a single loudspeaker in any of the locations LSI, LS2, LS3, LS4, LS5,
or LS6. In a multiple arrangement, two loudspeaker mechanisms LS2 and
LS3 are placed and angled to increase the coverage. In another multiple
arrangement two loudspeaker mechanisms LS4 and LS5 are placed in the
two sides of the cabinet. In a multiple arrangement of three loudspeaker
mechanisms, the loudspeaker mechanisms LSI, LS4, and LS5 are placed
in the front and the two sides of the cabinet, respectively, to obtain wide
angle coverage. In another multiple arrangement of three loudspeakers,
the loudspeaker mechanisms LSI, LS2, and LS3 are used.
Since the cabinet size of a table model receiver is relatively small, the
response falls off rapidly in the low-frequency region. See Sec. 6.9. Suit
able compensation may be employed in the audio amplifier to maintain the
low-frequency response. Employing this expedient, the limitation is the
maximum allowable excursion of the loudspeaker cone. The amplitude
of the cone for a certain sound power output is inversely proportional to the
area of the cone. Thus, it will be seen that the low-frequency response that
can be obtained is a function of the total radiating surface.
Personal radio receivers employ very small enclosures in the form of a
rectangular parallelopiped. The cubical content of the smallest receivers
is of the order of 25 cubic inches. The loudspeaker mechanism is usually
placed in the front panel of the radio receiver. The order of power available
for feeding the loudspeaker is of the order of 5 to 50 milliwatts. Since the
power is limited to a very low value and the loudspeaker mechanism and
cabinet are both small, it is impossible to obtain adequate low-frequency
DIRECT RADIATOR LOUDSPEAKERS 177
response for natural sound reproduction. In the personal radio the principal
use is for the reproduction of speech. Therefore, the acoustical problem is
to obtain good intelligibility on speech together with adequate sound level
output. This can be accomplished in the personal radio receiver by proper
and suitable design of the loudspeaker and case. See Sec. 6.9.
The acoustic problem in the larger portable radio receiver is similar to
the personal radio receiver. Of course, the larger the case and loudspeaker
the greater the low-frequency response.
Automobile radio receivers are housed in an enclosure of about 1000
cubic inches or less. In general, a loudspeaker mechanism is also placed in
the same housing. In some cases the radio chassis and loudspeaker are
I
'--~/ '\..i,,;--..J
_...I
I
I(
I,
1,_"
~""=-~l5
\8
I 0
I
1
I
I
TOP VIEW
l -.::......,.. .!
FRONT VIEW SIDE
A VIEW
r~ ~
r~\)d:':-,
I : ~!
I tli.... . _ j
1
Lr' _____
LS4 _
I
1
L
0 0 I
I
,z
CHASSIS
® ®
®
® ® ® ® ®
® ®
® ® ® ® ® ® ®
® ® ®
®®® ®
® ®
® ® ® ® ®
®
® ®®® ® ®
® ®
® ® ® ® ®
®
® ® ® ® ®
® ®
® ® ®
FRONT VIEW SECTIONAL VIEW
where x = displacement,
/MO = mechanical driving force, in dynes,
m = total mass, in grams,
eM = compliance of the suspension system, m centimeters per
dyne, and
YMT = total mechanical resistance, in mechanical ohms.
The total mechanical resistance is
YMT = YMS + YMR + YME 6.18
where YMS = mechanical resistance due to losses in the suspension system,
etc., in mechanical ohms,
YMR = mechanical radiation resistance, in mechanical ohms, and
YME = mechanical resistance due to the electrical system, in mechan
icalohms.
The mechanical resistance, YMS, is the sum of all the losses in the suspension,
the viscosity of the grill and cloth coverings and the viscosity loss due to the
air forced through the slit formed by the air gap and voice coil.
From equation 6.10 the mechanical impedance, ZME, due to the electrical
circuit is
(Bl) 2
ZME = YME = - 6.19
YET
where B = flux density, in gausses,
l = length of the voice coil conductor, in centimeters,
rET = YEC + YEG,
rEa = damped electrical resistance of the voice coil, in abohms, and
YEG = internal electrical resistance of the generator (the vacuum
tube), in abohms.
The inductance of the voice coil is negligible. The mechanical radiation
resistance, rMR, is given by the equation 5.10. It may be obtained directly
from the graph of Fig. 5.2.
180 ACOUSTICAL ENGINEERING
The mass, m, is the sum of the cone mass and the mass of the air load
upon the cone. The mechanical mass reactance of the air load upon a
cone may be obtained from equation 5.10. It may be obtained directly
from the graph of Fig. 5.2.
Heaviside's un extended problem 36 ,37,38 is as follows: Given a linear
network of n meshes in a state of equilibrium, find its response when a
unit force is applied to any mesh. The unit function is defined to be a
force which is zero for t < 0 and unity for t > O.
The indicial mechanical admittance of the mechanical circuit of Fig. 6.1 is
1
A(t) = 1 6.20
YMT + mp + GMp
where p is employed as a symbol for the differentiation with respect to the
independent variable, time.
Let
YMT
a = 2m
w = Jm~M - a2
Fig. 6.56 shows the effect of the electrical impedance of the vacuum tube
upon the transient response of a loudspeaker. In this case the loudspeaker
is connected to the following generators: a very high electrical impedance
corresponding to pentode or Class" B" operation; a generator of one-half
the electrical resistance of the loudspeaker corresponding to class "A"
operation; and to a generator of :very low electrical impedance corresponding
to inverse feedback operation. The electrical impedance characteristic of
the loudspeaker is shown by the uppermost left-hand graph of Fig. 6.57.
This example shows that the damping exerted by the electrical system is
of consequence. However, there is very little difference between Class
" A" and feedback operation. When a loudspeaker operates from a high
36 Carson, " Electric Circuit Theory and Operational Calculus," McGraw-Hill Book
Company, New York, N.Y., 1926.
37 Bush, "Operational Circuit Analysis," John Wiley and Sons, New York, N.Y.,
1937.
38 Berg, "Heaviside's Operational Calculus," McGraw-Hill Book Company, New
York, N.Y., 1936.
DIRECT RADIATOR LOUDSPEAKERS 181
electrical impedance vacuum tube amplifier, the internal mechanical resist
ance of the loudspeaker is the major factor influencing the transient response.
Fig. 6.57 shows response of a 12-inch (to-inch diameter cone) loudspeaker
to a unit force for various values of mechanical resistance. In order to
correlate the response with actual systems, the electrical impedance frequency
characteristic for each system is also shown. These characteristics are for
a loudspeaker coupled to a generator with very high internal electrical
!~?MfHlH
A
!ff~!-tlllill
B
iJtH j-IIIIII
o
TIME
0.05
IN SECONDS
40
20
f-r-"'"
V
100
r-
300
!twi4+HffH
t 1.00 0.05 0.1
!~N1d+H J \ \ I
z
20
1---
100 300 cl 1.00 0.05 0.1
20
I - -I--"
~o
FIG. 6.57.
100
rREQUENCY
300
iJtH IIIII11
0
TIME
0.05
IN SECONDS
speaker, with a 12-inch diameter cone, for different values of the suspension
mechanical resistance. The electrical impedance frequency characteristic
indicates the degree of internal damping.
system shown in Fig. 6.74A, it will be seen that there is a peak in the response
at 800 cycles and a dip at 1100 cycles. The sound outputs from this loud
speaker with applied tone burst signals of 800 cycles and 1100 cycles are
shown in Fig. 6.58. In the case of the peak in the response frequency
characteristic there is a slow growth and a slow decay in the response of
the loudspeaker to an applied tone burst signal. In the case of a dip in the
response frequency characteristic there is a rapid growth followed by a
decrease in output then followed again by an increase in output in the
response to an applied tone burst signal. It has been suggested that a
39 Olson, H. F., Audio Engineering, Vol. 34, No. 10, p. 5, 1940.
40 Olson, Preston, and May, Jour. Aud. Eng. Soc., Vol. 2, No.4, p. 219, 1954.
41 Corrington, M. S., Jour. Aud. Eng. Soc., Vol. 3, No. 1, p. 35, 1955.
-+ ---1
TRANSIENT DISTORTION TRANSIENT DISTORTION
FIG. 6.58. The transient response of a loudspeaker having the
response frequency characteristic of Fig. 6.74A at 800 cycles
and 1100 cycles. The waves show the tone burst input to the
loudspeaker, the sound output from the loudspeaker, and the
output from the loudspeaker after the input to the loud
speaker has been stopped.
43 Olson, Preston, and May, Jour. Aud. Eng. Soc., Vol. 2, No.4, p. 219,1954.
generated in the cone when the operation exceeds the linear portion of the
stress-strain characteristic of the cone. The strength of the cone depends
12 upon the thickness of the paper.
The weight in turn is proportional
to the thickness of the cone. The
9
.......
sound output of a typical direct
III
.,.;'
radiator loudspeaker for a certain
o
... 6
/ value of nonlinear distortion as a
V o
1.0 1.5 2.0 2.5
RELATIVE CONE MASS
function of the weight of the cone
is shown in Fig. 6.4. A review of
3.0 Figs. 6.4 and 6.59 shows that high
sensitivity by the use of a light
FIG. 6.59. The sound output of a typical cone is not compatible with low
direct radiator loudspeaker, for a certain fixed
value of nonlinear distortion, as a function of nonlinear distortion. In order to
the mass of the cone. obtain low nonlinear distortion, a
relatively heavy cone must be used.
B. Nonlinear Suspension System. 49-The force displacement charac
teristic of a typical, direct radiator loudspeaker cone suspension system
X
DISPLACEMENT
is shown in Fig. 6.60. It will be seen that for small amplitudes the suspen
sion system is linear. However, for large amplitudes the suspension system
is nonlinear.
44 Olson, H. F., Jour. Acous. Soc. Amer., Vol. 16, No. 1, p. 1, 1944.
DIRECT RADIATOR LOUDSPEAKERS 185
The force defiection characteristic of the loudspeaker cone suspension
system of Fig. 6.60 may be approximately represented by the expression
/M = /(x) = ax + f3x 3, 6.23
where a = constant> 0, f3 = constant> 0, and /M = applied force which
produces the displacement x.
The compliance of the suspension system of Fig. 6.60 may be obtained
from equation 6.23 as follows:
x 1
eM = JMT =
a
f3 2
x+ 6.24
F = Eli,
w = 21r/,
/ = frequency, and
t = time.
Substituting the expression for eM of equation 6.24 in equation 6.25,
the differential equation becomes
mx + YMX + aX + f3x 3 = F cos wt 6.26
Since the mechanical resistance, YM, is quite small compared to the
mechanical reactance, save over a very narrow frequency range near the
resonant frequency, equation 6.26 can be written as follows:
mx + aX + f3x 3 = F cos wt. 6.27
A number of investigators have obtained an approximate solution of this
differential equation.
If f3 is considered to be small, the relation
w2 =~ + !f3A 2 _ £ 6.28
m m Am
between the arbitrary amplitude A and w may be obtained,
186 ACOUSTICAL ENGINEERING
The sections which follow will show that these equations predict the
perfonnance of a loudspeaker with a nonlinear cone suspension system.
C. Distortion Characteristics of Nonlinear Suspension System.-The well
known experimental result of a nonlinear cone suspension system is the
production of odd order harmonics when a sinusoidal input is applied to the
loudspeaker. The wave shape under these conditions is shown in Fig. 6.61.
FIG.6.61. The wave shapes of the electrical input and the sound pressure
output of a loudspeaker with a nonlinear suspension system.
5Il. zo
\ \ olWIAITTS
~ 15
\ ~WIATTS
z \ \ ' ~~ATT
2 10
I-
ao
1\ ~
~ 5
ell
Q "- I"- ...I' ~
~o 100 400 1000
fREQUENCY IN CYCLES PER SECOND
CROSS-SECTIONAL VIEW
-
f
FIG. 6.63. A system consisting of a mass ml driven by a crank at a
frequency f and a second mass m2 supported by a spring coupled to
ml vibrating with a frequency tf. The cross-sectional view of the
cone shows a similar system and illustrates how subharmonics may be
produced by a loudspeaker.
2
jfr an increase in frequency is shown in
Fig. 6.65. The velocity frequency charac
/ l~ teristic for a decrease in frequency, is also
shown in Fig. 6.65. These characteristics
I
""
~ ~--.!. are quite similar to the theoretical charac
30 60 100 200 teristic of Fig. 6.64.
fREQUENCY E. Distortion Due to Inhomogeneity of
FIG. 6.64. Theoretical response the Air-Gap Flux.-Inhomogeneity of the
frequency characteristic of a direct flux density through which the voice coil
radiator loudspeaker mechanism moves is another source of distortion.
with a nonlinear suspension system.
The unstable portion of the response The result is that the driving force does
frequency characteristic is indicated not correspond to the voltage developed
by a dashed line. by the generator in the electrical driving
system. Furthermore, the motional elec
trical impedance is a function of the amplitude. This type of distortion is
similar to that due to a nonlinear suspension system. The wave-shape dis
tortion is similar to that of Fig. 6.61.
The force, in dynes, developed by the interaction of the current in the
voice coil and the magnetic field is
f= Bli 6.30
where B = flux density, in gausses,
1 = length of the voice coil conductor, in centimeters, and
i = current, in abamperes.
Equation 6.30 shows that the force is directly proportional to the current
if Bl is a constant. If the Bl product varies with the position of the voice
DIRECT RADIATOR LOUDSPEAKERS 189
8 8
7 7
6 6
5 5
~ 1/ >
t
t
U4 U4
o
g II -'
w
w
>3 > 3
/ I
j j
2 2
I
/ I
V \
"" V r----_
30
V
60 100
'-- 200 ~o
V
60 100 200
fREQUENCY fREQUENCY
~ ~ 00
FIG. 6.65. Experimental response frequency characteristic of a direct radiator
loudspeaker mechanism with a nonlinear suspension system for an increasing and
a decreasing applied frequency.
coil, the force will not be proportional to the current and distortion will
be produced.
A typical flux distribution in an air gap is shown in Fig. 6.66. A con
sideration of the flux distribution shows that the Bl product will be practi
cally a constant if the voice coil is made longer than the gap, as shown in
Fig. 6.66B, because, as the coil moves into the weaker tufting field on one
end, it moves into a stronger field on the other end. From the standpoint of
efficiency at the higher frequencies, this method is not particularly desirable
11)20
'"
::J ~ ~
'"" 16
If \
3
'" 12
>- 8
/ AIR GAP
AXIAL \
...
~
w V
4 L LENGTH
I \
o
x 0
3 -;4 ,2 0 .2
\ ,.
" AXIAL LENGTH IN INCHES
FIG. 6.66. Graph of the flux distribution in an air gap. A. Typical field map
of the flux lines in an air gap. B. A voice coil longer than the air gap. C. A
voice coil shorter than the air gap.
190 ACOUSTICAL ENGINEERING
because part of the voice coil is in a weak field. This type of distortion
can also be eliminated by making the air gap of sufficient axial length so that
the voice coil remains at all times in a uniform field as shown in Fig. 6.66C.
The latter method is used for high-frequency loudspeakers of high efficiency.
F. Frequency-Modulation Distortion. 45 -The amplitude of the cone in a
direct radiator loudspeaker for constant sound output, in the frequency
range below the ultimate radiation resistance, is inversely proportional to
the square of the frequency. If the cone is radiating both at high- and low
frequency, the source of high-frequency energy may be considered to be
moving back and forth at the low frequency. The high-frequency energy
will be modulated. The resulting frequency-modulated wave may be
represented by a carrier and a double infinity of sidebands.
The square root of the ratio of the power in the sidebands to the total
power in the sound wave, in per cent, is
. 8r----+----+----++----I
..,..
u 8 u
Q. 6 / w
Q. 6r----+----+----++---:Il
IWAT~V ~
z 4 z 4f---t-----t----t--h'<---j
- Y ~~
o o
j: j:
.. 2 ~ 2r----+-----r-~~:
o
>
VI
I--"V~ ~I--" ....
5 ~02 4 a 10' ~ o~-==:F-~~"'*::t:;:::::::=-!
• 104 0 103 2
FREQUENCY
Equation 6.32 applies to any diverging wave system in which the sound
pressure varies inversely as the distance. It may be mentioned in passing
that the pressure in a sound wave in free space varies inversely as the
distance.
The second harmonic pressure,46 in dynes per square centimeter, gen
erated in a distance x, in centimeters, in a plane wave is
{y + l)w Pl2x
P2 = 6.33
2 V2ypoc
where PI = fundamental pressure, in dynes per square centimeter, and
the other quantities are the same as equation 6.32.
Equation 6.33 applies to"a plane wave, as, for example, a sound wave in
a pipe.
In the case of a direct radiator loudspeaker the wave is diverging. At
a distance equal to the radius of the cone the system can, from the stand
point of distortion, be replaced by a spherical radiator equal to the radius
of the cone. The distortion generated in the volume between the cone
and the spherical surface may be determined by approximations by em
ploying equation 6.33. The complete expression for the second harmonic
distortion, in per cent, generated between a cone of radius rl, in centimeters,
and an observation point at a distance r, in centimeters, from the front of
the baffle is
D = P2r 100
{y = +
l)wPl [.85r +
r loge ':.-] 100 6.34
Plr 2V2ypoc rl
where Plr = pressure at the observation point, in dynes per square centi
meter.
46 Thuras, Jenkins, and O'Neil, Jour. Acous. Soc., Amer. Vol. 6, No.3, p. 173, 1935.
192 ACOUSTICAL ENGINEERING
...z 2 /
w
U I / / /
.
..,It
.8
" /
V /.1
/ /
B .2 ~/// r-
,O~VV V.I
i=
a:
o .I
,0bQ~/ /
.02 L/ '//7,,~
.0 ,L// / /I
4
8 103 2 4
V/ V V
8 104 10 2 2 8 103
,REQUENCY IN CYCLES PER SECOND
A'
FRONT VIEW FRONT VIEW
~ ~ ~......-------~
A B c
FIG. 6.70. Sectional views of felted cones for direct radiator, dynamic loud
speakers. A. Conical shape. B. Flared Shape. C. Conical shape with corruga
tions.
propagation of the wave in the cone and thereby broadens the directional
pattern.
The three types of suspension systems, shown in Fig. 6.71, are in general
use. The leather or kidskin suspension system shown in Fig. 6.71A has
gradually gone out of use. It has been displaced by the one-piece felted
cone and suspension system shown in Fig. 6.71B. The latter system is
much simpler and less costly in manufacture. One of the principal dis
advantages of the felted suspension system is the nonlinear characteristics
which introduce distortion (see Sec. 6.16). The stiffness of the suspension
system may be decreased and the distortion reduced by means of a folded
or double suspension system as shown in Fig. 7.61C. The reduction in
194 ACOUSTICAL ENGINEERING
A B c
FIG. 6.71. Sectional views of cone suspension systems for direct radiator loudspeakers.
A. Leather suspension. B. Felted corrugated suspension. C. Folded or double-felted
corrugated suspension system.
the damping ring can be compared with that of Fig. 6.74B taken with the
damping ring. It will be seen that the response frequency characteristic
without the damping ring exhibits a peak and a dip at 800 and 1100 cycles,
respectively.
In the direct radiator loudspeaker the sound vibrations start at the voice
coil, flow out in the cone, and then into the suspension system. In the low
frequency range the phase shift, in degrees, along the cone is relatively small
and the cone behaves essentially as a piston. However, in the high-fre
quency range the phase difference between the voice and suspension may
be several radians. In this frequency range it is important that the wave
that travels into the suspension system is absorbed and not reflected back.
The latter condition would lead to standing waves which would produce a
ragged response frequency characteristic. The sponge-rubber damping ring
serves as a suitable acoustical termination, thereby absorbing the vibrations
which flow into the suspension system. The response frequency charac
teristics without and with the damping ring are shown in Figs. 6.7SA and
6.7SB, respectively. It will be seen that response is smoother with the
damping ring.
47 Olson, H. F., Radio and Television News, Vol. 51. No.2, p. 69,1954.
48 Olson, Preston, and May, Jour. Aud. Eng. Soc., Vol. 2, No.4, p. 219, 1954.
30
25
, .. ~
-
AI' \ B
I
-\ /1 ~
rJl20
0 ~
/
I ~ -
!
III
III 15
j
2
0
11 I
D-
...
c/)
11:10
\j
y"
o
400 600 800 1000 2000 3000
FREQUENCY IN CYCLES PER SECONO
25
L. -.
~ ~'rr<7 '_F~
~~
'A "
o
2000 3000 4000 6000 800010000 15000
FREQUENCY IN CYCLES PER SE.COND
Centering suspensions for keeping the voice coil aligned in the air gap
are shown in Fig. 6.76. An inside slotted centering suspension, usually
made of fiber, is shown in Fig. 6.76A. An inside felted paper centering
DIRECT RADIATOR LOUDSPEAKERS 197
suspension with corrugations is shown in Fig. 6.76B. This type of suspen
sion is usually employed where the amplitude is small. An outside felted
paper suspension with corrugations is shown in Fig. 6.76C. The outside
INSIDE
SLOTTED INSIDE OUTSIDE
CORRUGATED CORRUGATED
CENTERING CENTERING DUST
01 SK DISK CAP
FIG. 6.76. Sectional views of voice coil centering systems for direct radiator dynamic
loudspeakers. A. Inside slotted fiber disk. B. Inside felted corrugated centering disk.
C. Outside felted corrugated centering disk.
A B c D
FIG. 6.77. Sectional views of voice coil constructions. A. Round enameled wire
wound on a paper form. B, C, and D. Self-supporting voice coils held together with
thermosetting cement. B. Round wire. C. Square wire. D. Edgewise wound ribbon.
A --~
'.~..""
SECTION A-~
t1.
$
A
I
-
1
I
Ii
c/>=M 6.35
R
where c/> = total lines of flux, in maxwells,
M = magnetomotive force, in gilberts, and
R = reluctance, no unit.
Equation 6.35 is analogous to Ohm's Law in electrical circuits expressed
as follows:
. e
2 = 6.36
YE
The quantities, units, and symbols in electrical and magnetic systems are
shown in Table 6.2.
TABLE 6.2
Electrical Magnetic
0.
T T
L-=-.
ffi R:
RL
CPe
MT
<p.
~
22
T..--L
MAGNETIC NETWORK
COIL
l\ TURNS
PERMANENT
MAGNET
SCHEMATIC VIEW SCHEMATIC VIEW
1>1 = M1 6.40
Rl
where Rl = reluctance of the air gap, and
M1 = magnetomotive force across the air gap in gilberts.
DIRECT RADIATOR LOUDSPEAKERS 203
The reluctance of the air gap is
1
RI = 6.41
Al
where 1 = length of the air gap in the direction of the flux in centimeters,
and
Al = cross-sectional area of the air gap, in square centimeters.
14
12
~ --....::
III
~~ ~~
'" D
III
~IO
"9 ( /Y
>:: 6 0/ ~\ 1)/
I ~
V
--
I
~ t-
~
;....
z 6 ~
-Irt- / ~~n J ~ V
l);;?
Q
r
u
J
04
~
V
/; ~
Z
/'
2
1/
o
600
/
V
700 600 500
Iii
400
DEMAGNETIZING fORCE IN
300
I 200
OERSTEDS
100 o
V 2 3
EXTERNAL ENERGY
4
B x H xlO-"
The curves 49 of Fig. 6.82 show the relation between the intrinsic flux
density and the magnetizing force for different magnetic materials. The
permeability may be obtained from the curves of Fig. 6.82 and the following
relation
B
fA-= - 6.43
H
where B = flux density, in gausses, and
H = magnetizing force in oersteds.
49 Kentner, A. E., Gen. Etec. Rev. XL V, No. 11, p. 633,1942.
204 ACOUSTICAL ENGINEERING
.-!-
V
Y /'/ ~
en
w 18
!3=B-H V %~
z
.J
9 16
j f-- ~ :::::: cY .,,
"zl4 VV k::: ~ V
/ J0 ///
<ll.. 12
)0 10
.,,- V ~f
IL /
7 V
/ V1 r/ • j /
t-
iii /
Z 8
w
a
/ ~
q: rI11" '0/
V
• I II
x 6
::J
.J
... 4 ~
r-
2
iI' r-- V
..? I- V I
II
Vi
z 2 ~ Yy / sf 'Ji v V
"j ';I I
it
t- ~ V.. . . / / ~ I' V
.- V I
~ 0
.01 4 8.1 8, 810 4 81022 4 81032
MAGNETIZING FORCE H IN OERSTEDS
ELECTRICAL NETWORK
Zltl2
STATIONARY DIAPHRAGM
PLATE
Assume that the polarizing voltage is eo and that the alternating voltage
is e = emaxsin wt. The force, in dynes, between the plates is
The first and third terms in the numerator of equation 6.49 represent
steady forces. The fourth term is an alternating force of twice the frequency
of the impressed voltage. The second term is an alternating force of the
frequency of the impressed voltage. If the polarizing electromotive eo is
large compared to the alternating electromotive force emaxsin wt, the fourth
term will be negligible. The useful force, in dynes, then is the second term
DIRECT RADIATOR LOUDSPEAKERS 207
which causes the moving surface to vibrate with a velocity which cor
responds to the impressed electromotive force.
fM = eoemax sin wt A = eoe
J, 4n-a2 4n-a2 6.50
The electrical motion impedance of this system will now be considered. The
charge, in statcoulombs, on the condenser is
q = CEeo 6.51
where eo = potential difference between the plates, in statvolts, and
CE = capacity per unit area, in statfarads.
The current, in statamperes, generated due to motion is
. dq
1 = dt 6.52
From the mechanical circuit of Fig. 6.83, the mechanical rectilineal im
pedance of the vibrating system is
ZM = ZMl +
ZM2 6.60
where ZM = total mechanical rectilineal impedance of the vibrating system,
in mechanical ohms,
ZMl = mechanical rectilineal impedance of the vibrating plate, in
mechanical ohms, and
ZM2 = mechanical rectilineal impedance of the load, in mechanical
ohms.
The mechanical rectilineal impedance at the plate is
ZM =
1M
-; 6.61
X
From equations 6.50 and 6.59
e 1617 2«41M
6.62
2 eo 2A2 i
From equations 6.60, 6.61, and 6.62
16172«4
ZEM = 6.63
eo 2A2 ZM
.---------lFiK
LOUDSPEAKER
PLATE
ELECTRICAL CIRCUIT
DIAGRAM
PERFORATED
BACK PLATE
METALLIC COATED
PLASTIC DIAPHRAGM
6
DIAPHRAGM
increased the efficiency because the mass of the air load was the major
factor in the mechanical impedance below 100 cycles. Using more than
4000 volts polarizing voltage leads to corona and other leakage effects.
These tests substantiate the theoretical consideration, namely, that the
operation of a practical electrostatic loudspeaker must be confined to the
upper portion of the audio frequency range, say above 1000 cycles.
A commercial electrostatic loudspeaker 53 for the frequency range above
7500 cycles is shown in Fig. 6.84. The diaphragm is a .001-inch plastic
with a thin metallic coating. The diaphragm rests directly upon the
perforated metal back plate. The electrostatic loudspeaker is coupled
53 Bobb, Goldman, and Roop, Jour. Acous. Soc. Amer., Vol. 27, No.6, p. 1128, 1955.
210 ACOUSTICAL ENGINEERING
directly to the vacuum tube; the plate supplies the polarizing voltage.
Since the effective spacing is about .001 inch, the sensitivity is comparable
to conventional dynamic loudspeakers. This small spacing limits the
maximum amplitude and confines the operation to the frequency range above
7500 cycles.
Front and sectional views and the electrical circuit diagram of a bilateral
or push-pull electrostatic loudspeaker54 is shown in Fig. 6.85. In commer
cial versions of this loudspeaker the area of the diaphragm is one square
foot. The spacing between the movable and fixed plates is about /6 inch.
INSULATOR A
roO-To 00-0-0-0-0-0-0-.0-01
DIAPHRAGM 10 0 0 0 0 0 0 0 0 0 0 0 0 0 0I
STATIONARV 10 0 0 0 0 0 0 0 0 0 01
PLATES 10 0 0 0 0 0 0 0 0 0 0 0 0 01
Ie. 0 0 0 0 0 0 0 0 0
10 0 0 0 0 0 0 0 0 0 0 0 0 01
POLARIZING 10 0 0 0 0 0 0 0 0 0 0 0 01
VOLTAGE 10 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
:0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 01
10 0 0 0 0 0 0 0 0 0 0 0 0 01
10 0 0 0 0 0 0 0 0 0 0 011
lo 0 0 0 0 0 0 0 0 0 0 0 0 01
,0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 01
10 0 0 0 0 0 0 0 0 0 0 0 01
AUDIO 10 0 01
INPUT
~ 0 0 0 _o_~!....~..2---..5'_o_~2....-!_oJ
A'
ELECTRICAL DIAGRAM SECTION A- A' FRONT VIEW
.)<I!!13t-t------T-+-y
x
FIG. 6.86. The geometry for obtaining the total sound
output of a loudspeaker.
Q =P AN 6.67
PAD
. where PAN and PAD are given by equations 6.65 and 6.66.
The directivity index can be computed from the directional characteristics
of the loudspeaker.
The directivity index of a loudspeaker can be obtained from the directivity
factor as follows:
DI DB = 10glO Q 6.68
7
HORN LOUDSPEAKERS
OF THE
FIG. 7.1 Cross-sectional view of a horn loudspeaker, the electrical circuit and mechanical
network of the vibrating system. In the voice coil ci.rcuit; e = the internal voltage of
thc generator. YEG = the internal electrical resistance of the generator. L = the
inductance of the voice coil. YED = the damped electrical resistance of the voice coil.
ZEM = the motional electrical impedance. In the mechanical network; mA and YMA =
the mass and mechanical radiation resistance due to the air load on the back of the
diaphragm. mo and mD = the masses of the voice coil and diaphragm. CMS and CMt
= the compliances of the suspension and air chamber. ZMH = the mechanical impe
dance at the throat of the horn. ZME = the mechanical impedance due to the electrical
circuit. 1M = the force generated in the voice coil. /MO = the force of the mechanical
generator.
where mA = mass of the air load on the back of the diaphragm, in grams,
me = mass of the voice coil, in grams,
mD = mass of the diaphragm, in grams,
rMA = mechanical resistance load on the back of the diaphragm, in
mechanical ohms,
CMS = compliance of the suspension, in centimeters per dyne,
CMl = compliance of the air chamber, in centimeters per dyne,
ZMH = rMH +jXMH = mechanical impedance of the throat of the
horn, in mechanical ohms,
rMH = mechanical resistance of the throat of the horn, in mechanical
ohms, and
XMH =
•
mechanical reactance of the throat of the horn, in mechanical
ohms.
HORN LOUDSPEAKERS 215
For initial efficiency considerations, the mechanical reactance of the
mechanical system is assumed to be negligible compared to the radiation
mechanical resistance, that is, mA, me, mD, CMI , IjCMs, and XMH are zero.
rMA is also negligible. Then
AD2
ZM = rMH = 42 - 7.4
AT
where AD = area of the diaphragm, in square centimeters, and
8
4 8
2
~ ~ 4
"~
w
z
!:! 10
U
i;:
...
I
0.5
0.25
Z
~IO
U
i;:
...
" 2
"' ["
0.5
I
0.1
I ~ 0.25
10 10000 10 10000
FIG. 7.2. The initial efficiency, in per cent, of a horn loudspeaker as a function of
AI)2/A1' for aluminum voice coils having masses of 0.1, 0.25, 0.5, 1.2, 4, and
8 grams and flux densities of 22,000 and 14,000 gausses. AI) and A l' are the areas
of the diaphragm and throat, respectively. in square centimeters. The above
graphs may be applied to a copper voice coil by multiplying the ratio AI)2/A l' by
one-half.
B. The Effect of the Mass of the Vibrating System upon the Efficiency.-In
the preceding section the mechanical reactance of the vibrating system
was assumed to be negligible compared to the mechanical resistance. The
mechanical mass reactance of the diaphragm and voice coil influences
the efficiency when this mechanical reactance becomes comparable to the
mechanical resistance. It is the purpose of this section to consider the
effect of the mechanical reactance of the vibrating system upon the effi
ciency.
The real part of the motional electrical impedance, equation 7.1, is
p. = (
rED rMH
2
+(Bl):;~~
XM + (Bl)2rMH X 100 7.13
This expression shows that the efficiency is a function of the flux density,
the coil mass and material, the diaphragm mass, the throat resistance, and
I 22000 GAUSSES COIL·DIAPHRAGM = I' I 10022000 GAUSSES COIL'DIAPHRAGM-I' 2-.
10 ~
\ \
I
20 100 1000 10000
I
20 100 1000
\
10000
FREQUENCY FREQUENCY
10 14000 GAUSSES
0
-.
COIL· DIAPHRAGM - 1'1
100
14000 GAUSSES COIL • DIAPHRAGM I' 2
~
r;
\~
z
!!!I 0
Il
...
b
"
\ \
FIG. 7.3. The efficiency. in per cent. as a function of the frequency of horn loud
speaker systems having ratios of voice coil mass to diaphragm mass of 1:2 and
1: 1. flux densities of 22.000 and 14.000 gausses. and initial efficiencies of 20. 40.
60. and 80 per cent for an aluminum coil. The above graphs may be applied to a
copper voice coil by multiplying the frequency by one-half.
the frequency. The efficiency characteristics for ratios of voice coil mass
to diaphragm mass of 1 : 1 and 1 : 2, and flux densities of 22,000 and 14,000
gausses for an aluminum voice coil are shown in Fig. 7.3. The character
istics of Fig. 7.3 are applicable to a copper voice coil by multiplying the
abscissa by 0.5. In order to connect with the characteristics of initial
efficiency of Fig. 7.2, these curves are depicted in terms of the initial
efficiency (20, 40, 60, and 80 per cent). These data show that it is a
218 ACOUSTICAL ENGINEERING
7.14
11 Hanna and Slepian, Jour. A.I.E.E., Vol. 43, No.3, p. 251, 1924.
12 Wente and ThUfas, Bell Syst. Tech. Jour., Vol. 7, No. 1, p. 140, 1928.
13 Olson, H. F., Jour. Acous. Soc. Amer., Vol. 2, No.4, p. 242, 1931.
14 Wente and Thuras, Jour. A.I.E.E., Vol. 53, No. 1, p. 17. 1934.
16 Smith, Bob H., Jour. Acous. Soc. Amer., Vol. 25, No.2, p. 305, 1953.
100 100
~
-"~
0
I
C B A
I
\C B
20 100 1000 10000 20 100 1000 10000
rREQUENCY fREQUENCY
100 100
\'
~
1\
I
20 100 1000
B f<'.
10000
I
20 100 1000
C B
~
10000
FREQUENCY fREQUENCY
FIG. 7.4. The efficiency, in per cent, as a function of the frequency of a voice coil
and diaphragm having a mechanical reactance of 1 mechanical ohm at 1000 cycles
coupled to a throat of a horn having a mechanical resistance of 1 mechanical ohm
and an air chamber having the following mechanical reactances: A. An infinite
mechanical reactance. B. A negative mechanical reactance of 1 mechanical ohm
at 2000 cycles. C. A negative mechanical reactance of 1 mechanical ohm at
1000 cycles for initial efficiencies of 20, 40, 60, and 80 per cent. These charac
teristics are applicable to other frequencies by multiplying the frequency by any
number and multiplying the mass and the compliance by the reciprocal of the
number.
abscissa by any number and, of course, multiplying the mass and the com
pliance by the reciprocal of the number. These characteristics are also
applicable for other values of mass and mechanical resistance by simply
multiplying these two factors by the same number and the compliance by
the reciprocal of that number. The characteristics shown in Fig. 7.4 have
included mass-compliance products which cover the useful range of values.
220 ACOUSTICAL ENGINEERING
where the motional electrical impedance, ZEM, in ohms, from equation 7.1, is
From the voice coil electrical circuit, Fig. 7.1, the total electrical impedance,
ZET,in ohms, at e is
ZET = rED + rEG + jwL + ZEM 7.17
where rED = voice coil resistance, in ohms,
Equation 7.16 shows that the throat acoustical resistance may vary over
wide limits without introducing large variations in the power output. As a
specific example, Fig. 7.5 shows the power output as a function of the
frequency lor a horn, having all dimensions two and one-half times
HORN LOUDSPEAKERS 221
that of Fig. S.l1B and driven 0
by a vacuum tube having the
constants indicated by the cap
tion of Fig. 7.5. Although the
variation in acoustical resistance 1--4
a:J
o
I 2 /
:>
Q.
\ /
...... --
is 6 to 1, the variation power out o:>
put is only 2 db.
E. The Effect of the Voice Coil
Temperature upon the Ejjiciency.l7
100 200 500
-The effect of the temperature fREQUENCY
of the voice coil upon the effi FIG. 7.5. Acoustical power output frequency
ciency is usually ignored in con characteristic of the horn (Fig. 5.11B with all
siderations of the performance dimensions multiplied by 2!) coupled to a
of a loudspeaker. In high power 10!-inch diameter, 10-gram diaphragm driven
loudspeakers, where the tem by a 5-gram aluminum voice coil in a field of
20,000 gausses. Damped electrical resistance
perature of the voice coil becomes of voice coil 20 ohms. Electrical impedance
quite high, considerable loss in effi of vacuum tube through a transformer,
ciency may result as will be shown 35 ohms.
in the discussion which follows.
The efficiency, in per cent, of a loudspeaker, when the temperature cor
rection is added, may be expressed.
60
!Z 40 --r-r-r-
w
U 30
a: ---- I-- I--
I-- I-- r-
- - -
--
w
0.. 20
-I--
~
>
u
3 10
r--
r--
r-- r-- I- -- - I-
--
r-
t--......
V
i:: 8
"
uJ 6
-
- -I-- i-
I'--
-
50 100 ISO 200 250 300
TEMPERATURE RISE IN DEGREES CENTIGRADE
gao
'" :2 80
3
10 900 1000 10000 10900 1000 10000
FREQUENCY fREQUENCY
THROAT THROAT
DIAPHRAGM DIA'81N AREA'SO 2SQ IN AREA DIAPHRAGM DIA=16IN AREA=201 SQ IN AREA
I
0 SQIN o so IN
z '0 6.28 ,""", ........ r-- 25.1
~2 0 I 2.6 50.2
a: "- 25.1 '\ "".1' I 00.
:;'4
0
50.2 t'- 201.
I00. 402.
'"
(g80 '"
"'8 0
o
..J --'
100100 10900 10000
1000 10000 1000
FREQUENCY fREQUENCY
FIG. 7.7. Characteristics depicting the loss in per cent of the total sound radiation,
due to the radiation of sound from the back of the diaphragm of a horn loudspeaker
for diameters of 2, 4, 8, and 16 inches and throat areas of 2,1. !, t, and t times the
diaphragm area.
19 Thuras. Jenkins, and O·Neil. Jour. Acous. Soc. A mer .• Vol. 6. No.3. p. 173. 1935.
20 Goldstein and McLachlin. Jour. Acous. Soc. Amer.. Vol. 6. No.4. p. 275, 1935.
The second harmonic distortion, at the mouth, in per cent of the funda
mental, generated in an exponential horn is given by
III
f-
!(
.1
8
'" " ..... ,
I......
"'"
""- 1010
~ 4
f-'
I
2
I"
"-
"-
I"-
'" '"
'" '"
::J
Q. • 3 "(0
!; 4
o
a:
z " I'
.....1'..
"
~1o
Q.
I"'
.00 I
I
2 4
FREQUENCY f
5 8 7 8 9 10 '"
CUTOFF
2
'-.
FREQUENCY
f'..
I"
of 1, 3, and 10 per cent distortion is shown in Fig. 7.8. For the sake of
generality the curves shown in Fig. 7.8 refer to an infinite horn. How
ever, the increase in power which may be transmitted by a practical finite
21 Olson, H. F., RCA Review, Vol. 2, No.2, p. 265, 1937.
HORN LOUDSPEAKERS 225
horn is only a few per cent greater than that shown in Fig. 7.8, because
very little distortion is generated in the large cross-sectional area near the
mouth of the horn.
It may be mentioned in passing that the multiple flare horn (see Sec. 5.25)
provides a means of decreasing the distortion because the rate of flare is
very rapid near the diaphragm and, therefore, the pressures are rapidly
reduced with respect to the distance from the diaphragm.
B. Distortion Due to Variation in Volume of the Air Chamber.22-In
general, acoustical, mechanical, and electrical networks are assumed to be
invariable; that is, the constants and connections of the network do not
vary or change with time. A network which includes a circuit element that
varies continuously or discontinuously with time is called a variable net
work. In some cases the variable elements are assumed to be certain
functions of the time; that is, the variations are controlled by outside forces
which do not appear in the equations or statement of the problem. In
another type of variable circuit element the variation is not an explicit
time function, but a function of the current and (its derivatives) which is
flowing through the circuit.
An example of the latter type of circuit element in an acoustical system
is the air chamber capacitance in a horn loudspeaker. The excurSIOns
of the diaphragm change the acoustical capacitance. The acoustical
capacitance of the air chamber of Fig. 7.9 is given by
The acoustical network of the acoustical system shows the effect of the
nonlinear element upon the sound power output. In the case of a single
frequency the distortion which this element introduces is small, because,
for constant sound power output, the amplitude of the diaphragm is in
versely proportional to the frequency. At low frequencies where the
amplitude of the diaphragm may be so large that the volume of the air
chamber becomes alternately zero and two times the normal volume, the
acoustical reactance of the acoustical capacitance is very large compared to
the acoustical resistance of the horn (see Fig. 7.9). At the high frequencies
22 Olson, H. F., RCA Review, Vol. 2, No.2, p. 265, 1937.
226 ACOUSTICAL ENGINEERING
A~ .~
~A ~
ACOUSTICAL NETWORK
LOW FREQUENCY
CROSS-SECTIONAL VIEW
FIG. 7,9. A mechanism with an air chamber coupling the diaphragm to the
horn. In the acoustical network: M I = the inertance of the diaphragm and
voice coil. CAl = the acoustical capacitance of the air chamber. YAI = the
acoustical resistance at the throat of the horn. p = the driving pressure.
p = BliIA. B = the flux density in the air gap. I = the length of the voice
coil conductor. i = the current in the voice coil. A = the area of the dia
phragm. The variation in volume of the air chamber introduces a nonlinear
element in the form of the acoustical capacitance CAl. The wave shapes of the
electrical input and acoustical output for a low, high, and a combination of a
high and a low frequency illustrates the effect of the nonlinear element upon
the acoustical output.
obtains, particularly with close spacing between the diaphragm and the
front boundary of the air cha!llber, the distortion may be tremendous.
Physically the low frequency modulates the high frequency.
In the above discussion the air chamber is assumed to be a pure acoustical
capacitance. This assumption is not correct at the higher frequencies
where the dimensions of the air chamber are comparable to the wavelength.
Regardless of the form of the acoustical impedance, it is, nevertheless, a
function of the spacing between the diaphragm and the air chamber and is
therefore a nonlinear element.
e. Distortion Due to the Diaphragm Suspension System. 23_The outside
suspension is another example of a variable circuit element in a mechanical
23 Olson, H. F., RCA Review. Vol. 2. No.2, p. 265. 1937.
HORN LOUDSPEAKERS 227
system. In certain types or, as a matter of fact, for unlimited amplitudes
in all types of suspension systems the stiffness is not a constant, but a
function of the amplitude and, in general, increase for the larger ampli
tudes (see Sec. 6.26).
In the case of a horn loudspeaker the amplitude of the diaphragm for
constant sound power output is inversely proportional to the frequency.
Furthermore, the mechanical impedance of the suspension system becomes
comparable to the other mechanical impedances in the system at the lower
frequencies. Consequently, the greatest distortion due to the suspension
system will occur at the low-frequency end of the working range.
The mechanical circuit of the mechanical system, Fig. 7.10, shows the
effect of the nonlinear element. When the stiffness of the suspension
system increases with amplitude, the third harmonic is the preponderant
distortion (see Sec. 6.26C). The wave shape under these conditions is
shown in Fig. 7.10. A distortion frequency characteristic of a diaphragm
coupled to a large throat horn is shown in Fig. 7.10.
M,",
MECHANICAL CIRCUIT
ELECTRICAL
v INPUT
WAVE
SOUND
SHAPES
OUTPUT
0- 1
~
U
... B\
0:
Q.
6
~
z
Q 4
\.
0
0:
o
1\,
0 2
0/)
r
CROSS-SECTIONAL VIEW o ~o 100 1000
rREQUENCY
160
...o
~140 /
"~
z
tjl20
/
If)
~ 100 /'
V /
,/ ,/
~ /' ./
~ 80
... V ~
/ ...... . - '
~
w
a:
j:!40
60
./ v:V ......
c .-'
V
../. ~. - 'V
/
~
w
~ 20
V
~Y
w
I
5 10 15 20 25
POWER INPUT IN WATTS
••
VI
."" ...... '" I'...
is'" " r-,
"" I'...
"r-.
'" r...... i'..
z 2 "
~ .1
•• " .l.SQ.IN.
...'"
C
~ 4
'-.:
.l.SQJ~
'<: . ,
~
" "....... .......
:J 2
'"
~ r......
I'..
""
-< i"
t'-r-,
I' I'..
~.Ol SQJn.
8
'" •
II.
4
~I. I'.. ......
,
.......... ....... r-, t-....
2
,
~SQ.I"il'.,....
"- " ,
.......
r-,
'"
.001
FIG.
2 4 5 ' 78V102
FREQUENCY
:I •
IN
4
CYCLES PER
2
SECOND
3 4 5 '7 8Vl04
'" 2
IJ
a:
..,20
I, IC /B "I" i"
a.
,
I
1\
!: 10
8
>
IJ
'\
~ 4
U
\.A "\.
i: \
...... 2
I
20 100 1000
\ 10000
rREQUENCY IN CYCLES PER SECOND
~I 0 ~
8
t
...z 4
...Q ,
~
....., 2
I
20 100 1000 10000
\
rREQUENCY IN CYCLES PER SECOND
The two preceding horn loudspeakers are suitable for high quality repro
duction of speech and music. For certain types of announce installations
it is desirable to project intelligible speech over very great distances (1
~~,.,
100
80
/"
/'
I M2 60
1,\
MECHANICAL NETWORK
>-
~ 40
\
'" 30
U V \
lo.
lo.
W 20
V \
/
CROSS-SECTIONAL VIEW
10
400
/ 1000
FREQUENCY
\ 4000
CROSS SECTION
LOW FREQUENCY UNIT
~
H.F.
TO
L.f. AMPLifiER
WIRING DIAGRAM
ASSEMBLY
~c~
~\. ~
'"
X
SEcnON A-X FRONT VIEW
horn unit for reproduction from 40 to 300 cycles and a multicellular horn
unit for reproduction from 300 to 8000 cycles.
In order to minimize time delay and phase distortion due to a large path
length difference between the low- and high-frequency horns, the effective
length of the low- and high-frequency horns must be practically the same.
The difference in path length in the system shown in Fig. 7.16 is made
relatively small by employing a short folded horn coupled to a large diameter
dynamic speaker mechanism. A further reduction in path length between
27 Wente and Thuras, Jour. A.I.E.E., Vol. 53, No. 1. p. 17, 1934.
HIGH
FREQUENCY
UNITS
LOW
FREQUENCY
UNIT
FRONT VIEW
high- and low-frequency units. The filter introduces phase shift as well as
a loss in power of 2 db or more.
The efficiency frequency characteristics of the high- and low-frequency
units of this loudspeaker without the filter are shown in Fig. 7.13, charac
teristics B. and A.
The low-frequency loudspeaker in the system depicted in Fig. 7.16
employs a short folded horn. Although the horn is short, there is still a
path difference between the low- and high-frequency horns of about 1
wavelength at the overlap frequency of 300 cycles. The path difference
can be obviated by the use of a high- and low-frequency horn of the same
length. In order to conserve space the over-all depth must not be too
HORN LOUDSPEAKERS 235
great. Under these conditions the flare cutoff at the low-frequency horn
will be about 80 cycles. The radiation mechanical resistance can be increased
and the output in the frequency range below the flare cutoff maintained by
the use of a phase inverter system in combination with the horn as shown in
Fig. 7.17. The action of the system may be determined from the mechanical
network of Fig. 7.17. By a suitable choice of constants uniform response
may be maintained in the low-frequency range down to 40 cycles.
A theater loudspeaker system 29 employing a low-frequency loudspeaker
of the type shown in Fig. 7.17 is shown in Fig. 7.18. The low- and high-
FRONT VIEW
frequency horns are of the type with straight sides and, therefore, exhibit a
reasonably uniform directivity pattern in the horizontal plane. See Sec.
2.19C. This design makes it possible to obtain the directional characteristics
of cellular horns without the complex construction. The low- and high
frequency horns are of the same length which obviates the transient distor
tion inherent in two channel systems in which there is a path length difference
between the low and high frequency. This feature also simplifies the
problem of obtaining uniform directivity in the overlap frequency region.
29 Volkmann, J., Unpublished Report.
236 ACOUSTICAL ENGINEERING
\
\ "-L
\ \
\,
- __
/
-t: \ ' ___...Y
/
1/
1/
I,
\\ l'
'\~ ,:'
~'v"-;:' -"-"''-.'f.~
-~+~- -
\oJ
B B' Z
:::;
'~'
\
'/
'/
,.........."':..-:.,
;'"
I
\\
\\
\\
a::
\oJ
z
a::
"" \\
o
u
II \\\\
1/ I \\
II
, /:'"--
\ ---7' \
, "
.t. \\
~ I /
I
A'
FRONT VIEW
30 Frayne and Locanthi. Jour. Soc. Mot. Pic. Tel. Eng.• Vol. 63. No.3, p. 82, 1954.
31 Klipsch. Paul W .• Trans. IRE. Prof. Group on Audio. Vol. AU-I. No.3. p. 16. 1953.
HORN LOUDSPEAKERS 237
is of the folded type. The acoustical radiation resistance presented to a
loudspeaker operating in the corner of a room at the intersection of the
floor and two walls is two times the acoustical radiation resistance presented
to a loudspeaker operating along a wall of a room at the intersection of the
wall and floor. See Sec. 2.2. This increased radiation can only be realized
at the very low-frequency portion of the frequency range.
Some wide frequency range systems divide the frequency range into
three sections, namely, low, mid, and high frequency. Direct radiator,
dynamic and electrostatic, and horn loudspeaker units are used to cover the
high-frequency range. A high-frequency horn loudspeaker for the frequency
range above 10,000 is shown in Fig. 7.21. The diaphragms are of the order
VOICE COIL
MAGNET
of i inch in diameter. The horns are about two inches in length. The
mouth area of the horn is about three square inches.
C. Compound Horn Loudspeaker. 32-The compound horn loudspeaker
consists of a single diaphragm mechanism with one side of the diaphragm
coupled to a straight axis horn and the other side coupled to a long folded
hom, Fig. 7.22. The equivalent of the system is shown in Fig. 7.22. The
functional acoustical network of the vibrating system is also shown in Fig.
7.22. At the low frequencies the acoustical reactance of the acoustical
capacitance, CA2, is large compared to the throat acoustical impedance,
ZA2, of the low-frequency horn and sound radiation issues from the low
frequency horn. At the high frequencies the acoustical reactance of the
acoustical capacitance, C A2, is small compared to the acoustical impedances,
ZAl and ZA2, and, therefore, shuts out the low-frequency horn and radiation
issues from the high-frequency horn. In the mid-range, radiation issues
from both horns. The response frequency characteristic, Fig. 7.22, shows
the response range of the two horns. The throats of the two horns may be
chosen so that the efficiency characteristic of this loudspeaker will be the
same as that of the two-channel system discussed in the preceding section.
However, the power handling capacity is somewhat smaller because the
size of the diaphragm must be a compromise between high-and low-fre
quency requirements.
32 Olson and Massa, Jour. Acous. Soc. Amer., Vol. 8, No. 1, p. 48, 1936.
238 ACOUSTICAL ENGINEERING
ACOUSTICAL NETWORK
FIG. 7.22. Cross-sectional view of a compound horn loudspeaker, the developed equiv
alent of the high- and low-frequency horns, and the acoustical network of the acoustical
system. In the acoustical network: M = the inertance of the diaphragm. GAl = the
acoustical capacitance of the diaphragm suspension system. ZAI = the acoustical
impedance at the throat of the small horn. ZA2 = the acoustical impedance at the
throat of the large horn. GA 2 = the acoustical capacitance of the chamber behind
the diaphragm. P = the driving pressure. p = Bli/A. B = the fiux density. I = the
length of the conductor in the voice coil. i = the current in the voice coil. A = the
area of the diaphragm. The sections A-A and B-B refer to the horizontal and vertical
cross sections of the front view. The graph shows the frequency ranges of the high
frequency and low-frequency horns and the over-all pressure response frequency charac
teristic.
the same function as the system shown in Fig. 7.15. The stresses in the
diaphragm and the voice coil system are reduced by the use of a number of
smaller units as contrasted to a single large unit. The possibility of failure
of the system is reduced by the use of a multi-unit driving system. The use
DIAPHRAGM
! {i'"
HORN
TO
AMPLIFIER
WIRING
DIAGRAM
/J{
FRONT VIEW SECTION A-/J{
MECHANICAL NETWORK
PERSPECTIVE VIEW
SECTIONAL VIEW
when the lid is open. With the horn collapsed, that is, the lid placed against
the case, the dimensions are the same as those of a direct radiator loud
speaker system. The experimentally determined response frequency
characteristic is shown in Fig. 7.26. The efficiency is of the order of 25
per cent. It is possible to obtain a sound level of 84 decibels at a distance
of three feet with 10 milliwatts.
F. Folded Horns.-There are innumerable ways of folding or curling a
horn. The different types of folded horns are shown in Figs. 7.16 and 7.22.
The principal purpose of folding or curling a horn is to use the volume
occupied by the horn more efficiently. Three more different types of folding
HORN LOUDSPEAKERS 241
are shown in Fig. 7.27. A simple folded horn is shown in Fig. 7.27A.
A folded horn with a ring-shaped mouth is shown in Fig. 7.27B. The
directional characteristics of a ring-shaped mouth are sharper than those
of the rectangular or circular shapes having equivalent areas (see Sees. 2.9
and 2.10). The horn shown in Fig. 7.27C is used for sending out radiation
o
,.. l/1\
dI
c
-10
I
/ "" ,
~
....
~ -15
I
~
gj
II: -20 /
-25 J
-30
200
/ 400 600 1000 2000 4000
FREQUENCY IN CYCLES PER SECOND
~{~}e
~\17~
BOTTOM VIEW
A
SECTION
B
A-A' SECTION
C
A-A
A B c
FIG.7.28. Horn loudspeaker driving mechanisms. Mechanisms A, B, C, and
D depict various types of air chambers and diaphragms for coupling to a small
throat horn. Mechanism E depicts a large diaphragm coupled to a large
throat horn.
different methods for reducing ihterference in the air chamber are shown in
Fig. 7.28A, B, C, and D and Figs. 7.1, 7.9, 7.15, and 7.21. These expedients
are necessary for efficient reproduction at the high-frequency portion of
the audio range where the wavelength is relatively small. For the low
frequency portion of the audio-frequency range a large-throat horn may be
coupled to a large diaphragm, as shown in Fig. 7.28E, without incurring
any loss due to interference, notwithstanding the large size, because the
dimensions are small compared to the wavelength.
H. Diaphragms and Voice Coils.-The diaphragms or cones of horn
loudspeaker mechanisms are made of aluminum alloys, molded bakelite
with various bases, molded styrol, fiber, paper, and felted paper. Typical
diaphragm shapes are shown in Figs. 7.1, 7.9, 7.10, 7.15, 7.23, 7.24 and 7.28.
HORN LOUDSPEAKERS 243
Round, square, and ribbon wire voice coil conductors are used as shown in
Fig. 6.77.
1. Field Structures.-Permanent magnet and electromagnetic field
structures used in horn loudspeaker mechanisms are shown in Figs. 7.1,
7.9, 7.10, 7.15, 7.23, 7.24, 7.28, and 6.79. In general, it is customary to
use higher flux densities in the gap in horn loudspeakers than in direct
radiator loudspeakers. Soft iron may be used for the pole tips for flux
densities up to 20,000 gausses (See Fig. 6.82). For flux densities from
20,000 to 23,000 gausses, a special alloy, Permandur 34 (see Fig. 6.82), is
employed for the pole tip material in order to obtain these high densities
with tolerable efficiency.
J. Horn Walls. Vibration and Absorption. 35-In the theoretical analysis
carried out in this chapter it has been assumed that the horn walls are
rigid and nonabsorbing. In the case of certain materials such as wood,
paper, and fiber the absorption of sound by walls of the horn may introduce
an attenuation of several decibels. The absorption may be reduced by the
application of lacquers and varnishes. The attenuation in metallic horns
due to dissipation is negligible. The vibration of the walls of the horn
distorts the response frequency characteristic and introduces "hangover"
and reverberation. The response to transients is usually poor when the
walls of the horn vibrate. This vibration may be reduced by increasing
the thickness of the walls and by suitable bracing. The vibrations and ring
in metallic horns may be reduced by coating the outside of the horn with
deadening material such as asphalt or pitch compounds.
7.5. Throttled Air Flow Loudspeaker.-A throttled air flow loud
speaker consists of a valve mechanism, actuated by the electrical signal,
which modulates a steady air stream so that the undulations in the throttled
air stream correspond to the variations in the electrical input, see Fig. 7.29.
The throttled air stream is usually coupled to a horn to improve the efficiency
of the system. In order to obtain a constant relationship between the
electrical input and the acoustical output as a function of the frequency,
the ratio of the volume current to the applied current must be independent
of the frequency. This means that the ratio of the amplitude of the valve
to the current must be independent of the frequency. This is, in general,
difficult to accomplish in the high-frequency region in view of the fact that
the valve mechanism must be stiffness controlled. This in turn means that
the mass of the valve must be small and at the same time be sufficiently
rugged to withstand the steady air pressure. The outstanding advantage
of this system is the large acoustical output which can be obtained for a
small electrical input. Efficiencies of more than 100 per cent can be realized
if the ratio of acoustical output to the electrical input is considered. How
ever, in addition to the electrical power which must be supplied to the
throttling mechanism, there is the power that must be furnished in
supplying the steady stream of air. If this mechanical power is added
34 Elmen, G. W., Bell Syst. Tech. Jour., Vol. 15, No. 1. p. 113, 1936.
35 Phelps, W. D., Jour. Acous. Soc. Amer., Vol. 12, No. 1, p. 68, 1940.
HORN
VALVE
AIR INLET
CYLINDRICAL
SHIELD
FIG. 7.30. Schematic and sectional view of an audio modulated corona dis
charge loudspeaker. (After Klein.)
MICROPHONES
~~tL
~ AH CA , CM y
ACOUSTICAL NETWORK
MICROPHONE~rE2
ELECTRICAL CIRCUIT
-20
M,
III
fA' °-30
'"z-40
III
/\
0
Q.
:3-50
V \v ~
'" £
CROSS - SECTIONAL VIEW 1000 10000
FREQUENCY
FIG. 8.1. Cross-sectional view, the electrical circuit and the acoustical network
of a single-button carbon microphone. In the electrical circuit, rEi = the elec
trical resistance of the carbon element, rE2 = the electrical resistance of the load,
and e = the polarizing voltage of the battery. In the acoustical network
Ml and 1'.41 = the inertance and acoustical resistance at mouthpiece opening.
ZAH = the acoustical quadripole representing the horn or mouthpiece. M2 and
1'A2 = the inertance and acoustical resistance of the holes in the mouthpiece.
CAl = the acoustical capacitance of the air chamber in front of the diaphragm.
Ms, 1'A3, and C A 2 = the inertance, acoustical resistance, and acoustical capaci
tance of the diaphragm. M4, 1'A4, and CAS = the inertance, acoustical resistance,
and acoustical capacitance of the carbon element. CA 4 = the acoustical capaci
tance of the case. p = sound pressure. The graph shows the open circuit
voltage response frequency characteristic for constant sound pressure in free
space.
the displacement. Consider the electrical circuit of Fig. 8.1, for sinusoidal
motion of the diaphragm, the current, in amperes, in the circuit is given by
. e
t= 8.1
rEO + hx sin wt
where e = voltage of the battery, in volts,
rEO = total electrical resistance of the circuit when x = 0, in ohms,
x = amplitude of the diaphragm, in centimeters,
h = constant of the carbon element, in ohms per centimeter,
w = 27Tj, and
f = frequency, in cycles per second.
248 ACOUSTICAL ENGINEERING
MECHANICAL NETWORK
0
'"
~"'40 - B ,'/ ~
i'".J\
.
o
:3"'50
,
, '1""
A
'" ,
a: :
-60
CROSS-SECTIONAL VIEW 100 1000 10000
FREQUENCY
MECHANICAL NETWORK
-zo
.,
0-30
\oJ
,./........ ..... i\.
'"z-40
!?
'"~-50 ~
-60
100 1000 .0000
CROSS-SECTIONAL VIEW FREQUENCY
FIG. 8.3. Cross-sectional view and the mechanical network of a new type single-button
carbon microphone. The electrical circuit is the same as that of Fig. 8.1. In the
mechanical network, mo and t'MO = the mass and mechanical resistance of the holes in
the outer grill. CMO = the compliance of the air chamber between the inner and outer
grids. ml and t'Ml = the mass and mechanical resistance of the waterproof membrane.
CMl = the compliance of the air chamber behind the diaphragm. C M3 = the compli
ance of the air chamber behind the waterproof membrane and the diaphragm. m2 and
t'M2 = the mass and mechanical resistance of the cloth. m3 and t'M3 = the mass and
mechanical resistance of the hole in the diaphragm. m5 = the mass of the diaphragm.
1'M4 and C M4 = the mechanical resistance and compliance of the diaphragm to carbon
cup couples. C M2 = the compliance of the air chamber behind the microphone unit.
1'M5 and C M5 = the mechanical resistance and compliance of the diaphragm suspension
system. m6, 1'M6, and CM6 = the mass, mechanical resistance, and compliance of the
carbon cup and granules. 1M = the driving force. 1M = pA. P = the sound pressure.
A = the area of the diaphragm. The graph shows the open circuit voltage response
frequency characteristic for constant sound pressure on the diaphragm.
The free-space response shown in Fig. 8.2 indicates the diffraction effect
of the microphone as an obstacle in increasing the pressure on the diaphragm
(seeSec.1.11 and Fig. 1.5).
In addition to the smoother response the sensitivity of the new unit is
higher because of the reduction in mass of the vibrating system. Due to
the shape of the carbon chamber the performance of the microphone is less
affected by angular position.
MICROPHONES 251
A new carbon microphone,2 shown in Fig. 8.3, has been developed which
in appearance is similar to the microphone of Fig. 8.2. However, there are
many important features which lead to improved performance. One of the
most important elements that has been added to the microphone of Fig. 8.2
is the mechanical resistance, rM2. The vibrating system is stiffness controlled
below the resonant frequency. At the resonant frequency the mechanical
resistance provides the controlling element. As a consequence, the response
in the frequency region 2000 and 4000 cycles, depicted in Fig. 8.3, is smoother
than that of the microphone of Fig. 8.2. The sensitivity of the microphone
of Fig. 8.3 is somewhat greater than the microphone of Fig. 8.2. This has
been accomplished by an improved design of the carbon cup.
m.~~~
~r-------,J
"'.,
MECHANICAL NETWORK
MO<'O~B,'j~C
ELECTRICAL SYSTEM
;~mllilit
1 fIllflllEflI
,~ .~ '00 f'REQUENCY
CROSS-SECTIONAL VIEW
FIG. 8.4. Cross-sectional view. the electrical circuit. and the mechanical
circuit of a double-button. stretched diaphragm, carbon microphone. In the
mechanical circuit. ml and rMI = the mass and mechanical resistance of the
air load. ZMH = the mechanical quadripole representing the cylindrical
cavity or pipe. m2 and CMI = the mass and compliance of the diaphragm.
ma, rM2, and CM2 = the mass, mechanical resistance, and compliance of the
carbon granules. m4. rM3, and C M3 = the mass, mechanical resistance. and
compliance due to the damping plate. JM = the driving force. JM = pA.
A = the area of the diaphragm. p = the sound pressure. The graph shows
the open-circuit voltage response-frequency characteristic for constant sound
pressure at the diaphragm.
3 Jones. W. C.• Bell Syst. Tech. Jour., Vol. 10, No. 1, p. 46. 1931.
- (hX sin wt + -
3- x3.
h 3S t h 3x 3 . 3 t )
=2e Illw---SIllw . . . 8.5
rEO rEO 4 rE0 3 4rE0 3
MICROPHONES 253
Comparing equation 8.5 with equation 8.2 shows that the large second
harmonic term has been eliminated by the use of a push-pull two-button
microphone.
One common cause of faulty operation of the carbon microphone is due
to the cohering of the carbon granules caused by the breaking of the circuit
when the current is flowing. The use of electric filters as shown in the
circuit diagram will protect the microphone against cohering.
The frequency range and response of the double-button carbon micro
phone compares favorably with the condenser microphone. The carbon
microphone is several times more sensitive than the condenser microphone.
However, the limitation is carbon noise.
~~'h
~ MECHANICAL NETWORK
~
CROSS-SECTIONAL VIEW
(CEO + CEI sin wt)rE ~ + (1 + rECElw cos wt)i - eOCElw cos wt = 0 8.8
The solution of equation 8.8 is
t
.= eoCEl . (t
sm w + '/'1
-I. )
CEO"v(1/CEow)2 + rE 2
eoCEirE sin (2wt + CPl - CP2)
CE02V[(1/C Eo w)2 + 4rE2] [(1/C Eo w)2 + rE2]
+ terms of higher order 8.9
where CPl tan-1 1/CEOwrE and CP2 = tan-1 1/2CEOwrE.
=
For small diaphragm amplitudes, the generated voltage, in statvolts, is
,.
e = rEt =
CEO
J eoCEl rO
1
___ + r02
. (
sm wt
-I. )
+ ,/,1 8.10
CE0 2w 2
w = 27T1, and
Equation 8.12 shows that the sensitivity below the resonant frequency
is inversely proportional to the stiffness and the mechanical resistance.
For the same fundamental resonant frequency the stiffness can be reduced
6 Crandall, I. B., Phys. Rev., Vol. 11, No.6, p. 449, 1918.
7 Crandall, " Vibrating Systems and Sound," D. Van Nostrand Company, Princeton,
N.j., 1926.
256 ACOUSTICAL ENGINEERING
by decreasing the mass. This procedure also reduces the amount of mechan
ical resistance required to damp the fundamental resonance and thereby
obtain uniform response. Aluminum alloys, due to the low density and
high tensile strength, are the logical materials for use in diaphragms. The
minimum diaphragm thickness suitable for the manufacture of condenser
microphones is about .001 inch. The electrical capacitance of a microphone
with a diaphragm diameter of 1% inches and a spacing of from .001 to .002
inch is from 400 to 200 mmfds. Due to the high electrical impedance of this
capacitance it is necessary to locate the microphone near the vacuum tube
amplifier. The electrical capacitance of a long connecting cable reduces
the sensitivity without frequency discrimination because the internal
electrical impedance of the microphone is also an electrical capacitance.
The response frequency characteristics of a condenser microphone for
k\\....,.....,.,~--BACK PLATE
CLAMPING
constant sound pressure on the diaphragm and for constant free wave
sound pressure are shown in Fig. 8.5.
The condenser microphone 8 ,9 shown in Fig. 8.5 employs a diaphragm with
a diameter of 1% inches. The over-all diameter of the condenser microphone
unit is about 3 inches. These microphones were developed about twenty
five years ago and were employed in the early days of sound reproduction.
The condenser microphone was replaced by the electrodynamic (voice coil
and ribbon) and piezoelectric microphone. During the past decade, smaller
condenser microphones have been developed. A miniaturized version of the
microphone shown in Fig. 8.5 is shown in Fig. 8.6. The over-all diameter of
the microphone unit is a little less than 1 inch. The fundamental resonant
frequency of the diaphragm is about 9000 cycles. The system is highly
damped so that uniform response is maintained to over 15,000 cycles. The
8 Harrison and Flanders, Bell Syst. Tech. Jour. , Vol. 11, No.3, p. 451, 1932.
9 Veneklasen, Paul S., Jour. Acous. Soc. Amer., Vol. 20, No.6, p. 807,1948.
MICROPHONES 257
deviations in response are smooth and can be easily compensated by electrical
means to obtain a response frequency characteristic which is independent of
frequency. The amplifier which may be used with this microphone is
shown in Fig. 8.7. The cathode follower type of operation provides a
DIAPHRAGM
II LOUTPUT
250V
system having a high input electrical impedance. This is necessary for the
small condenser microphones in which the capacitance is only about SO
mmfds in order to maintain the response in the low-frequency region. See
equation 8.11. The condenser microphone shown in Fig. 8.6 is used as a
standard microphone in pressure calibration of laboratory standard micro
phones. See Sec. 1O.2Ald.
Another miniature condenser microphone 1o is shown in Fig. 8.8. This
microphone employs a plate instead of a stretched diaphragm. See Sec.
CASE
~~~~~~~~~~~~~~ QUARTZ
DIAPHRAGM
3.5. The over-all diameter of the microphone unit is about ! inch. The
amplifier used with these microphones is of the type shown in Fig. 8.7.
c. Piezoelectric (Crystal) Microphones.ll.12.13_A piezoelectric micro
phone is a microphone which depends upon the generation of an electro
motive force by the deformation of a crystal having piezoelectric properties.
10 Hilliard. J. K.. and Noble. J. J .. Trans. IRE. Prof. Group on Audio. Vol. AU-2.
No.6. p. 168. 1954.
11 Sawyer. C. B .• Proc. Inst. Rad. Eng.• Vol. 19. No. 11. p. 2020.1931.
12 Williams. A. L.. Jour. Soc. Mot. Pic. Eng.• Vol. 18. No.4. p. 196. 1934.
13 Nicolson. U.S. Patent 1.495.429.
258 ACOUSTICAL ENGINEERING
~
BIMORPH
ELEMENTS
TERMINALS
TWISTER BIMORPH
BASE-"~--'I
DIRECT
BENDER BIMORPH CRYSTAL ACTUATED DIAPHRAGM
MICROPHONE CRYSTAL ACTUATED CRYSTAL
CRYSTAL ELEMENTS UNIT OR CELL MICROPHONE MICROPHONE
FIG. 8.9. Crystal elements and sound cells. A direct actuated crystal micro
phone. In the electrical network, CEO = the electrical capacitance of the
crystal. "EO = the electrical resistance of the crystal. Z filL = the electrical
impedance of the load. eo = the open circuit voltage developed by the
crystal. In the mechallical circuit, m, "M, aud C M l = the mass, mechanical
resistance, and compliance of the crystal. C M2 = the compliance of one half
of the air chamber. 1M = the driving force, 1M = pA. A = the effective
area of the crystal. p = the sound pressure. A diaphragm actuated crystal
microphone. In the mechanical circuit, ml, "Mlo and C M1 = the mass,
mechanical resistance, and compliance of the diaphragm. m2, "M2, and
CM2 = the mass, mechanical resistance, and compliance of the crystal.
CM3 = the compliance due to the case volume. 1M = the driving force.
1M = pA. A = the effective area of the diaphragm. p = the sound pressure
at the diaphragm.
1M
( +. 1 1) .
x=~----------~~------~~ 8.14
rM Jwm + JW
-:---C + -;---C JW
Ml JW M2
i = _ _ _..::..1M;;;;;;...._--:-_ 8.15
. 1
rMl + JWml + JW
-·-C-
Ml
The most common materials used for the diaphragms of pressure micro
phones are aluminum alloys, Bakelite, styrol, and paper. In order to obtain
a minimum density-resistivity product, aluminum is almost universally
used for the voice coil (see Table 6.1). Both edgewise wound ribbon and
round wire have been used for the voice coil (see Sec. 6.27 and Fig. 6.76).
p
DIA~SION I
V
IY
'"
Cw,Tw, GOIL
A CROSS-SECTIONAL VIEW I
./ ""
fN 5 ~
Z
C ~
I~ -...;:::
MECHANICAL CIRCUIT
~
~J
ELECTRICAL CIRCUIT
I
> S
!: 4
u
~
g.OI/( A
!oJ
> 1
.oo~o 100 1000 10000
fREQUENCY
MECHANICAL
~
ELECTRICAL CIRCUIT ~'
m C-.., r.
loll
u'~:"r:':=!!!!!!!;k~.",;B~~T m3 r"3
SILK
f.., f"l!
I I
MECHANICAL NETWORK
a: 30 100 1000
fREQUENCY
10000
located in the bottom of the "V." The mechanical network of this micro
phone is the same as that of the dynamic microphone in the preceding section
and the action is the same. A transformer, housed in the magnet structure,
is used to step up the low electrical impedance of the conductor to that
suitable for transmission over a line of several hundred feet.
3. Ribbon Microphone.-The pressure ribbon microphone 21 ,22 consists
of a light metallic ribbon suspended in a magnetic field and freely accessible
to the atmosphere on one side and terminated in an acoustical resistance
on the other side. The essential elements are shown schematically in Fig.
8.12. These elements may take various forms as, for example, the pipe is
usually coiled in the form of a labyrinth (see Fig. 8.14).
The acoustical network 23 of the pressure ribbon microphone is shown in
Fig. 8.12.
The inertance and acoustical capacitance of the ribbon are designated
by MR and CAR.
The acoustical resistance and mass of the air load upon the ribbon are
designated by r AA and M AA. The expression for the air load upon the
ribbon will now be derived. The pressure, in dynes per square centimeter,
at a distance a in centimeters, from an elementary source is (see Sec. 2.2)
p = dS jwpumaxEiwtE-ika 8.17
4-rra
where dS area of the source, in square centimeters,
=
Umax = maximum velocity of dS, in centimeters per second,
p = density of air, in grams per cubic centimeter,
w = 2",j,
j = frequency, in cycles per second,
U = velocity over the surface dS, in centimeters per second,
t = time, in seconds,
k = 2",/>.., and
>.. = wavelength, in centimeters.
The pressure at any point on the ribbon due to a velocity umaxEiwt of the
ribbon is
8.18
where al = radius vector having the shortest air distance from the point
1 to the surface element dS. To compute the total force, the above integra
tion must be performed and then the resulting pressure integrated over the
surface of the ribbon.
22 Olson, H. F., Jour. Soc. Mot. Pic. Eng., Vol. 27, No.3, p. 285, 1936.
MICROPHONES 265
The total force is
JMA = jwPU:;X€iwtJJdSfJ ~~ €i ka 1 8.19
where dS'
= surface element at 1.
The acoustical impedance due to the air load is
.
ZAA=r AA +JXAA = A2
JMA
Umax€
it
w
8.20
[lf
_
:
=
=-
MsrAS
M~
to MA
rAA
CAR
RIBBON
R~:rNZ[2 i~i· LINE i~(i) :~[:TI[J
ELEC'==T=-R-IC-A-L-S"":Y=-STEM ELEC. EQUIV.
"RONT PIPE
VIEW
FIG. 8.12. Schematic view, electrical system and its equivalent and
acoustical network of a pressure ribbon microphone. In the electrical
circuit, rEI = the electrical resistance of the ribbon. ZE2 = the external
electrical impedance load presented to the ribbon. eo = the open circuit
voltage developed by the ribbon. In the acoustical network, MA and rAA =
the inertance and acoustical resistance of the air load on the ribbon.
MR and CAR = the inertance and acoustical capacitance of the ribbon.
Ms and rAS = the inertance and acoustical resistance of the slit. ZAE =
the acoustical impedance due to the electrical circuit. ZAP = the acousti
cal impedance of the pipe. p = the sound pressure.
The ribbon is spaced from the pole pieces of the magnetic structure to
allow freedom of motion. This slit or aperture, r AS and MAS, gives rise to
an acoustical impedance (see Sec. 5.4),
ZAS = rAS + jwMs 8.21
where r AS = acoustical resistance of the slit, in acoustical ohms, and
Ms = inertance of the slit, in grams per (centimeter).4
The back of the ribbon is terminated in an acoustical resistance in the
form of a finite pipe damped with tufts of felt. The acoustical network of
the pipe shows that for the mid- and high-frequency range the acoustical
impedance is an acoustical resistance.
The acoustical resistance of the pipe referred to the ribbon is
42
rAP = Ap 8.22
The acoustical impedance due to the electrical circuit may influence the
motion of the ribbon. The acoustical impedance due to the electrical
circuit is
(El)2
ZAE= - - 8.23
AR2ZET
/~ 2
4>, +v Io
A.OI4 0 100
r-- ~~
1000 10000 B 040 100
.........
r- ./
1000
0
10000
FREQUENCY FREQUENCY
FIG. 8.13. A. The acoustical impedance characteristics of the elements of the ribbon
pressure microphone. XAR = the ribbon acoustical reactance. XAA = the air load
acoustical reactance. r AA = the air load acoustical resistance. rAP = the acoustical
resistance of the pipe. XAP = the acoustical reactance of the pipe. XAE = the
acoustical reactance due to the electrical system. p = the sound pressure. B. U y =
the volume current of the ribbon for a sound pressure of 1 dyne per square centimeter.
4> = the phase angle between the ribbon volume current and the driving pressure.
4>1 = leading. .p2 = lagging.
coupled to the ribbon. This provides a small pickup area and, therefore, a
nondirectional characteristic. The surge acoustical impedance of the pipe
is the same as the acoustical impedance of the damped pipe. Under these
conditions a smooth response frequency characteristic is obtained because
the sound flows into the pickup pipe, past the ribbon and then into the
damped pipe without reflections. A small horn is used at the pickup point
to accentuate the response in the frequency region above 5000 cycles.
4. Probe Microphone. 25 ,26-In some acoustical measurements, a micro
phone equipped with a small sound pickup system with a high acoustical
impedance so that it will not disturb the sound field is a useful tool. A
probe-type microphone shown in Fig. 8.15 satisfies these requirements.
The probe-type microphone consists of a tube about 4 inches in length and
an inside diameter of from .020 to .1 inch coupled to a small condenser
microphone. The consideration of the attenuation of sound in tubes as a
function of the frequency and diameter will be found in Sec. 5.32. The
24 Olson and Preston, Audio Engineering, Vol. 34, No.7, p. 18, 1950.
25 Hilliard, John K., Trans. IRE, Prof. Group Audio, Vol. AV-2, No.6, p. 168, 1954.
26 Benson, Robert W., Jour. Acous. Soc. Amer., Vol. 25, No. 1. p. 128, 1953.
It
CONNECTOR
RIBBON
MAGNET
DAMPED LABYRINTH
CONDENSER
MICROPHONE
PROBE TUBE
COUPLING
CHAMBER
FIG. 8.14. Cross-sectional view of a rib- FIG. 8.15. Cross-sectional view of a
bon-type pressure microphone. probe-type microphone consisting of a
tube of small bore coupled to a condenser
microphone.
'" "
is zero, namely, the conductor also In
0 _3
acts as the diaphragm. There
~
fore, the ultimate sensitivity in ...
~ -6
the high-frequency region will ...
always be higher in a ribbon 0-9
system than in a diaphragm and CI
moving coil system.
A diaphragm-type dynamic
::I
...
of
~-12
g
"" r-.....
----
..... r-
-15
microphone requires a rigid dia o .5 I 1.5 2 2.5 3
phragm system in order to pre MASS OF THE DIAPHRAGM
MASS OF THE COIL
vent spurious responses due to
relative motion of different parts FIG. 8.16. The voltage output of a dynamic
of the diaphragm. Since the microphone as a function of the ratio of the
mass of the diaphragm to the mass of the coil.
ribbon serves a dual purpose of The ribbon microphone corresponds to the
diaphragm and voice coil, rigidity value zero for this ratio.
is not required. Therefore, the
response frequency characteristic of a ribbon system is smoother than the
diaphragm-voice coil system.
Another consideration is the low-frequency response. In well-designed
ribbon-type microphones, the resonant frequency of the ribbon can be
placed below the audible range. Therefore, above the audible range, the
diaphragm and conductor system is mass controlled. This simple mass
characteristic makes it a particularly simple task to develop suitable
phase shifting networks for ribbon transducers.
The transient response is another important characteristic of a microphone.
It can be shown that in a well-designed system, the one with the lowest
inherent mass will exhibit the most faithful response to transients. There
fore, the ribbon system will exhibit the best transient response. This has
been substantiated by large- and small-scale explosion tests.
Closely allied to transient response is wind noise. In order to determine
the wind response of different microphones, a wind generator, described in
Sec. 1O.2G, was developed. This wind machine delivers a combination of a
steady and a fluxuating air stream and, therefore, simulates wind under
actual conditions. Tests have been made comparing the wind-noise response
of different microphones obtained with natural wind and the wind machine.
These results show that there is practically a perfect correlation. The
advantages of the wind machine are as follows: wind response can be obtained
at any time and the wind noise produced by the machine remains constant.
Using the wind machine, tests have shown that with the same screening
and the same response frequency characteristic, the response to wind of all
microphones is directly related to the sensitivity, which means the signal to
wind noise is the same.
1
DIAPHRAGM
COIL
~W~_MAGNET
DIAPHRAGM
CDIL
N TURNS
e tP
SPACER
MM
MAGNET
RM ---t::t!:_ _---1...J
Employing equation 8.27 or 8.29, the change in flux with respect to time is
Acp MAAx
8.30
At = 4a 2 At
or
dcp MAdx
8.31
dt 4a 2 dt
272 ACOUSTICAL ENGINEERING
AUDIO
OUTPUT
T
SECTIONAL VIEW ELECTRICAL SYSTEM
~
m T rMT rMO mO C UO ! m, r M3 m3 C M3
f~~~---'
T
M2 ·~MC_L ~~c mz Me i '
I _ ML 1.11
~------~----~,----~-----+----------~~~
MECHANICAL NETWORK
. 1
ZM3 = YMD + JWmD + JW
-.C-
MD
8.38
1
ZM4 = YML + JWCML
-.- 8.39
- !M2ZM2(ZM4+ ZM5)
jw[ZMI (ZM2+ ZM3) (ZM4 +ZM5) +ZM4ZM5 (ZMI +ZM2) +ZM2ZM3(ZM4 +ZM5)] 8.43
The amplitude, in centimeters, of m2 is
X3 =
!MZM7 ZM9 8 44
jW[ZM6 (ZM7 +ZMS) (ZM9+ ZMlO) +ZM9ZMIO(ZM6 +ZM7) +ZM7ZMS(ZM9+ZMIO)] .
where
r (ZM4ZM5) (Xl+ X2) .
JM =
ZM4 + ZM5 JW
The amplitude response charac
teristic can be obtained from equa
tion 8.44 and the constants of the
system. The voltage output can be
obtained from the amplitude and
equation 8.33.
The electrical connections for the
MECHANICAL
electronic microphone are shown in
31 Olson, H. F., Jour. Acous. Soc. Amer., Vol. 3, No. 1. p. 56, 1931.
32 Olson, H. F., Proc. Inst. Rad. Eng., Vol. 21. No.5, p. 655,1933.
33 Massa, F., Jour. Acous. Soc., Amer., Vol. 10, No.3, p.173,1939.
34 Anderson and Wigginton, Audio Engineering, Vol. 34, No. 1. p. 12, 1950.
A comparison of equations 8.45 and 8.50 shows that for a wave of constant
sound pressure the driving force is proportional to the frequency.
The velocity of the mechanical system, for D.x small compared to the
wavelength, is
i = !M
Jwm
= S,?m D.x cos
Jem
e cos ket = SPm D.x cos
em
e sin ket 8.51
i =
2SPm . (k et ) sm
mw sm 2 cos uD)
. (kD.x 8.52
1.0
r---...
.9
.8
" \
r\
\
t
{
!; .3
o
.2
.1 IL'1
II
~
I \ If\.
o
.01
2 34$6789
.1
2 3
o
4 ~ «578.
I
2 3
"""
45&78'
10
X
FIG. 8.22. Computed open-circuit voltage response frequency charac
teristic of a pressure gradient, mass-controlled, electrodynamic micro
phone.
D=kA D=.!A
'0 '.0
H-++3~E'E'+'f-t-1" 270H-++3~!iE'+--1f-t--jo'
D=A
'00 '00
36 Olson, H. F., Jour. Acous. Soc. Amer., Vol. 3, No. 1, p. 56, 1931.
37 Olson, H. F., Proc. Inst. Rad. Eng., Vol. 21, No.5, p. 655, 1933.
MAGN
~
1 =1
··
=
POLE - =
= R IBBON -~
RIBBON
=
=
,·
,,
I
=
1= 1
0 0
TERMIN AL- f--
' ..I
fRONT VIEW END VIEW
I - vI - sin 2 (J
where IL = .
SIll
(J ,
IOU = 1 when u = 0,
IOu = 2 when u =F 0,
(J = angle of the incidence,
R = radius of the plate, in centimeters,
k = 27T/>',
>. = wavelength, in centimeters, and
] u = Bessel function, of the order u.
38 Sivian and O'Neil. Jour. Acous. Soc. Amer., Vol. 3, No.4. p. 483.1932.
MICROPHONES 281
The pressure at the center on the front and back of a circular plate for
normal incidence 8 = 0° or 180°, from equation 8.58, is
IP~801 = 1 8.60
3.0
/ 11
I
/
2.0 /
o· /
...ii!
0.
'/
.. 1.5
V
~
1.0 - 180·
~
.5
o a 3 .. ~ e 7 89. 2 2
.01 1
~
FIG. 8.25. Computed pressure frequency characteristic,
at the center, on the front, and the back of a circular
baffle for normal incidence of the impinging sound wave.
of three times that in free space at RIA = .5, then falls back to the same
as the free space pressure at RIA = 1, and repeats for RIA = 1.5 and
RIA = 2, etc. The pressure at the back is the same as the free space pres
sure for all frequencies. The measured pressure at the center on the front
and back of a circular baffle is shown in Fig. 8.26. In order to reduce
errors in measurement to a minimum, baffles of different diameters were
used. In addition, several different pressure measuring arrangements
were used. The results shown in Fig. 8.26 represent an average of these
determinations. It will be seen that theory and experiment are in fairly
good agreement. Some of the deviation may be attributed to finite size
of the pressure measuring system.
282 ACOUSTICAL ENGINEERING
The phase angles at the front and back of a circular baffle computed from
equation 8.58 are shown in Fig. 8.27. A point in the plane wave cor
responding to the plane of the baffle is the reference plane for the phase.
It will be seen that for RIA less than .5 the phase of the pressure at the front
of the baffle leads that of the pressure in the wave. For values of RIA less
3.5
1"'\
3
2.5 II n
2 .0 /
If
c..
+
21.5
I
~V i-" -r--, ~
I
c:f
;;;...
0 2 3. s,e78!l 2 2
.01 .1
R
1\
FIG. 8.26. Measured pressure frequency characteristic
at the center, on the front, and the back of a circular
baffle for normal incidence of the impinging sound wave.
than .1 the phase on the front leads by the same amount as the phase on the
back lags the pressure in the wave.
Equation 8.58 may be used to compute the difference in pressure between
the two sides of a relatively small ribbon located in a large baffle (Fig. 8.28).
The difference in pressure between the two sides of the ribbon in a circular
baffle, Fig. 8.28, is
t1p = Pe - PB 180 + 8.61
where PB and Po +
180 may be obtained from equation 8.58. The acoustical
impedance of the ribbon, Fig. 8.28, is given by
ZAR =JW
. MR --
j 8.62
WCAR
f MA =
jWpUmaxEJwt
47T
If Ifa;:-EdS' ds ika
1 8.63
The above integration extends over both sides of the ribbon and cog
nizance must be taken of the 180 0 difference in phase between the front
90
Cf'
o
r--. V\
V
-90
~ ~
""",
\
-180
In
\
'"
'"
Ir
Cl
~-Z70
~
~-360
-<
\
-540
\
-630
\
-720
\
2 3 4 SITal 2 3 4 5 6789 2
.01 .1 I
R
X
FIG. 8.27. Computed phase frequency characteristic
at the center, on the front, and the back of a circular
baffle for normal incidence of the impinging sound wave.
and back when integrating between the two surfaces. The integration of
equation 8.63 may be carried out by dividing the ribbon into small elements
and carrying out the indicated integration.
The acoustical impedance of the air load is
. JMA
ZAA = r AA + JXAA = A R 2Umax€ jwt 8.64
284 ACOUSTICAL ENGINEERING
The acoustical impedance, ZAS, of the slit between the ribbon and pole
pieces is given by equation 8.21. The acoustical impedance due to the
electrical circuit is given by equation 8.23.
The resonant frequency of the ribbon is usually placed below the audible
limit. Therefore, the acoustical capacitance of the ribbon may be neglected.
The acoustical resistance, r AA, of the air load
is negligible save at the very high frequencies.
When the fundamental resonant frequency of
the ribbon is located below the audible
frequency range, the negative reactance term
in equation 8.62 may be neglected. Under
these conditions the acoustical impedance of
the vibrating system is
ZAT = jwMR + jwMAA 8.65
where M AA = inertance of the air load.
FIG. 8.28. A velocity micro The velocity, in centimeters per second, of
phone with a large circular the ribbon is
baffle. . I:1p
X=-- 8.66
ARZAT
where AR = area of the ribbon, in square centimeters.
The voltage output in abvolts is
e = Eli 8.67
where B = flux density, in gausses,
l = length of the ribbon, in centimeters, and
i = velocity of the ribbon, in centimeters per second.
The response characteristic of a mass-controlled ribbon located in a large
circular baffle, Fig. 8.28, computed from equation 8.67 is shown in Fig. 8.29.
The experimental response of a ribbon microphone with a circular baffle is
shown in Fig. 8.30. The agreement between the measured response and
the computed response is quite good. There is some deviation between
RIA = .5 and RIA = .8. There is also some discrepancy in this region
between computed and measured pressures (Figs. 8.25 and 8.26). It is
interesting to note that the theoretical response of the pressure gradient
microphone and the ribbon in a baffle is practically the same, Figs. 8.22
and 8.29.
The measured directional characteristic of the ribbon microphone with ,
a circular baffle is shown in Fig. 8.31. It will be seen that, for small values
of RIA, the directional characteristic corresponds to a cosine function. Be
tween RIA = i and t the directional pattern is sharper than a cosine char
acteristic. Then for RIA larger than t the characteristics broaden and
assume irregular shapes. The theoretical directional characteristics em
ploying equations 8.58, 8.61, 8.66, and 8.67 are shown in Fig. 8.32. It will
be seen that the agreement with the experimental results of Fig. 8.31 is
MICROPHONES 285
quite good. There is some deviation for DIA = t. It is in this region that
deviations occurred between the theoretical and experimental results for
the pressure, Figs. 8.25 and 8.26, and for the response, Figs. 8.29 and 8.30.
The theoretical directional characteristic for a doublet, Fig. 8.23, becomes
progressively broader for RIA =i, t, and i and does not agree at all with the
experimental results. For RIA = !, t, and 1 the shape of the theoretical
directional characteristic of the doublet does not correspond with the
experimental results. Summarizing, the theoretical directional characteris
tics of a ribbon microphone with a circular baffle agree within a few per
0
--.....
.9 ""
.e r\
~
'" .7
~
~ .6
...J
...J 1\
~.5
Z
-,4
f
~
~.3
o
.2
-
.1 I\"
1/ W \~ 'I\~
o 2 3 4 5 • 7 a8 2 3 4 56789 2 3 4 5 e 788
.01 .I 1 10
1.0
- ..... ~
.9
i\
.8
.2
.1 II \\II,,..
II
o 234S178' 23451789 2 3
hn
4 5 . 788
.01 .1 I 10
I-t+HMtCH++-+oo
0
r-r-.
60 ............
30
0
'" \
\
1\
6
30 -I
9 I 2 3 4 $, 67891 2 3 -4567811 2 3 4
.01 .1
FIG. 8.33. The phase angle, in degrees, between the actuating force and the
particle velocity for a mass-controlled ribbon with a circular baffle as a function of
RIA. The phase angle between the voltage output of a mass-controlled, electro
dynamic ribbon located in a magnetic field as a function of RIA.
The above analysis has been concerned with a ribbon located in a cir
cular baffle. Irregular baffles instead of circular baffles are used in com
mercial microphones for two reasons: first, a suitable magnetic field results
in an irregular baffle and, second, the sound path lengths between the two
sides of an irregular baffle differ and, as a consequence, it is possible to
obtain uniform directional response characteristics over a wide frequency
288 ACOUSTICAL ENGINEERING
270'H-++~~+-r4-JgO 270
11000~
1.0
270f-+-+-t--al1!E-+-t-H
FIG. 8.34. The directional characteristics of the velocity microphone shown in Fig. 8.24.
The polar graph depicts the output, in volts, as a function of the angle, in degrees. The
maximum response is arbitrarily chosen as unity.
are shown in Fig. 8.34. Further, the deviation from a cosine characteristic
is very small.
The above considerations have been concerned with a plane wave. As in
the case of the pressure gradient microphone, it can be shown that the
output of a baffle-type velocity microphone corresponds to the particle
velocity in a spherical wave. The response of a baffle-type velocity micro
phone as a function of the distance from a point source and the frequency
is shown in Fig. 8AOA.
The response of the baffle-type velocity microphone may be obtained
from the acoustical network of Fig. 8.35 and the acoustical impedance
characteristics of the acoustical elements of the system of Fig. 8.36. The
ribbon is 2.2 inches in length and .2 inch in width. The flux density is
9000 gausses. The open-circuit voltage generated of the ribbon is given
by equation 8.26. The computed voltages are indicated by the dots in
MICROPHONES 289
RIBBON TRANS. LINE TRANS. V.T.
RIBBON TERMINAL
r[1811~,=-----=,III~
'
M::~~E
~
M .rA2
~
ELECTRICAL SYSTEM
MsfAS
MRCAR
~J
A'
,RONT SECTION
VIEW A-I(
ACOUSTICAL NETWORK
FIG. 8.35. Front and sectional views of the vibrating system of a velocity
microphone. In the electrical circuit, e = the open circuit voltage. rEI =
the electrical resistance of the ribbon. ZE2 = the electrical impedance of the
load upon the ribbon due to the transformers and vacuum tube. ZEA = the
electrical impedance due to the acoustical system. In the acoustical network,
rAI. M 1, rA2, and M 2 = the acoustical resistances and inertances due to the
air load on the front and back of the ribbon. rAS and Ms = the acoustical
resistance and inertance due to the slit between the ribbon and the pole
pieces. MR and CAR = the inertance and acoustical capacitance of the
ribbon. ZAE = the acoustical impedance due to the electrical system. PI and
P2 = the sound pressure at the front and back of the ribbon.
IOOr---~-----r----'-'-----'-----~---'-'-----r----~----'-~---'
10
'"
u
z
«0
'"
Q.
..J
« .Q.
~
f-
iI:
<l
(/)
=>
0
u
«
01
~--~----~-----L-IOLO~~~~~L-~~LIO~OLO~--L-----L---~'~O~00~O~-200~~~
FREQUENCY IN CYCLES PER SECOND
the voltage in the line. This is .0085 volt per dyne per square centimeter
at the grid of the vacuum tube.
In general, the electrical load, ZE2, of the transformer and vacuum tube
is large compared to the electrical resistance of the ribbon and the motional
electrical impedance, ZEA. Under these conditions the voltage delivered at
the grid is the same open-circuit voltage developed by the ribbon multiplied
by the step-up ratio of the transformers. However, if the electrical resist
ance of the ribbon and the motional electrical impedance are comparable in
magnitude to the electrical load impedance, cognizance must be taken of the
electrical resistance of the ribbon and the motional electrical impedance
when computing the voltage developed across the load. In the equivalent
electrical circuit of Fig. 8.35, the motional electrical impedance, in abohms,
is given by
(Bl)
ZEA = A'RZAP
8.68
89 Olson, H. F., Jour. Acous. Soc. Amer., Vol. 3, No.3, p. 315, 1932.
40 Weinberger, Olson, and Massa, Jour. Acous. Soc. Amer., Vol. 5, No.2, p. 139, 1933.
41 Olson, H. F., Jour. Soc. Mot. Pic. Eng., Vol. 27, No.3, p. 284,1936.
A'
fRONT VIEW CROSS- SECTION A-A
Response Ratio =
j(~r + cosr 8)2 + (A cos 8)221T1'2 8.71
(~r
This ratio for 8 = 0, 30°, 60°, 90 120°, 150 , and 180 0 for 1, 2, and 5 feet
0 0
is shown in Fig. 8.40. The same ratio for a conventional velocity micro
phone for 1, 2, and 5 feet is shown in Fig. 8.40. It will be seen that the
accentuation in the unidirectional microphone is smaller than in the case
of the velocity microphone.
2. Efficiency of Energy Response to Random Sounds of the Unidirectional
Microphone as a Function of the Relative Sensitivities of the Bidirectional
and N ondirectional Microphones. 44_The unidirectional microphone consists
of the combination of a bidirectional microphone, in which the output is a
function of the cosine of the angle of incidence, and a nondirectional micro
phone. In general, it is customary to make the output of the bidirectional
microphone, for 8 = 0, equal to the nondirectional microphone. For this
condition the directional characteristic is a cardioid of revolution. In the
case of both the bidirectional and the cardioid unidirectional microphones,
the ratio of energy response to generally reflected sound is one-third that of a
nondirectional microphone. It is interesting to investigate the efficiency
of response to random sound of other ratios of sensitivity of the bidirectional
to the nondirectional unit.
The voltage output of a microphone consisting of a bidirectional and non
directional unit is given by
eUD = Rl + R2 cos 8 8.72
where Rl = voltage putput of the nondirectional microphone, and
R2 = voltage output of the bidirectional unit for 8 = O.
w 41\
II)
;Z
~
,
\ 1\
:Jl4 \ l'~
~ \\
~~
,
:33 \ r-..'
0.
w3 a:
a:
\ i\\
w i\ I FT. w ~
rr'
..t!. 0·
>2 2:2
~ "- I
«...J ~ 30~YI If
W
...J ....fJF . r--- ....... .... w
a: I l'l...~0 90· I
a: I
Ill~ ~ .I,r~~o·
A 0 00 100o B~
I
100
-rI'#1000
FREQUENCY FREQUENCY
:;j4
;Z
0
l'~
0.
II)
l1 3
\' ~~
2:2 \.' ~'
w
I
«
, ~ "-30
o·
1'60~
w
2:2
« ~
I-
-";::t-. O·
.....30" ..-I6C1'
...J
I~ "-
...J
w 180· 90" w 9O·
a: I a:
U;O· ~ 120"... ) f::::!=I. I , 120"
cOzo
I ~
DO20
IsO" .....
100 1000 100
fREQUENCY FREQUENCY °
FIG. 8.40. A. The relative voltage output of a velocity (or pressure
gradient) microphone as compared to a nondirectional pressure micro
phone for distances of 1, 2, and 5 feet. E, C, D, the relative voltage
output of a unidirectional microphone as compared to a nondirectional
pressure microphone for distances of 1, 2, and 5 feet and for various
angles of the incident sound.
8
V ~
~ I- II'" -- V
.7
/"
V
c .6 - -
z
V
-I'" _5 /"
~8 ~e-=
c
'" .4 /'
V
3,V T--!-' -
-
I
.1
0
o 2 3 4 5 6 7 8 9 10 20 30 40 <Xl
R,+ R2
and the cardioid unidirectional is the same. However, for 0 < Rl/R2 < 1
the response to random sounds is less than in the case of either of these two
microphones.
3. Efficiency of Energy Response to Random Sounds of a Unidirectional
Microphone as a Function of the Phase Angle between the two Units. 45 -The
preceding discussions have assumed that the phase angle between the out
puts of the two units did not change with frequency. There are two
principal sources of phase shift between the two units, namely, a phase
shift due to a finite separation, and a phase shift due to a difference in the
phase frequency characteristics.
Consider the case in which there is a phase shift c/> between the output
of the bidirectional and nondirectional units. The output of each separate
unit is eo volts. The output of the combination is
27Te02 fo" [(cos 28 + 2 cos 8 cos cP + cos 2 cp) + sin 2 CPJ sin 8 d8
8.75
167Te0 2
Directional efficiency
BIDIRECTIONAL-~~~----L-7>~--.--71"'~-----r.~-----"'tt---
NONDIRECTIONAL------"tt-----i1H-f--.+--+-t------''--t--t_-f---'Lf------\--+_
R, R,
-=1 -=00
R2 R2
ENERGY RESPONSE .L
TO RAfilDOM SOUND 3
27Te02 Jo" {[cos 8 + cos (K cos 8)]2 + [sin (K cos 8]2} sin d8 1
16n-e0 2 3 8.79
That is, the efficiency is independent of the separation between the units.
Of course, for very large distances the separation disturbs the response for
8 = O. However, in the conventional microphone this does not occur.
Therefore, the effect of finite size has no effect on the efficiency of energy
response to random sounds.
4. Distortion of the Directional Pattern in the Unidirectional Microphone.
-Deviations from the cardioid characteristic in the unidirectional micro
phone are due to
1. Phase shift in the velocity microphone due to deviation from a pure
mass reactance
2. Phase shift in the velocity microphone due to diffraction
3. Phase shift in the pressure microphone due to deviation from resistance
control
4. Phase shift in the pressure microphone due to diffraction
5. Deviation in the output from a cosine directional characteristic in the
velocity microphone
6. Deviation in output with angle in the pressure microphone
7. Unequal sensitivity of the two elements
The phase angle between the output of a velocity microphone and the
particle velocity in a plane wave has been considered in Sec. 8.3B. The
phase shift in a pressure ribbon microphone has been considered in Sec.
8.2D3. It is possible to adjust these phase shifts and those due to diffrac
tion so that the cancellation for 180 0 will be of the order of -30 db up to
10,000 cycles. In the case of the dynamic pressure unit the problem of
maintaining appropriate phase shifts is more difficult.
B. Single-Element Unidirectional Microphones. 46 -Unidirectional micro
phones consisting of the combination of a nondirectional and a bidirectional
microphone have been described in Sec. 8.4A. It is the purpose of this
section to describe single-element unidirectional microphones in which a
unidirectional pattern is obtained by combining a single-element electro
acoustic transducer with a phase-shifting network.
1. Phase-Shifting Unidirectional Microphone.-A unidirectional micro
phone consisting of a nondirectional and bidirectional microphone has been
described in the preceding section. It is the purpose of this section to
describe other means for obtaining directional response.
The elements of a phase-shifting microphone are shown in Fig. 8.43.
The open ends of the pipes are separated by a distance D. A bend of
46 Olson, H. F., Jour. Soc. Mot. Pic. Eng., Vol. 52, No.3, p. 293, 1949.
298 ACOUSTICAL ENGINEERING
length d is placed in the shorter pipe. The ribbon element measures the
difference in pressure between the two pipes. The difference in pressure
between the two pipes is given by
D..p =
. (d-D
2po SIll -A- 7T + D7T
T cos 8) 8.80
....... .............
d=D d=2D
- 110 180
,"
180 180 180
FIG. 8.43. A directional microphone employing a phase shifting system. The polar
graphs show the directional characteristics for various ratios of diD. The polar graph
depicts the output, in volts, as a function of the angle, in degrees. The maximum
response is arbitrarily chosen as unity.
If the distances D and d are small compared to the wavelength, D..p will
be proportional to the frequency. If a mass-controlled, electrodynamic
element is used, the output will be independent of the frequency.
A series of directional characteristics for various ratios of D to d is shown
in Fig. 8.43
2. Polydirectional Microphone.-The single element polydirectional
microphone47 is shown in Fig. 8.44. The ribbon is located in the air gap
formed by the pole pieces. A permanent magnet supplies the flux to the
air gap. The entire one side of the ribbon is covered by the labyrinth
connector. The connector, in turn, is coupled to a damped pipe or labyrinth.
The type of directional characteristic is governed by the size of the aperture
in the labyrinth connector.
The action of this microphone can be obtained from Fig. 8.45 which
shows the schematic view of the microphone and the acoustical network.
The sound pressure acting on the open side of the ribbon may be written
Pi = POlEi(wt+4>l) 8.81
47 Olson, H. F., Proe. Insf. Rad. Eng., Vol. 32, No.2, p. 77, 1944.
MICROPHONES 299
where POI = amplitude of the pressure, in dynes per square centimeter,
w = 27Tj,
The sound pressure acting on the aperture in the labyrinth connector may
be written
8.82
where P02 = amplitude of the pressure, in dynes per square centimeter, and
q,2 = phase angle with respect to a reference point, in radians.
M. r.s
Ms TAS
MArAA
r
MRCAR:
~2rA2 rA •
P,
ZA2 1.12 TA •
~
r' 3 z., r' 3
VIEW VIEW
FIG. 8.45. Front view, cross-sectional view, and the acoustical network of
a polydirectional microphone. In the acoustical network, ME and CAR =
the inertance and acoustical capacitance of the ribbon. MA and 1'AA = the
inertance and acoustical resistance of the air load on the front of the ribbon.
M sand r AS = the inertance and acoustical resistance of the slit between the
ribbon and pole pieces. M 2 and r A 2 = the inertance and acoustical resis
tance of the aperture in the pipe. l'A3 = the acoustical resistance of the
damped pipe. ZAE = acoustical impedance due to the electrical system.
PI = the sound pressure at the front of the ribbon. P2 = the sound pres
sure at the back of the connector.
The phase angle 1>3 is a function of the angle of the incident sound as
follows:
1>3 = 1> cos (J 8.85
where (J = angle between the normal to the surface of the ribbon and the
direction of the incident sound,
1> = phase angle for (} = 0, and
1> = function of the frequency.
The volume current, in cubic centimeters per second, of the ribbon due to
the pressure PI is
Xl = PI(ZA2 + ZA3)
8.86
ZA1ZA2 + ZA1ZA3 + ZA2ZA3
where I:1p = Pic/> the difference in pressure between the two sides of the
ribbon. Equation 8.91 will be recognized as that of the velocity micro
phone. The directional characteristic is bidirectional.
In Fig. 8.46E the aperture is closed. In this case the acoustical im
pedance ZA2 is infinite. Under these conditions the pressure pz is ineffective.
From equations 8.86, 8.87 and 8.88 the volume current of the ribbon is
given by
XR = Xl ~ X2 = Xl = PI 8.92
ZAI + ZA3
Equation 8.92 will be recognized as that of the pressure ribbon microphone.
The directional characteristic is nondirectional.
Using an aperture which may be varied, it is possible to obtain any
limacon characteristic between the cosine bidirectional Fig. 8.46A and the
302 ACOUSTICAL ENGINEERING
"u
M, ZAI rAI
~
r"' l rA2 ZA.
~.
r.,
ZA'
r.,
A
M, ZA'
rA,
·'n'
"'""""'"
rAI TA2 !.. ZA.
r..
~. 'W
Z. . rAJ
rAJ
B
M, ZA'
TA,
·'n'
rAI rA2
TA•
p.
M.
lA'
M,
C
lA.
rAJ
lA'
TA,
TA•
·'tr
rAI 1"2
TA•
p.
M.
ZAl
D
lA. TA•
TAJ
., 0·'
M, ZA' TA,
ZAl TAl
rA•
E
ACOUSTICAL ACOUSTICAL DIRECTIONAL
SYSTEM NETWORK CHARACTERI STIC
ACOUSTICAL NETWORK
ACOUSTICAL NETWORK
SECTION A-A
FRONT VIEW
MAGNET
RIBBON
POLE
ACOUSTICAL NETWORK
FRONT VIEW
FIG. 8.50. Front and sectional views and the acoustical network of
a ribbon-type unidirectional microphone with lumped acoustical
elements. In the acoustical network, Pl = the sound pressure on
the front of the microphone. MA = the inertance of the air load
upon the ribbon. Mil and CAll = the inertance and acoustical
resistance of the ribbon. CAl = the acoustical capacitance of the
cavity behind the ribbon. M 1 and I'Al = the inertance and acousti
cal resistance of the cloth covering the chamber. pz = the sound
pressure on the back of the microphone.
SECTIONAL VIEW
r Mo Co rA2 M3 MI
M4
fA4
rA3
I I CA2
ACOUSTICAL NETWORK
two vibrating systems are placed back to back. The cavities behind the
diaphragm are interconnected by small holes. The phase shift in the
vibrating system combined with the electrical polarizing system makes it
possible to obtain a variety of directional characteristics as well as a non
directional characteristic. With the potentiometer at the full negative
position a bidirectional pattern is obtained. With the potentiometer set
at the zero position a cardioid type of directional pattern is obtained. With
the potentiometer set at the full positive position a nondirectional pattern is
obtained.
7. Dipole Microphone. 56-A dipole microphone is a microphone in which
the response is a function of the sound pressure between two distinct points.
56 Olney, Slaymaker, and Meeker, Jour. Acous. Soc. Amer., Vol. 16, No.3, p. 172,
1945.
MICROPHONES 309
FRONT
DIAPHRAGM
..-__--ll-
CIRCUIT DIAGRAM
FRONT VIEW SECTION A-A'
FIG. 8.52. Front and sectional views and circuit diagram of the vacuum tube
amplifier of a directional condenser microphone.
ACOUSTICAL NETWORK
MlcAlrAI
ZA2 Pz
ZA4
ACOUSTICAL SYSTEM
~ CAl
the talker's mouth and yet retain the acoustical advantage of a close
talking microphone. The microphone and telephone receiver are made
an integral unit in a telephone operator's set. A disk of silk cloth covers
the end of each tube. The acoustical resistance termination practically
eliminates the resonance in the tubes.
The performance of the system may be determined from a consideration
of the acoustical network. The two pipes are represented as acoustical
310 ACOUSTICAL ENGINEERING
~~~
~ I C
.' T C. a ~2
ACOUSTICAL NETWORK
A
,---..,
/' , \
I \
I \
I I -,=:,--:",-C. 2
\
, /
I
M2r.a 1>2
" ........ _-/ /
A'
FRONT VIEW SECTION A-I<
FIG. 8.54. Front view, sectional view and acoustical
network of the differential microphone. In the
acoustical network, M I , rAI, and M 2, rA2 = the
inertances and acoustical resistances of the two holes
in the case. C,i1 and CA2 = the acoustical capaci
tances of the air chambers on the two sides of the
diaphragm. M3, rA3, and CA3 = the inertance, acous
tical resistance, and acoustical capacitance of the
diaphragm. M 4 , rA4, and CA 4 = the inertance, acous
tical resistance, and acoustical capacitance of the carbon
elements. PI and P2 = the sound pressures at the two
holes in the case.
8.98
where Dl = distance between the points in the pairs of points, and
D2 = distance between the two pairs.
A second-order gradient microphone may be made up of two oppositely
phased first-order gradient microphones as shown in Fig. 8.55.
The acoustical network of the acoustical system of one of the units in
the second-order microphone is shown in Fig. 8.55. The controlling element
in the system is an acoustical resistance. The transducer is of the dynamic
type. Therefore, in a plane wave, the voltage output of a single unit will be
proportional to the frequency. Connecting two of the units in opposition
the voltage output of the second-order gradient microphone will be pro
portional to the square of the frequency. However, as a close-talking
microphone the output will be independent of the frequency. It possesses
the antinoise characteristics of a close-talking, second-order, gradient
microphone. The directional characteristics of the second-order gradient
microphone, as equation 8.98 shows, are bidirectional and proportional to
the square of the frequency.
B. Gradient Microphones of Any Order.62-The general expression for
the actuating pressure for a microphone of any order n for any two points
separated by a distance 8r is is
I5np =
onp n 0
or n 8r =orn (- J
.PM
r
Eik(ct-r») (I5r cos 8)n 8.99
59 Olson, H. F., Jour. Acous. Soc. Amer., Vol. 17, No.3, p. 192, 1946.
61 Olson, H. F., Jour. Acous. Soc. Amer., Vol. 17, No.3, p. 192, 1946.
62 Olson, H. F., Jour. Acous. Soc. Amer., Vol. 17, No.3, p. 192, 1946.
CAl
M3f A3 C A3
rAI
~ 101, UNIT I
eA21t~rtli~~~~~~
DIAPHRAGM VOICE
COILS
UNIT 2
VOICE
COIL
SECTIONAL VIEWS
ELECTRICAL CONNECTION
FIG. 8.56. The directional characteristics of gradient microphones of order zero, one,
two, three, and four.
Equation 8.99 shows that the pressure available for driving the micro
phone is proportional to the nth power of the frequency. The directional
characteristics are bidirectional cosine functions, the power of the cosine
is the order of the gradient. The directional characteristics for gradient
microphones of orders zero, one, two, three, and four are shown in Fig. 8.56.
C. Noise Discrimination of Gradient Microphones. 63-Gradient micro
phones of order one and higher are directional. Therefore, these micro
63 Olson, H. F., Jour. Acous. Soc. Amer., Vol. 17, No.3, p. 192, 1946.
MICROPHONES 313
phones discriminate against sounds from random directions. The magnitude
of the discrimination is given by the expression in equation 8.117 as follows,
if!
...J
W
III
Li
w
o
40
30
'"'"
~
""~"
SECOND
~DER
~20 b-,.
~
w
if!
~ FIR~
--
ti 10
ORDE
~
Q.
if!
W
0::
ORD
r I---
o
.2 .4 .8 I 2 4 8 10
2rrr
DISTANCE T
~:: 280 5&0 1120 2800 5600 11200
IYz 140 280 560 1400 2800 5600 11200
3" 70 140 280 700 1400 2800 5600
FREQUENCY IN CYCLES PER SECOND
The above equation assumes that the distance between the origin of the
sound and the microphone is greater than nA, where n is the order of the
gradient, and A is the wavelength.
314 ACOUSTICAL ENGINEERING
ZERO
ORDER
o
...
III
~ I
.... ~
..J
!oJ
m
U -to FIRST ~
!oJ
o
V V
!!: -20
./ ~
!oJ V /'
V
III
~ -30
/' 5~g~~~
V
V
A
rt)
!oJ
a:
-40
-50
10 2 2
/
4 8 103 4 8 104
FREQUENCY IN CYCLES PER SECOND
ZERO
ORDER
o
III
..J
r-
l--- -
W
III
8o -10
~ -20
fiRST
ORD;!Y'
,..,. ~ ------
/
,/
w
III
~ -30
/ /
III
w
" -40 V SECO~
ORDER
/
-50
102 2
V
4 a 103 2 4 8 10"
fREQUENCY IN CYCLES PER SECOND
65 Olson, H. F., Jour. Acous. Soc. Amer., Vol. 17, No.3, p. 192, 1946.
in Sec. 8.4. It is the purpose of this section to consider higher order com
bination gradient microphones with unidirectional characteristics.
A higher order unidirectional gradient microphone may be obtained by
combining two first-order gradient microphones with a delay system as
shown in Fig. 8.60. The voltage output of this system is given by
ez = eo(Dz + Dl cos 8) cos 8 8.102
where eo = reference voltage output,
Dl = distance between the first-order gradient elements, and
Dz = path length of the delay.
Equation 8.102 holds for the frequency range in which Dl and Dz are small
compared to the wavelength. The reference voltage output is a function
EBCXJ
e,,=e~ (1+3cose)cose
of the frequency and the type of electro acoustical generating system. The
maximum discrimination against random sounds occurs when Dz = iDl .
For this condition the energy response to random sounds is one-eighth
that of a nondirectional microphone. This is a very high order of directivity.
The directional characteristics for two different conditions are shown in
Fig. 8.60.
The system of Fig. 8.61 consists of two combination pressure and pressure
gradient microphones, described in Sec. 8.4, and a delay system. A number
of combinations are possible in this system, as for example, combination
units with various delays and dissimilar combination units with various
delays. The directional characteristics for two different conditions are
shown in Fig. 8.61.
E. Second-Order Gradient Uniaxial Microphone. 66-A second-order
gradient microphone with a unidirectional directivity pattern consisting of
the combination of two unidirectional microphones each consisting of two
66 Olson and Preston, RCA Review, Vol. 10, No.3, p. 339, 1949.
MICROPHONES 317
~
=
===
=
c::=> =
= 0
==c:=:::=:::::::: =
o c::::==::::::=> c:::=::::J.
== =
= = =
===
FRONT SCREEN CONNECTOR SCREEN
microphones of the type described in the preceding section could be used for
the elements. It also appeared that a high order of directivity was not
required in the high-frequency range. The microphone which was developed
operates as follows: The system is of the second-order gradient type up to
1000 cycles; there is a transition from second- to first-order gradient from
2000 to 4000 cycles, and above 4000 cycles the directivity pattern is the
same as the uniaxial microphone. As described in the preceding section,
the directional pattern of the uniaxial microphone is sharper than a cardioid.
Specifically, the directivity pattern up to 2000 cycles is given by
Typical microphones of this classification are reflector, lens, and line micro
phones. The second class of microphones has been considered in Secs.
8.3, 8.4, and 8.5. The dimensions of gradient microphones, as contrasted
to wave-type microphones, are small compared to the wavelength. It is
the purpose of this section to consider two examples of wave microphones,
namely, the parabolic reflector microphone and the line microphone.
A. Parabolic Rejlector. 67 ,68,69,70,71-Reflectors have been used for years
for concentrating and amplifying all types of wave propagation. The
surface of the parabolic reflector is shaped so that the various pencils of
incident sound parallel to the axis are reflected to one point called the
focus (Fig. 8.65). To obtain an appreciable gain in pressure at the focus,
the reflector must be large compared to the wavelength of the incident
sound. This requirement of size must also be satisfied in order to obtain
200 '" 1000'"
SECTIONAL VIEW
68 Hanson, O. B., Jour. Acous. Soc. Amer., Vol. 3, No. 1, Part 1, p. 9, 1931.
69 Hanson, O. B., Jour. Acous. Soc. Amer., Vol. 3, No. 1, Part 1, p. 81, 1931.
70 Dreher, Carl, Jour. Soc. Mot. Pic. Eng., Vol. 16, No. 1, p. 29, 1932.
71 Olson and Wolff, Jour. Acous. Soc. Amer., Vol. 1, No.3, p. 410,1930.
MICROPHONES 321
that the directivity increases with frequency. For example, the system
is practically nondirectional at 200 cycles. On the other hand, the direc
tional characteristiC is very sharp at 8000 cycles.
B. Lens Microphone. 72, 73_A lens microphone consists of an acoustic
lens arranged so that the pencils of sound arrive in phase at a common
point termed the focus, and a microphone located at focus. A schematic
sectional view of a pressure microphone located at the focus of an acoustic
lens is shown in Fig. 8.66. The directivity pattern is determined by a
INCIDENT SOUND
ACOUSTIC
LENS
relationship between the dimensions of the lens and the wavelength similar
to that of the parabolic reflectormicrophone.
C. Large-Surface Microphone.-A large-surface microphone,74 in the form
of a large number of dynamic microphone units arranged on a spherical
surface, is shown in Fig. 8.67. A curved surface source has been considered
in Sec. 2.20. The microphone shown in Fig. 8.67 is based upon the funda
mental principles of a curved surface radiator or receiver. The angular
spread of the microphone shown in Fig. 8.67 is about 50°. The diameter is
four feet. The directivity pattern follows that of a curved surface system
of this diameter. The directivity pattern is reasonably uniform above 300
cycles. The low-frequency limit of uniform directivity could be extended
by employing a microphone of a larger diameter.
A large-surface microphone 75 in the form of a condenser microphone
consists of a nonstretched diaphragm designed to vibrate in phase over its
entire surface. The operating elements consist of an aluminum foil front
electrode cemented to a 3J32-inch foam rubber sheet mounted on an
aluminum back plate. The whole assembly is fitted in a picture frame
18 inches square and protected by a perforated metal face plate. The
directivity pattern is that of a square plate, considered in Sec. 2.17.
72 Clark, M. A., Jour. Acous. Soc. Amer., Vol. 25, No.4, p. 829, 1953.
73 Clark, M. A., Trans. IRE, Prof. Group on Audio, Vol. AU-2, No. 1. p. 5, 1954.
75 Aamodt and Harvey, Jour. Acous. Soc. Amer., Vol. 25, No.4, p. 825, 1953.
MICROPHONE UNITS
0
0
0
0 0
0
0
0
0
0
0
0
~
0 0 0 0 0 0
0 0 0 0 0
0 0
0 0 0 0 0 0
0 0
0 0 0 0 0 0
0 0
0 0 0 0 0 0 0
0 0 0
0 0 0 0
0 0 0 0 0
0
0 0 0
0 0 0
0 0
0 0
0 0 0
FRONT VIEW SECTIONAL VIEW
76 Olson, H. F .• Jour. Inst. Rad. Eng.• Vol. 27. No.7. p. 438. 1939.
77 Mason and Marshall. Jour. Acous. Soc. Amer., Vol. 10. No.3. p. 206. 1939.
MICROPHONES 323
minated in an acoustical resistance. Under these conditions the output of
the pipes can be added vectorially.
1. Line Microphone: Useful Directivity on the Line Axis. Simple Line.
This microphone consists of a number of small pipes with the open ends,
as pickup points, equally spaced on a line and the other ends joined at
a common junction decreasing in equal steps (Fig. 8.68). A ribbon element,
connected to the common junction and terminated in an acoustical resistance
in the form of a long damped pipe, is used for transforming the acoustical
vibrations into the corresponding electrical variations.
The contribution, in dynes per square centimeter, by any element n at
the common junction of the microphone may be expressed as
Re = ~
1
IJ-1/2
12
/ £21Tj(X-X cos e)/A dx I 8.109
. ~ (l - 1 cos 8)
sin 1\
8.110
~ (l - l cos 8)
The directional characteristics of the microphone of Fig. 8.68 for various
ratios of length of the line to the wavelength are shown in Fig. 8.69. These
characteristics are surfaces of revolution about the line as an axis. This
microphone is useful for collecting sounds arriving from directions making
small angles with the microphone axis.
324 ACOUSTICAL ENGINEERING
180 180
'80
sin i (l - I cos (J + d)
Re= 8.112
X(l - I cos (J + d)
The directional characteristic of the microphone of Fig. 8.70 for various
ratios of the length of the line to the wavelength, and for a delay path of
one-fourth times the length of the line is shown in Fig. 8.71. Comparing
LENGTH =~ LENGTH =~
2
LENGTH =A LENGTH =2}. LENGTH =4.>.
w.•
'00 .
,
FIG. 8.71. The directional characteristics of the microphone shown in Fig. 8.70 for a
time delay of one-quarter the length of the line as a function of the ratio of the length of
the lines to the wavelength. The polar graph depicts the output, in volts, as a function
of the angle, in degrees.
•
Fig. 8.71 with Fig. 8.69, it will be seen that the same directional characteristic
can be obtained with a shorter line by introducing appropriate delay. In
the case of a delay path comparable to the wavelength, loss in sensitivity
occurs.
3. Line Microphone: Useful Directivity on the Line Axis. Two Lines
and a Pressure Gradient Element.-This microphone consists of two lines
of the type shown in Fig. 8.70 arranged so that the ribbon element measures
the difference in the pressures generated in the two lines (Fig. 8.72). The
centers of the two lines are displaced by a distance D. In the line nearest
the element, a bend of length D is inserted between the junction and the
ribbon element.
To show the action of the pressure gradient system, assume that the
length of all the small pipes is the same and the openings between the two
sets are separated by a distance D. Under these conditions the line systems
are nondirectional.
The difference between the forces on the two sides of the ribbon, assuming
326 ACOUSTICAL ENGINEERING
that the mass mechanical reactance of the ribbon is large compared to the
mechanical resistance of the damped pipes, may be expressed as
fM = . (7TD A
A cos (27Tft) sm cos fJ) 8.113
Equation 8.113 shows that the force available for driving the ribbon is
proportional to the frequency and the cosine of the angle fJ.
Employing mass-controlled ribbon of mass m r , the velocity is given by
~~ DELAY
~",Q
FIG. 8.72. Line microphone. Useful directivity on the line axis. This microphone
consists of two lines of the type shown in Fig. 8.70 displaced by a distance D along the
axis. In the line nearest the ribbon element a bend is inserted which introduces a path
length D. The ribbon element measures the difference in pressure in the two lines.
FIG. 8.73.
'8.
-. '80
H
'"zlOO
VI
!> @ ® ® 0
~ r-.. /
'"
o l'\.
~
/ / /
80
eo J '\7 '\ J ,,/ J
I
~40
A P\ IA l)\.
u
a: ZO
7 ~ [7
"-I"; F- ~ / ~
<oJ
0.
J;'
X ~v r-.....
3 '" .5 6 7 8 ' IO~ ... , 6 7 8.'0.. .
FREQUENCY I N CYCLES PER SECOND
FIG. 8.75. Voltage response frequency characteristics of the units and electrical
filter system shown in Fig. 8.74.
700 CYCLES
A
FIG. 8.76. A. The directional characteristic of line 3 of Fig. 8.74 at 700 cycles. B. The
directional characteristics of lines 2 and 3 and the resultant at 950 cycles. C. The direc
tional characteristics of lines 2 and 3 and the resultant at 1250 cycles. D. The
directional characteristics of the microphone shown in Fig. 8.72 for the range from 85
to 8000 cycles fall within the shaded area.
MICROPHONES 329
lines 2 and 3 is a directional characteristic very close to Fig. 8.76A. The
directional characteristics of the microphone shown in Fig. 8.76 for the range
from 85 to 8000 cycles, except for the small lobes for angles greater than 90°,
fall within the shaded area of Fig. 8.76D. Considering that this microphone
has a frequency range of 6l octaves, it is a remarkably uniform directional
characteristic.
S.7. Throat Microphone. 8o ,81,82,83-The throat microphone is a micro
phone actuated by direct contact of the diaphragm with the throat. A
perspective view and a sectional view of a carbon-type throat microphone
are shown in Fig. 8.77. Since the acoustical impedance of the flesh of the
UNIT
~~=~~
FIG. 8.77. Complete throat microphone and sectional view of the carbon
type throat microphone unit.
81 Martin, D., Jour. Acous. Soc. Amer., Vol. 19, No. 1, p. 43, 1947.
82 Greibach and Pacent, Elec. Eng., Vol. 65, No.4, p. 187, 1946.
83 Erickson, J. R., Bell Lab. Record, Vol. 23, No.6, p. 193, 1945.
84 Olson and Carlisle, Proc. Inst. Rad. Eng., Vol. 22, No. 12, p. 1354, 1934.
85 Jones and Bell, Jour. Soc. Mot. Pic. Eng., Vol. 19, No.3, p. 219, 1932.
=0
8.9. Hot-Wire Microphone.-The hot-wire microphone consists of a
MOTION OF
AIR -----...:
LOOP
ANTENNA
where 1(if;) = ratio of the voltage output for incidence at the angle if; to that
for if; = 0, and
dn", = element of solid angle at the angle if;.
The directional efficiency of a microphone is a measure of the energy
response to reverberation noise and other undesirable noise.
In many systems in which the directional pattern cannot be expressed
in simple terms which can be integrated, the determination of the direc
tional efficiency must be carried out by numerical integration. The direc
tional efficiencies of cosine functions are easily determined. Directional
patterns which are powers of the cosine function are plotted in Fig. 8.80.
The directional efficiency for these patterns is also given. For the same
signal to random noise, reverberation, etc., the directional microphone
may be operated at Ilv'directional efficiency times the distance of a non
directional microphone.
86 Phinney, Thomas A., Trans. IRE, Prof. Group Audio, Vol. AU-2, No.2, p. 44,
1954.
87 Olson, H. F., Jour. Soc. Mot. Pic. Eng., Vol. 16, No.6, p. 695, 1931.
88 Olson, H. F., Jour. Acous. Soc. Amer., Vol. 3, No. 1, p. 56, 1931.
89 Olson, H. F., Proc, Inst. Rad. Eng., Vol. 21, No.5, p. 655, 1933.
90 American Standards Association Sectional Committee Z-24, Report on Calibra
tion of Microphones, Jour. Acous. Soc. Amer., Vol. 7, No.4, p. 300, 1936.
91 Baumzweiger, B., Jour. Acous. Soc. Amer., Vol. 11, No.4, p. 447, 1940.
92 Bauer, B.B., Trans. IRE, PGA 8, July, 1952.
332 ACOUSTICAL ENGINEERING
cos e
3 ..L cos3 e 3.7
14
cos4 e I
18 cos4 a 4.2
150 150
FLOW LINES
A B
FIG. 8.81. Wind screens for microphones. A. Bernoulli wind screen applied to a
dynamic microphone. B. Wind screen consisting of a wire frame covered with sheer
silk.
made in carbon materials in recent years and the distortion, although still
high, has been materially reduced.
8.14. Transient Response of Microphones.-The subject of transient
response of vibrating systems, together with applications to loudspeakers
has been considered in Sec. 6.25. The measurement of transient response
of loudspeakers will be considered in Sec. 1O.3G. The transient response of
a microphone may be predicted from the mechanical or acoustical network
of the vibrating system.
In the case of the vibrating system of the mass-controlled velocity micro
phone the response to transients is very good. The acoustical circuit of
Fig. 8.35 may be reduced to the simplified acoustical circuit of Fig. 8.82A,
provided the elements, M A and M R, the inertances due to the mass of the
A B c
ACOUSTICAL CIRCUITS
air load and the mass of the ribbon are the controlling elements. For the
audio-frequency range, the microphone may be designed so that the dif
ference in pressure, I:!..p, between the two sides of the ribbon is proportional
to the frequency (see Sec. 8.3). Under these conditions,
I:!..p = jwp' 8.118
where w = 27Tj,
j = frequency in cycles per second, and
P' = a sound pressure proportional to the free-field sound pressure,
in dynes per square centimeter.
Equation 8.118 shows that the acoustical circuit 8.82B is equivalent to
acoustical circuit 8.82A. From acoustical circuit 8.82B the volume current
lS
Ti
I NOISE
I
I
I
0 0 I
I
11 11
I
I
.,.. I
I I I I
I I I
AIR RIBBON ITRANSFORMERS I VACUUM TUBE I PLATE AND I
I I I I GRID RESISTORSI
I I
STUDIO MICROPHONE I AMPLIFIER
I
noise in the studio. The noise due to the random pressures upon the
diaphragm caused by the thermal agitation of the air molecules. The noise
due to the thermal agitation of the atoms in the diaphragm. The noise due
to the thermal agitation of the electrons in the conductor. The noise
due to the Barkhausen effect in the core of the transformer. The noise due
to shot effect, secondary emission, ionization, hum, etc., in the vacuum
tube. The noise due to the thermal agitation of the electrons in the plate
resistor.
A. Ambient Noise in the Studio.-The ambient noise in the studio is
usually one of the most important factors in determining the lower limit
of reproduction from the standpoint of the pickup system. The general
amhient noise level in a studio varies from 10 db for a very quiet studio
to 35 db for a noisy studio, as in the case of an audience. The spectrum
of room noise is shown in Fig. 12.38. It will be seen that room noise is not
uniform with respect to frequency. In the case of thermal noise the
generated voltage is proportional to the square root of the width of the
frequency band regardless of the position in the frequency spectrum.
B. Noise Due to Thermal Agitation of the Air Molecules.-Superposed
on the average atmospheric pressure there are fluctuations caused by the
distribution of thermal velocities of air molecules. The rms thermal sound
93' Olson and Preston, RCA Review, Vol. 7, No.2, p. 155, 1946.
336 ACOUSTICAL ENGINEERING
- Jf!2
P= J!l Pt 2 df = Jf!2J! 4kTr A df 8.120
where f2M! df = square of the thermal mechanical force in the interval df.
df = frequency interval, in cycles per second,
rM = mechanical resistance, in mechanical ohms,
k = Boltzmann's constant, 1.37 X 10-16 , and
T = absolute temperature, in degrees Kelvin.
D. Noise Due to Thermal Agitation of the Electrons in the Conductor.
The thermal agitation of the electrons in the conductor of the electrical
system of a microphone generates a fluctuating voltage. 94 ,95 The voltage,
94 Johnson, J. B., Phys. Rev., Vol. 32, No. 1, p. 97,1928.
MICROPHONES 337
e, in abvolts, due to the thermal agitation of the electrons in a conductor
is given by
8.122
where k = Boltzmann's constant, 1.37 X 10-16 ,
T absolute temperature, in degrees Kelvin,
12 - /I = width of the frequency band, in cycles per second, and
rE = electrical resistance of the conductor, in abohms.
E. Noise Due to Barkhausen Effect in the Transformer.-In the magnetiza
tion of a piece of ferromagnetic material by continuously varying magneto
motive force the resultant flux does not vary in a continuous manner but
is made up of small steps. This phenomenon is termed the Barkhausen
effect. In a well-designed transformer the only source of Barkhausen noise
of any consequence is in the leakage reactance. Since the leakage reactance
is small the Barkhausen noise will be relatively small. Furthermore in
most high-grade transformer alloys the Barkhausen effect is also quite
small.
F. Noise in the Vacuum Tube.-There are a large number of sources of
noise in the vacuum tube. A few of these are shot effect, thermal noise
in the plate impedance, ionization, and hum. These noises are treated at
length in books 96 on vacuum tubes. The voltage generated in the plate
of a well-designed triode, with an amplification of 20, from all sources
except hum, is 2.8 X 10-5 volt. This is 1.4 X 10-6 volt at the grid terminals.
G. Noise due to Thermal Agitation of the Electrons in the Plate Resistor.
The noise voltage generated in the plate resistor can be obtained from
equation 8.122 in Sec. 8.15D.
H. Example of Noise in a Sound Pickup System.-It is the purpose of
this section to give the actual magnitude of the noise in each element of
a sound pickup system. For the studio a very low level will be assumed
namely, 10 db. The microphone will be the velocity type with a sensitivity
of 600 microvolts per dyne per square centimeter at the 2S0-ohm terminals
(see Sec. 8.3B). The final step-up transformer raises the impedance to
50,000 ohms at the grid of the triode vacuum tube. All noise voltages
will be referred to the grid terminals of the vacuum tube. The frequency
range is 30 to 15,000 cycles.
1. Ambient noise in the studio, 5.0 X 10-6 volt.
2. Noise due to thermal agitation of the air molecules, 2.5 X 10-6 volt.
3. Noise due to thermal agitation of the atoms in the ribbon vibrating
system, negligible.
4. Noise due to thermal agitation of the electrons in the ribbon, 3.5 X 10-6
volt.
5. Noise due to the Barkhausen effect in the transformer, negligible.
6. Noise in the vacuum tube, 1.4 X 10-6 volt.
96 Terman, .. Radio Engineers Handbook," McGraw-Hill Book Company, New
York, N.Y., 1943.
338 ACOUSTICAL ENGINEERING
o Q R T
y
Z
FIG. 8.84. Microphone shapes. A. Pressure carbon. B. Pressure carbon. C.
Pressure magnetic. D. Pressure dynamic. E. Pressure dynamic. F. Pressure
dynamic. G. Pressure inductor. H. Pressure dynamic. I. Pressure crystal or
dynamic. J. Pressure crystal. K. Pressure dynamic. L. Pressure crystal.
M. Velocity ribbon. N. Velocity ribbon. O. Pressure dynamic. P . Pressure
dynamic. Q. Pressure condenser. R. Pressure condenser. S. Pressure dynamic.
T. Pressure ribbon. U . Velocity ribbon . V. Unidirectional ribbon. W. Uni
directional dynamic or crystal. X. Unidirectional combination ribbon and dynamic.
Y. Unidirectional ribbon. Z. Unidirectional dynamic. D. . Unidirectional ribbon.
MICROPHONES 339
The above data show that the noises from all sources are comparable
in magnitude. In a microphone of lower sensitivity the electrical noise
sources in the conductor, resistor, and vacuum tube would be the limiting
factors. For this reason it is very important to employ high-sensitivity
microphones in wide frequency range and high-quality reproduction of
sound.
8.16. Shapes of microphones.-Microphones may be classified in many
different ways. One classification involves the type of response, namely,
pressure, velocity, or combination of pressure and velocity. Another
classification involves the type of transducer used to convert acoustical
variations into the corresponding electrical variations, as for example,
carbon, magnetic, dynamic, electrostatic, crystal, ribbon, etc. The con
figuration of the elements of a microphone is determined to a large extent
by the type of response and the transducer. The outside shape in turn is
largely determined by the configuration of the elements. A few typical
examples of microphone shapes are shown in Fig. 8.84. The actual number
of microphone shapes commercially available today is somewhere around
500. For this reason it is impossible to depict all of the different shapes.
However, most of the microphones in use today follow the general patterns
shown in Fig. 8.84. Some of the shapes shown are not necessarily the most
common. The drawings are not replicas but reasonably accurate sketches.
The pressure microphones with different types of transducers are as follows:
A and B, Carbon; C, Magnetic; D, E, F, H, I, K, 0, P, and S, Moving Coil,
dynamic; G, Inductor dynamic; T, Ribbon; I, J, and L, Crystal; Q and R,
Electrostatic. The velocity microphones with ribbon transducers are as
follows: M, N, and U. Unidirectional microphones with different trans
ducers are as follows: Wand Z, Moving coil; W, Crystal; X, Combination
ribbon and moving coil; V, Y, and ~, Ribbon.
9
MISCELLANEOUS TRANSDUCERS
The reluctance RD of the permanent magnetic field circuit and the reluc
tance RA of the alternating magnetic field circuit can be obtained from the
magnetic network of Fig. 9.2.
The first and third term on the right-hand side of equation 9.1 represent
a steady force, the second term represents a force of the same frequency
and the last term represents a force of twice the frequency of the current
in the coil. Equation 9.1 shows the necessity for the polarizing field in
order to obtain high sensitivity and to reduce second harmonic distortion.
The diaphragm in the bipolar telephone receiver is a circular plate clamped
at the edge (see Sec. 3.5). The effective mass of the diaphragm, when it is
a clamped plate, is one-third the actual mass of the diaphragm. The
effective area of the diaphragm is one-third the total area of the diaphragm.
The first resonant frequency is usually placed at 1000 cycles. The effective
compliance of the diaphragm can be obtained from the effective mass
and the resonant frequency for the frequency region at and below the
first resonant frequency of the diaphragm. Referring to the mechanical
network it will be seen that the system is stiffness controlled in the region
below the resonant frequency. This means that, for a constant driving
force,1M, the force applied to the compliance, CM3, of the ear cavity will
be independent of the frequency and hence the sound pressure in the ear
cavity will be independent of the frequency.
The sound pressure delivered by a bipolar telephone receiver to a cavity
as a function of the frequency is shown in Fig. 9.1. In the range below
the resonant frequency the response is independent of the frequency. At
the first resonant frequency of the diaphragm the response is very high.
Above the resonant frequency the amplitude decreases rapidly with fre
quency. The peak at 3000 cycles is the second resonant frequency of the
diaphragm.
The pressure response frequency characteristic labeled A, Fig. 9.1, was
obtained with no leak between the ear and the earcap. In all hard earcaps
a leak occurs between the ear and the telephone receiver and the acoustical
342 ACOUSTICAL ENGINEERING
~CHC~
flO TCMZTcM3
ELEC. CIRCUIT MECHANICAL NETWORK
30
CD
i\
a A , , \
,,
20
'"z
II) \
o
.. 10
,'B \ ./1\
II)
'" \
'"
100 1000 10000
f"REQUE:NCY
CROSS -SECTIONAL VIEW
in actual practice. The artificial ear (see Sec. 10.4B) introduces a leak
which corresponds to the leak between the ear and the earcap. It will
be seen that the effect of this leak is to reduce the response at the lower
frequencies. Those familiar with telephone receivers have noticed that
the low-frequency response is increased when the leak is reduced by pressing
the telephone receivers tightly against the ears.
Since the development of the bipolar telephone receiver by Alexander
Graham Bell the construction has remained essentially the same. Improve
ments have been made in sensitivity and response by the use of better
materials. However, the clamped plate diaphragm characterized by
prominent resonant peaks was retained. Referring to Fig. 9.1, it will be
MISCELLANEOUS TRANSDUCERS 343
seen that the peaks due to the first and second resonance fall within the
response range. These resonances not only introduce frequency distortion,
but increase the intensity of reproduction of clicks due to the poor transient
response (see Sec. 6.15).
A bipolar telephone receiver 2 has been designed in which all the prominent
resonances within the response range have been eliminated and the response
frequency characteristic improved both from the standpoint of uniformity
as well as the frequency range. The new telephone receiver is of the bipolar
SCHEMATIC VIEW
The response of this receiver was taken by measuring the pressure generated
in a plain cavity. This cavity is designated by the compliance CM4. The
holes in the grid covering the receiver proper are large enough to have no
reaction upon the response. A resilient screen of silk is mounted on the
back of this grill. The mass of this screen is very small and is lumped with
the diaphragm mass mo.
The electrical portion of the circuit consists of the winding electrical
resistance rEI and inductance Ll. The eddy current elements are designated
ELECTRICAL
MECHANICAL NETWORK
NETWORK
30
'a" 20 --..... lA
'"
<II
Z
o
a.1 0
<II
'"0:
~o 100 1000 10000
CROSS - SECTIONAL VI EW FREQUENCY
as rE2 and L2. The electrical impedance3 due to the mechanical system is
designated by the motional electrical impedance ZEM. The force 1M can
be obtained from equation 9.1.
The pressure response computed by means of the mechanical network is
shown by the dots on the graph of Fig. 9.3. The measured pressure response
is given by the curve on this graph. The agreement is very good and shows
that it is possible to predetermine the response and to evaluate the effect of
3 Olson, "Dynamical Analogies," D. Van Nostrand Company, Princeton, N.J., 1943.
MISCELLANEOUS TRANSDUCERS 345
changes in the constants of the component parts. Comparing the response
of Figs. 9.1 and 9.3 it will be seen that large gains have been effected in
uniform response over the entire range and in sensitivity from 1500 to
3000 cycles.
The new magnetic telephone receiver4 shown in Figs. 9.4 and 9.5 differs
radically from any previous commercial telephone receiver. The novel
ELECTRICAL
MECHANICAL NETWORK
NETWORK
FIG. 9.4.
CROSS - SECTIONAL VI EW
@II I ~II~IIIIII
40 100 1000
FREQUENCY
10000
features are a ring-type armature and a new magnetic and vibrating system.
The driving armature is a disk of permandur. See Sec. 6.28. The action
and the performance of the magnetic system may be obtained from the
sectional view of the magnetic system and the magnetic network shown in
Fig. 9.5. The action and performance of the vibrating system may be
obtained from the mechanical network shown in Fig. 9.4. The diaphragm
is a dome-shaped phenolic impregnated fabric material. A major portion
4 Mott and Miner, Bell Syst. Tech. Jour., VoL 30, No. 1, p. 110. 1951.
34-6 ACOUSTICAL ENGINEERING
R~
SECTIONAL VIEW
FIG.
"L MO
MAGNETIC NETWORK
than the receiver shown in Fig. 9.3. The frequency range is extended about
700 cycles. With the lower mechanical impedance, the effect of holding the
receiver off the ear does not produce as marked loss in intelligibility as in the
case of the disk diaphragm armature types because the response frequency
characteristic is not altered to any appreciable extent.
B. Crystal Telephone Receiver.-A crystal telephone receiver 5 consists
of a light diaphragm connected to a Rochelle salt crystal (Fig. 9.6). The
three corners of a "bender" crystal are fastened to the case. The fourth
corner is connected to the diaphragm.
The electrical impedance of a crystal is primarily a capacitive electrical
reactance. The electrical network of Fig. 9.6 shows that the low-frequency
response can be raised relative to the high-frequency response by connecting
a high electrical resistance in series with the telephone receivers. A relatively
high electrical resistance must be used because the electrical impedance
of the crystal is relatively high, being 80,000 ohms at 1000 cycles.
S Williams, A. L., Jour. Soc. Mot. Pic. Eng., Vol. 32, No.5, p. 552, 1939.
MISCELLANEOUS TRANSDUCERS 347
The performance of the vibrating system may be obtained from the
mechanical network of Fig. 9.6.
A pressure response frequency characteristic with the telephone receiver
feeding a plain cavity is indicated by B, Fig. 9.3. The pressure response
frequency characteristic taken on an artificial ear is indicated by A, Fig. 9.6.
In dynamic telephone receivers the flux density is relatively low and ZEM
is small compared to rEI and may be neglected.
6 Wente and Thuras, Jour. Acous. Soc. Amer., Vol. III, No. 1, p. 44, 1932.
indicated by B, Fig. 9.7, shows that the response at the low frequencies is
reduced due to the leak.
D. Inductor Telephone Receiver.-An inductor telephone receiver 8 ,9 is
a telephone receiver in which a straight-line conductor, located in a magnetic
field, drives a "V" shaped diaphragm. An acoustical network is used to
compensate the response of the inductor-type telephone receiver shown in
Fig. 9.8. The acoustical network compensates for the leak between the
ear and the earcap. The effect of the leak between the ear and the earcap
upon the response of a telephone receiver has been outlined in the preceding
sections. Obviously, from a practical standpoint the performance of a
8 Olson and Massa, Jour. Acous. Soc. Amer., Vol. 6, No.4. p. 240, 1935.
9 Olson, H. F., Jour. Soc. Mot. Pic. Eng., Vol. 27, No.5, p. 537, 1936.
~C~AI~~~ ~ __ __- ,
P
I rAE
"-v~·,
, '
(~
:---L___ :1._;
: CAE
M_E____...s.__-'-__--I
ACOUSTICAL NETWORK
20~~---.--"-n~nr--'--r,,nnTn
'o" B
~o~HIOO~~~~~~I~OO~O--~~~~~IOOOO
CROSS -SECTIONAL VIEW
FREQUENCY
MICROSCOPE
the groove. Perspective and sectional views, the electrical circuit, and the
mechanical network of a lateral-type magnetic phonograph cutterll are
shown in Fig. 9.10.
A schematic view of the magnetic system and magnetic circuit of the
Z5,-,--,,-'--'-''--'-'--TT-'
~20~4--+~+--+~~-r~~~~
FIG. 9.10. Perspective and sectional views, mechanical network and velocity
response frequency characteristic of a lateral-type phonograph recorder. In the
mechanical network, JM = the mechanical driving force. ml = the mass of the
armature. C Ml = the compliance of the restoring spring. m2, rM, and C M 2 =
the mass, mechanical resistance, and compliance of the damping element. M 3
and C M 3 = the mass of the stylus and holder. ZM = the mechanical impedance
of the load presented to the stylus. In the electrical circuit: ZEM = the motional
electrical impedance. L and rEI = the damped inductance and electrical
resistance of the coil. rEO = the electrical resistance of the generator. e = the
voltage of the electrical generator. The graph depicts the velocity response
frequency characteristic of the recorder.
MAGNET
ARMATURE
COIL
TURNS
_--++-.....:jL.....P....1 1 MI
PIVOT -t--==~-I.o
~
R2 R4
SCHEMATIC VIEW MAGNETIC NETWORK
m rM
f~CM
I T
CIRCUIT DIAGRAM
.. 30
40
MECHANICAL CIRCUIT
V
/~ :-.,
a
~
/ ''(
... 20
(J) V
z
a
...g; 10
~ I\.
0::,
o
/ B
/
-10
"
20 100 1000 10000
FREQUENCY
SECTIONAL VI EW
FIG. 9.12. Sectional view, mechanical circuit, electrical system, and velocity
response frequency characteristic of a feedback lateral-type phonograph cutter.
In the mechanical circuit, 1M = the mechanical driving force. m, rM, and C M =
the mass, mechanical resistance and compliance of the vibrating system. In the
graph, A = the velocity frequency response characteristic without feedback.
B = the velocity response frequency characteristic with feedback.
'"'"'=TIt'"'""" tP
ELECTRICAL SYSTEM
50
., 4 0
E a
w 30
<I)
.
~ 20
/
" ~
<I)
w
10 "
0::
o ,. V 8
8 10 2 2 • 81032
FREQUENCY
11120
//
o ~
,.
l
/
i)
9WID
>
/
/'
./
100 1000 10000
FREQUENCY IN CYCLES PER SECOND
Fig. 9.14 was used up to about two decades ago in recording phonograph
records. To prevent overcutting the groove, the recording is made so that
the amplitude is essentially independent of the frequency below approxi
mately 800 cycles. The velocity under these conditions falls off 6 db per
octave with decrease of the frequency. Above approximately 800 cycles
the recording is made so that the velocity is independent of the frequency.
The amplitude in this frequency range falls off 6 db per octave with increase
of the frequency.
14 Vieth and Wiebusch, Jour. Soc. Mat. Pic. Eng., Vol. 30, No. 1, p. 96, 1938.
356 ACOUSTICAL ENGINEERING
0
2/
5
/:y
,- ., ;..
0
,,' k(.:
en
o
/~ ~
5
>
~
u
0
;::;::; ,,/
/~
o "..
oJ
!oJ _
5 "..
>
.......; V
-I 0
.. "'V
.... ..,. .'
-IS
-2<yo~
~ -;:.;".,'
~ 3
100 rooo 10000 20000
FREQUENCY IN CYCLES PER SECOND
STYLUS
A B
FIG. 9.16. Two systems for heating the cutting stylus of a phonograph re
corder. A. Direct current system. B. Radio frequency system.
the reproduction of 331 and 78 RPM records. The large spindle is used for
the reproduction of 45 RPM records.
Another type of record changer and player plays and changes a single type
of record. One of the most common is the 45 RPM record player and
changer. IS
A record player is the simplest type of disk-record reproducer. It is
manually operated. It ranges from the simplest of all disk-record players
RECORD
PICKUP ARM
".".....,..---++ PICKUP
45 RPM
SPINDLE
g""?}3 SPEED
SELECTOR
t:::======s;;;;======:::;;;;;=-~~~~~~t-START
STOP
LEVER
19 Maxfield and Harrison, Bell Syst. Tech. Jour., Vol. 5, No.3, p. 493, 1926.
MECHANICAL NETWORK
40
III
030
/1
w
'"Z20 .A r ~ 1\/
, \A N
~
'"~IO
IV
900 1000 10000
CROSS -SECTIONAL VIEW fREQUENCY
network of the mechanical system and the velocity of the generator obtained
from Fig. 9.15. The voltage output of the crystal is proportional to the
displacement. The internal electrical impedance of the crystal increases
with the decrease in frequency since the crystal is essentially an electrical
capacitance. The open circuit voltage characteristic renders the low
frequency compensation problem exceedingly simple.
Referring to the mechanical network of Fig. 9.19 it will be seen that the
velocity in the record, ZMR, is a function of the magnitude of the mechanical
impedance of the pickup. As the mechanical impedance of the pickup
becomes larger the vibration velocity of the record will be correspondingly
r
greater. Vibration of the record produces radiation of sound into the air.
CEG
--$-'ZEL
1-1
~~C"3
""3'1,,(: r" 2
Most of this radiation occurs at the high frequencies. The sound produced
in this manner is termed mechanical noise. It is undesirable because it
interferes with the sound from the loudspeaker and produces distortion.
To overcome this, a low-noise crystal pickup21 has been developed. The
essential elements and mechanical network of a low-noise crystal pickup
are shown in Fig. 9.20. A permanent sapphire stylus is used instead of
a replaceable needle. The mechanical impedance of the pickup in shunt
with the mechanical impedance of the record is very small. Therefore, the
motion or vibration of the record due to the pickUp is very small. The
mechanical noise of the low-noise pickup of Fig. 9.20 is about 20 db lower
than the replaceable needle pickup of Fig. 9.19. The open circuit voltage
output is proportional to the amplitude of the crystal. The electrical
generator may be considered to be the open circuit voltage in series with the
electrical capacitance.
21 Burt, A. D., Electyonics. Vol. 16, No. 1, p. 90, 1943.
MISCELLANEOUS TRANSDUCERS 361
FIG. 9.20. Cross-sectional views, mechanical network and electrical circuit of a low
noise crystal pickup. In the mechanical network, ZMR = the mechanical impedance
of the record. ml = the mass of the stylus and holder. CMl = the compliance of
the stylus arm. m2 = the mass of the vertical member. C M 2 = the compliance of
the chuck. m4, rM2, and C M 4 = the mass, mechanical resistance and compliance of
the crystal. CM 5 and rM3 = the compliance and mechanical resistance of the chuck
bearing. C M3 and rMl = the compliance and mechanical resistance of the crystal
support. ma = the mass of the pickup and tone arm. 1M = the force generated by
a velocity generator. In the electrical circuit, eo = the open circuit voltage de
veloped by the crystal. CEO = the electrical capacitance of the crystal. ZEL = the
electrical impedance of the load.
r~ELECTRICAL CIRCUIT
TOP VIEW (CRYSTAL EXPOSED)
MECHANICAL NETWORK
FIG. 9.21. Cross-sectional views, mechanical network, and electrical circuit of a low
noise, wide range crystal pickup. In the mechanical network, ZMR = the mechanical
impedance of the record . ml = the mass of the stylus and holder. C M1 = the com
pliance of the stylus arm. m2 = the mass of the vertical member. CM2 = the
compliance of the chuck. m4, rM2, and CM4 = the mass, mechanical resistance, and
compliance of the crystal. C M3 and r,l(l = the compliance and mechanical resistance of
the crystal support. ma = the mass of the pickup and tone arm. rM5 and C M6 = the
mechanical resistance and compliance of the front bearing. C M 5 and rM3 = compliance
and mechanical resistance of rear bearing. 1M = the force generated by a velocity
generator. In the electrical circuit, eo = the open circuit voltage developed by the
crystal. CEO = the electrical capacitance of the crystal. ZEL = the electrical im
pedance of the load.
362 ACOUSTICAL ENGINEERING
TURN OVER
LEVER
ELECTRICAL CIRCUIT
PERSPECTIVE VIEW
m,
m6 C"6 r M6
FIG. 9.22. Perspective views, mechanical network and electrical circuit of a turnover
ceramic phonograph pickup. In the mechanical network, ZMR = the mechanical im
pedance of the record. ml = the mass of the stylii and holders. m2 = the mass of
the connecting arm. C M2 = the compliance of the turnover arm. rMa and C M 3 = the
mechanical resistance and compliance of the turnover lever arm. C M4 = the com
pliance of the connecting arm. rM5 and C MS = the mechanical resistance and com
pliance of the ceramic support. m6. rM6, and C M6 = the mass, mechanical resistance,
and compliance of the ceramic transducer. m7 = the mass of the pickup and tone arm.
1M = the force generated by a velocity generator. In the electrical circuit, eo = the
open circuit voltage developed by the ceramic transducer. CEO = the electrical capaci
tance of the ceramic transducer. ZEL = the electrical impedance of the load.
arm and angularly displaced by 180°. A stylus with a .003" radius at the
tip is used to reproduce 78 RPM coarse groove records and a stylus with a
.001" radius at the tip used to reproduce 45 RPM and 33! RPM fine groove
records. By means of the turnover lever either stylus may be presented
to the record. The stylus arm is coupled to the ceramic transducer by means
of a cradle lever arm. The performance of the system may be deduced from
the mechanical network of the system. The open circuit voltage output is
proportional to the amplitude of the ceramic transducer. The internal
electrical element of the ceramic transducer is an electrical capacitance.
The electrical generator may be considered to be open circuit voltage in
series with an electrical capacitance.
22 Koren, Pearson, Klingener, and Sabol, Jour. Acous. Soc. Amer., Vol. 26, No. 1,
p. 15, 1954.
r
for the reproduction of 78 RP:!'.1 coarse groove records and the other stylus
r--eG~
cEGT ~
VIBRATING SYSTEM ELECTRICAL CIRCUIT
L I :L 2 Ls:L4
MECHANICAL NETWORK
arm is fitted with a stylus having a tip radius of .001" for the reproduction
of 45 RPM and 33t RPM fine groove records. The lever located at the
front of the cartridge is turned 180 0 to change from one stylus arm to the
other. The stylus arm under operation rests in the cradle of the lever arm
connected to the ceramic transducer. The use of the vibrating system
shown in Fig. 9.23 reduces the mechanical impedance at the stylus. The
lever system provides the proper stylus for the particular application. The
performance of the system may be deduced from the mechanical network.
23 Bauer, Gunter, and Steeler, Jour. A udio Eng. Soc., Vol. 2, No.4, p. 239, 1954.
364 ACOUSTICAL ENGINEERING
9.6
where a = spacing between the armature and pole pieces, in centimeters, and
A = area of the pole piece, in square centimeters.
The incremental change in reluctance is given by
~R=~x 9.9
A
where ~x = change in distance a, in centimeters.
From equations 9.7, 9.8, and 9.9,
~cp = AMMLlx 9.10
a2
24 Kellogg, E. W., Jour. A.I.E.E., Vol. 46, No. 10, p. 1041, 1927.
MAGNET
r
C M2 MI
ARMATURE
S:;~--;-ICOIL
m2
~ ELECTRICAL CIRCUIT
m, rMI
FIG. 9.24. Front and side views, mechanical network, and electrical circuit of a magnetic
pickup. In the mechanical network, ZMR = the mechanical impedance of the record.
GMO = the compliance of the needle. ml = the mass of the needle holder and armature.
GMl = the compliance of the needle holder pivot. C M2 and rMl = the compliance and
mechanical resistance of the damping material. m2 = the mass of the pickup and tone
arm. 1M = the force generated by a velocity generator. In the electrical circuit,
eo = the open circuit voltage developed in the coil. La and rEO = the inductance and
electrical resistance of the coil. ZEL = the el,ectrical impedance of the load.
STYLUS
R2 R,
ARMATURE R, R2
~ I
~
COIL
TURNS
e cp
SPACER
RM
MM MAGNET
MM
RM
e=N
de/>
9.11
dt
where N = number of turns in the coil.
From equations 9.10 and 9.11, the generated voltage in the coil is given by
NMMA.
e=~x 9.12
Equation 9.12 shows that the generated voltage will be independent of the
frequency if the velocity of the armature is independent of the frequency.
The mechanical network of the mechanical system is shown in Fig. 9.24.
Damping, represented by the compliance CM2 and the mechanical resist
ance rMI, is furnished by a suitable material such as viscoloid.
A more recent design of magnetic pickup 26 is shown in Fig. 9.26. The
horizontal stylus arm also serves as the armature. The pole pieces are
located at the stylus. This design makes it possible to obtain a relatively
ARMATURE
DAMPING BLOCK
-- -------~l
I
I
I
I
I
I
I
I
I
FIG. 9.26. Bottom and sectional views, mechanical network, and electrical
circuit of a magnetic pickup. In the mechanical network, 3MB = the mechanical
impedance of the pickup. ml = the mass of the stylus and stylus holder. YMI
and CMl = the mechanical resistance and compliance of the armature. m2 = the
mass of the pickup and tone arm. J.v = the force generated by a velocity
generator. In the electrical circuit, eo = the open circuit voltage developed in
the coil. LG and rEG = the inductance and electrical resistance of the coil.
ZEL = the electrical impedance of the load.
27 Stanton, W.O., Jour. Audio Eng. Soc., Vol. 3, No.2, p. 70, 1955.
rEG LG
[j['IT. - COIL
STYLUS:;.u
SECTIONAL VIEW
C)
ELECTRICAL CIRCUIT
TOP VIEW
PERSPECTIVE VIEWS MECHANICAL NETWORK
FIG. 9.27. Top and bottom perspective views, mechanical network, and the
electrical circuit of a magnetic pickup. In the mechanical circuit, ZMR = the
mechanical impedance of the record. ml = the mass of the stylus, stylus holder,
and armature. rMl and C Ml = the mechanical resistance and compliance of the
armature. rM2 and C M2 = the mechanical resistance and compliance of the
armature longitudinal support wire. rMa and CMa = the mechanical resistance
and compliance of the armature base support. ma = the mass of the pickup and
tone arm. 1M = the force generated by the velocity generator. In the electrical
circuit, ea = the open circuit voltage developed in the coil. La and rEG = the
inductance and electrical resistance of the coil. ZEL = the electrical impedance
of the load.
the mechanical network of Fig. 9.29. Since the system may be made very
small and light, it is possible to reproduce the entire audio-frequency range.
The open circuit voltage is proportional to the velocity of the coil. The
electrical generator may be considered to be the open circuit voltage in
MAGNET
VOICE COIL
FIG. 9.28. Cross-sectional view, mechanical network, and electrical circuit of a vertical
dynamic pickup. In the mechanical network, ZMR = the mechanical impedance of
the record. ml = the mass of the stylus and voice coil. GMl and YMl = the compliance
and mechanical resistance of the suspension system. m2 = the mass of the pickup and
tone arm. 1M = the force generated by a velocity generator. In the electrical circuit,
eo = the open circuit voltage developed in the voice-coil. LG and YEO = the inductance
and electrical resistance of the voice-coil. ZEL = the electrical impedance of the load.
BOTTOM VIEW
MECHANICAL NETWORK
series with the electrical impedance of the coil. The coil is practically a
constant electrical resistance over the audio-frequency range.
Another form of dynamic pickup, shown in Fig. 9.30, is capable of repro
ducing both lateral- and vertical-type phonograph records by merely chang
ing the transformer connections. The vibrating system consists of two
r
{::]J
CONNECTION CONNECTION
rEG LG
TRANSFORMERS
ELECTRICAL CIRCUIT
$ TIC
M, rM
f.------r--;M M
MECHANICAL NETWORK
PERSPECTIVE VIEW
r[Iim,l~:,
MECHANICAL NETWORK
PERSPECTIVE VI EW
that a lateral displacement of the stylus will produce a change in the spacing
between the ribbon and insulated black plate and thus produce a change in
electrical capacitance. The electrical capacitance formed by the ribbon and
insulated back plate is made a part of the oscillating circuit of a 30-megacycle
ELECTRICAL CIRCUIT
ELECTRICAL SYSTEM
FIG. 9.32. Sectional view, electrical system, mechanical network, and electrical circuit
of an electronic pickup. In the mechanical network, ZMR = the mechanical impedance
of the record. m l = the mass of the stylus. YM1 and C Ml = the mechanical resistance
and compliance of the stylus arm. YM4 and C M4 = the mechanical resistance and
compliance of the damping member. m2' YM2' and C M2 = the mass, mechanical
resistance, and compliance of the anode lever. ma, YMa, and C Ma = the mass, mechan
ical resistance, and compliance of the diaphragm. m4 = the mass of the pickup and
tone arm. JM = the force generated by a velocity generator. In the electrical circuit,
eG = the open circuit voltage developed in the tube. YEG = the internal electrical
resistance of the tube. ZEL = the electrical impedance of the load.
r
MECHANICAL NETWORK
SCHEMATIC VIEW
ala Bachman, W. S., Elec. Eng., Vol. 65, No.3, p. 159, 1946.
STYLUS
:t
:oJ 5 ~ .......
ELECTRICAL DIAGRAM
COIL
::E
-- 100
B r--.I.
1000
FREQUENCY
~
10000
FIG. 9.34. Perspective and bottom views, mechanical network, electrical diagram, and
mechanical impedance frequency characteristics of a feedback pickup. In the mechan
ical network, ZMR = the mechanical impedance of the record. ml = the mass of the
stylus and stylus holder. C Ml = the compliance of the stylus arm. m2 = the mass
of the front portion of the armature and ceramic transducer. f'M2 and C M2 = the
mechanical resistance and compliance of the damping block under the stylus arm. f'Ma
and C Ma = the mechanical resistance and compliance of the damping blocks on the
ceramic transducer and magnetic armature. ma = the mass of the rear portion of the
ceramic transducer and magnetic armature. CM4, = the compliance of the ceramic
transducer and magnetic armature. m4, = the mass of the pickup and tone arm.
rM5 and C M5 = the mechanical resistance and compliance of the support for the ceramic
transducer and armature. IMl and/M2 = the forces generated by the velocity generator
and the magnetic driving system. In the graph: A. mechanical impedance characteristic
without feedback. B. mechanical impedance characteristic with feedback.
9.16
TOP VIEW J.
rmym
A-~
mvm
B-~
rmyrn-
C-~
SECTiONAL VIEWS
FIG. 9.36. The shape of the groove cut by the stylus of a lateral
phonograph recorder in a lacquer disk.
walls of the groove is shown in Fig. 9.37. Referring to Figs. 9.36 and 9.37,
it will be seen that, when the reproducing stylus moves in this groove, it
will rise as the groove narrows. The frequency of the rise is twice the
frequency of the modulation. The narrowing of the groove is termed
the pinch effect. The two sides of the groove are symmetrical; therefore, the
stylus must execute symmetrical motion about the center line which means
that there should be no even harmonics. However, odd harmonics are
produced. The equations for the magnitude of the harmonics have been
developed.
With regard to nonlinear distortion, the two-frequency method of distor
tion testing has been found to agree quite well with subjective testing of
nonlinear distortion. 39 The nonlinear distortion, in per cent, which occurs
when two frequencies are combined is given by
36 DiToro. M. J.. Jour. Soc. Mot. Pic. Eng.• Vol. 29. No.5. p. 493.1938.
37 Pierce and Hunt, Jour. Acous. Soc. Amer.• Vol. 10. No. 1, p. 14. 1938.
38 Sepmeyer, L. W., Jour. Acous. Soc. Amer.• Vol. 13. No.3. p. 276.1942.
9.17
41 Max, A. M., Jour. Audio Eng. Soc., Vol. 3, No.2, p. 66, 1955.
42 Bauer, B. B., Jour. Acous. Soc. Amer., Vol. 16, No.4, p. 246, 1945.
43 Bauer, B. B., Trans. I.R.E., Prof. Group Audio, PGA-6, p. 11, 1952.
44 Comerci, Frank A., Jour. Soc. Mot. Pic. Tel. Eng., Vol. 64, No.3, p. 117, 1955.
45 Reid, J. D., Jour. Acous. Soc. Amer., Vol. 13, No.3, p. 274,1942.
and the walls of the record. The force, in dynes, developed by the inter
action of the pickup stylus and the record is given by
9.18
where w = 27rj,
j = frequency, in cycles per second,
x = amplitude of the groove, in centimeters,
ZMR = mechanical impedance of the record, in mechanical ohms, and
ZMP = mechanical impedance of the pickup at the stylus, in mechani
cal ohms.
Equation 9.18 illustrates the importance of a pickup with a small mechani
cal impedance. If the pickup mechanical impedance is comparable to
the mechanical impedance of the record, a considerable part of the amplitude
of the record groove will take place in motion ot the record. This motion
or vibration of the record produces sound which is radiated into the air.
The radiated sound corresponds somewhat to the sound recorded on the
record, but it is very much distorted due to the way in which it is produced
and is, therefore, disagreeable. Furthermore, there is interference between
this sound and the sound radiated from the loudspeaker. The force which
drives the stylus is a function of the record mechanical impedance, if the
mechanical impedance of the pickup at the stylus is relatively large. This
may produce distortion in the reproduced sound because the mechanical
impedance of the record varies over wide limits from the outside to the inside
groove and is a function of the mounting of the record supporting means.
A pickup with a high mechanical impedance also produces excessive record
wear. Equation 9.18 together with the above discussion shows that
record noise and wear and distortion can be reduced by making the
mechanical impedance of the pickup small compared to the mechanical
impedance of the record. The measurement of record noise is described in
Sec.lO.SD.
C. Selection oj Rotational Speed and Record Diameter. 46-The playing
time, the diameters of the start and end grooves of the recording, the rota
tional speed of the grooves per inch, and the terminal linear velocity are all
factors involved in the determination of a record. These relations are inter
connected by the following equations:
p=D02~DtN 9.19
and
9.20
46 Reiskind, H. I., Trans. I.R.E., Prof. Group Audio, PGA-S, February, 1952.
MISCELLANEOUS TRANSDUCERS 379
where P = playing time, in minutes,
Do = diameter of the start groove of the recording, in inches,
D t = diameter of the end groove of the recording, in inches,
R = rotational speed, in revolutions per minute,
N = grooves, per inch, and
St = terminal linear velocity.
Equations 9.19 and 9.20 show that there are many parameters involved
in the selection of the rotational speed and the playing time. One of the
most important of these is nonlinear distortion, discussed in the preceding
section. The other factors such as playing time, record diameter, grooves
per inch, rotational speed, etc. are determined by the particular application.
liNCH - 33.j.RPM
FIG. 9.38. Typical dimensions of the most common commercial-type disk phonograph
records. The dimensions are the diameter of the outside of the record, the diameter of
the outside and inside record groove, the diameter of the label, and the diameter of the
center hole.
maximum playing times are 25, 17, and 8 minutes, respectively. The 45
RPM records are made in a diameter of 7 inches. The nominal maximum
playing time is 8 minutes. The 161 RPM records are made in a diameter
of 7 inches. The nominal maximum playing time of the records with the
large center hole is 30 minutes. The nominal maximum playing time for
the small-hole records is 45 minutes for music and 60 minutes for speech.
The diameter of the outside, the diameters of the first and last program
grooves, the label diameters and the diameter of the center hole of the
different records are shown in Fig. 9.38. It may be mentioned in passing
that the specifications of Fig. 9.38 are given as representative and do not
include all the variations.
The dimensions of the coarse groove, the fine groove, and the ultra-fine
grooves and the dimensions of the corresponding stylii are shown in Fig.
9.37. The coarse groove is used in 78 RPM records. The fine groove is
used in the 45 and 33! RPM records. The ultra-fine groove is used in the
16i RPM records.
The maximum nominal grooves per inch for the different size grooves are
as follows: coarse groove, 125; fine groove, 275; and ultra-fine groove, 550.
The maximum amplitudes, in inches, in the frequency range 200 to 2000
cycles for the different size grooves are as follows: coarse groove, .004-.005
inch; fine groove, .0015-.002 inch; and ultra-fine groove, .0007-.001 inch.
9.4. Vibration Pickup.-Measurement and study of vibration have
become an important factor in the elimination of noise in machinery, vehicles,
and household appliances. Depending upon the requirements, it may be
desirable to measure amplitude, velocity, or acceleration.
Direct measurement of acceleration, velocity, or displacement of vibra
tion requires the establishment of a stationary body to serve as a reference
frame against which these functions may be determined. Any type of
transducer may be used to convert the motion into the corresponding
electrical current. It is the purpose of this section to describe a piezo
electric inertia-type vibration pickup.
The structure of a typical inertia-type piezoelectric vibration pickup 50
is shown in Fig. 9.39. The crystal is a Rochelle salt bimorph type. With
the crystal held at the three corners the output voltage is proportional to
the force acting on the free corner... The crystal is enclosed in a rigid metal
case. When the case is driven by a vibration normal to the plane of the
crystal element, a force is developed at the unsupported section of the
crystal due to inertia reaction. The mechanical network of the vibrating
system is shown in Fig. 9.39. The mechanical resistance is small and does
not influence the mechanical network save near the resonant frequency
which occurs at about 1500 cycles. The velocity, in centimeters per second,
is given by
9.21
50 Bauer, B. B., Jour. Acous. Soc. A mer., Vol. 11, No.3, p. 303, 1940.
MISCELLANEOUS TRANSDUCERS 381
where 1M = driving force developed at the free edge of the crystal, in
dynes,
m = effective mass of the crystal, in grams, and
eM = effective compliance of the crystal, in centimeters per dyne.
For frequencies well below the resonant frequency the velocity is given by
X,-...., !M
Jwm
9.22
The acceleration is given by
'.
JXw= 1M
- 9.23
m
The displacement is given by
x
X= 9.24
JW
Below the resonant frequency the force is proportional to and in phase
with the acceleration. The voltage output of the unit then corresponds
ELECTRICAL SYSTEM
10
SECTIONAL VIEW
[flu -
III
o
SUPPOR~S ...J 0
A-#P'
CRYSTAL
..
flO r. m
w
>
~ -10
V D'
LEADS MECHANICAL
NETWORK 10 100 1000
PERSPECTIVE VIEW ,REQUENCY
FIG. 9.39. Sectional view, perspective view of the crystal mounting arrangement,
mechanical network, electrical connection, and response frequency characteristic
of a vibration pickup. In the mechanical network, m = the mass of the crystal.
YM and eM = the mechanical resistance and compliance of the crystal and
supports. f M = driving force. In the electrical system, with the switch lever
on D, V, and A the response corresponds to displacement, velocity, and accelera
tion, respectively. The graph depicts the voltage response frequency charac
teristics for A, acceleration, V, velocity, D, displacement.
In these devices two different types are used, in one the armature or voice
coil is free and the field structure is coupled to the vibrating system under
test, in the other the armature or voice coil is driven by the vibrating system
under test. The electrical compensation in these devices differs from the
crystal type because the voltage output is proportional to the velocity.
A ceramic vibration pickup 51 similar to the crystal vibration pickup has
been developed. The vibrating system and the electrical characteristics
are similar to those of the crystal vibration pickup. The essential difference
being that the ceramic element is mounted on the four corners and the
acceleration acts upon the center.
An accelerometer 52 has been developed employing a mechano-electronic
transducer similar to that employed in the electronic microphone of Sec.
8.2F and the electronic pickup of Sec. 9.3B2f. A small weight is attached
to the anode rod. The stiffness and restoring force are supplied by the
diaphragm. The system exhibits a high order of sensitivity.
9.5. Sound-Powered Phones.-A sound-powered phone system is a
point-to-point telephone communicating system employing no batteries,
amplifiers or any other means of external power. The sequence of events
in a sound-powered telephone system is as follows: The human voice pro
duces a sound wave which actuates the microphone at the transmitting end.
The microphone converts the acoustical energy into the corresponding
electrical energy. This energy is carried by wires to the receiving end.
At the receiving end the electrical variations are transformed into the cor
responding sound vibrations by the receiver.
A sound-powered telephone is shown in Fig. 9.40. Cross-sectional views,
mechanical and electrical networks of the microphone and receiver are shown
in Fig. 9.40. In order to obtain a tolerable sound level at the receiver the
over-all efficiency of the system must be quite high. This high efficiency
is accomplished by the use of multi-resonant .elements which reduce the
mechanical impedance of the vibrating system. The transmission frequency
band is made relatively narrow so that a low value of mechanical impedance
can be obtained with a simple vibrating system. The response frequency
characteristics of the microphone, the receiver, and the combination of the
microphone and receiver are shown in Fig. 9.41. In the combination system
it will be seen that there is a gain in sound pressure over the useful trans
mission frequency range which means that the sound pressure in the ear
cavity is greater than that at the microphone. The transmission of sound
without pressure loss requires a very efficient system.
9.6. Electrical Megaphone.-The electrical megaphone 53 consists of
the combination of a microphone, an amplifier, and a horn loudspeaker
(Fig. 9.42). The microphone and horn loudspeaker form a single unit.
In use, the operator speaks into the microphone. The voice is reinforced
by the amplifier and loudspeaker. The resulting power output is many
51 Carlson, E. V., Trans. I.R.E., Prof. Group on Audio, PGA-10, p. 2, 1952.
52 Lewis, Robert C., Jour. Acous. Soc. Amer., Vol. 22, No.3, p. 357, 1950.
~
REC. ELEC. CIR. REC.
MECH. NETWORK
\
::l ::>
a.
I
::l
020
r'\ I
520
"'«j" I0
/ \ !oJ
a:
::>
~ 10
.
V
V
\ Q
!;(-20
II
/ \
\ \
!oJ
\
o a: a:
> V
0 o -30
500 1000 2000 3000 500 1000 2000 3000 500 1000 2000 3000
FREQUENCY FREQUENCY FREQUENCY
times that of the unaided voice or the voice and an acoustical megaphone.
The only theoretical limitation to the amount of reinforcing is the produc
tion of continuous oscillations due to regenerative feedback from the loud
speaker to the microphone. The directional pattern of horns shows that
the rear radiation is quite small compared to that directly in front when
the dimensions of the mouth of the horn are comparable to the wavelength
(see Sec. 2.19). By placing the microphone at the rear of the horn and
HORN
LOUD SPEAKER
MECHANISM
~~=L __ "::~
L.::===:::;;;;;;1'-===L-=~~~
MAGNETIC PLASTIC
COATING
BASE
N~\ N
COIL -I""'1I-+-~ RECORDED
e SIGNALS
RI
,...M
Rz
'" R2 ,
R'
R. R.
R' ''';y
• R4
R•
R.
Rs
•
R' e q;
"'5 •
tjd'
R7
"'. "",R.
'!1.
SCHEMATIC V I EW MAGNETIC NETWORK SCHEMATIC VIEW MAGNETIC NETWORK
FIG. 9.46. Schematic view and magnetic net FIG. 9.47. Schematic view and magnetic net
work of the head and tape of a magnetic tape work of the head and tape of a magnetic tap!
recording system. In the magnetic circuit, reproducing system. In the magnetic circuit
M and M' = the magnetomotive forces M = the magnetomotive force stored in th!
developed by the currents in the two coils. tape. R5 and R5' = the reluctances of th!
R5 and R5' = the reluctances of the magnetic magnetic material of the heads. Ra and R~
material of the heads. Ra and R7 = the = the reluctances of the top and bottom ail
reluctances of the top and bottom air gaps. gaps. R4 and R6 = the reluctances of tht
R4 and Rs = the reluctance of the magnetic magnetic leakage paths. R2 and R 2' = th!
leakage path. R2 and R 2' = the reluctances reluctances of the air gaps between the head
of the air gaps between the head and the and the tape. RI = the reluctance of tht
tape. RI = the reluctance of the magnetic magnetic tape. rP5 and rP5' = the flux in the
tape. rP = the flux through the magnetic coils. e and e' = the induced voltages in
tape. i and i' = the currents in the two coils. the coils.
the magnetomotive force. The loss due to the finite length of the gap in
the reproducing head is given by
. 7TdJ
Slll
o
-......
" r--
\
III
o
.z
- -10
.
(/)
~
:J; -15 \
'""
II:
-20
\
/ \ r'\
-25
n11\
II
-30
01 .02 .03 .04.05 .2 .3.4.5 I 2 3 4 5
d/ }.. EFFECTIVE GAP LENGTH ... WAVELENGTH
FIG. 9.48. The response loss due to the reproduction with a finite gap as
a function of the ratio of the gap length to the wavelength.
will be seen that the response is zero when the gap length equals a wavelength
and multiples of the wavelength.
The output of the reproducing head is
d~
e = N dt 9.26
35
'\, ('
30
~
25 v
CD
;!; 20
....
/
Ul
z
a'
'"a:
....
15 /
/
10 /
5
V
o
.01
V .02.03.04.05 .1 .2.3 .4.5 I 2 3 4 5
d/). EFFECTIVE GAP LENGTH - WAVELENGTH
The magnetic material used in the coating of the plastic tape is of necessity
nonlinear because it must pos!?ess retentivity to retain the magnetic signal
applied to the tape in recording. The characteristic which depicts the
magnetomotive force or magnetizing force H produced by the recording head
in the magnetic tape and the residual induction Br after the magnetic tape
leaves the head is depicted by the characteristic 1, 2, 0, 3, 4- of Fig. 9.50.
The nonlinear portion in the vicinity of 0 of Fig. 9.50 will produce distortion.
Various means have been developed for reducing the effects of this non
linear characteristic. The system which is universally used in sound repro
duction today is the alternating current bias. The high-frequency signal
50 to 150 kilocycles is added to the audio signal in the recording head, there
being no modulation of one signal by the other. The action of the addition
of the high-frequency bias in reducing the effect of the nonlinear charac
teristic is shown in Fig. 9.50.
The system used in recording and reproducing from magnetic tape is
shown in Fig. 9.51. The tape transport mechanism consists of the takeup
and payoff reels and the capstan drive. Three heads are used, namely,
erasing, recording, and reproducing heads. The reproducing head is used,
in recording, to monitor the recording. In reproducing the erasing and
MISCELLANEOUS TRANSDUCERS 389
recording heads are immobilized. In recording, any previous signal is
removed by the erase head which is accomplished by feeding a high-fre
quency signal of high intensity to the erase head.
When the recording process is carried out with constant current in the
Br
3/
H
0/ S
{
2/
1/
DRIVE
CApSTAN
REEL REEL
head and the reproducing process is carried out with an amplifier in which
the response is independent of the frequency, the overall response will be
given by the characteristic of Fig. 9.49. Therefore, suitable compensation
must be provided in order to obtain an over-all uniform response frequency
characteristic. The upper frequency limit is determined by the first dip,
where the gap is equal to the wavelength. In recording, suitable high
frequency accentuation is applied in the range above dj).. = .3 so that in
reproduction no additional compensation will be required in this frequency
range. In reproduction, the response is accentuated 6 db per octave with
decrease in frequency in the frequency range below dj).. = .3. In this way
a uniform recording-reproducing characteristic is obtained. The accentua
tion of high-frequency response in recording increases the signal-to-noise
ratio. There is no overload problem in this frequency range due to accentua
tion of response in recording, because the amplitude of speech and music is
lower in the high-frequency range compared to the mid- and low-frequency
ranges. In the frequency range below dj).. = .3 the amplitude of flux if> will
be constant for constant current in the recording head. Therefore, in order
to obviate overloading of the tape in recording, the low-frequency compensa
tion must be supplied in the reproducing amplifier.
The standard tape speeds are as follows: 30, 15, 7~, 31, and 1~ inches per
second. The higher speeds are used for high quality recording. The lower
speeds are used for speech reproduction. The standard tape width is ! inch.
Both single- and double-track recordings are used on the i-inch tape.
The upper frequency limit of reproduction will depend upon the air gap
of the head and the tape speed. With heads in use today, and at tape
speeds of 30 and 15 inches per second an upper frequency limit of 20,000
cycles can be easily achieved. At a tape speed of 71 inches per second an
upper frequency limit of 15,000 cycles per second is possible in a well
designed system. The upper frequency limit employing tape speeds of
31 and 1i inches is correspondingly lower.
9.8. Magnetic Tape Conversion Systems.-A. Frequency Conversion
System.-The fact that magnetic tape can be operated over a tremendous
range of speeds from a fraction of an inch per second to a thousand inches
per second provides a means for frequency conversion. The frequency
conversion may be either to a lower or a higher frequency. The use of a
magnetic tape system for providing frequency conversion is shown in
Fig. 9.52. The tape is recorded with a linear speed V and a maximum
frequency bandwidth of j. The system over which the signal is to be
transmitted is limited to a maximum frequency band of less than j. The
tape is reproduced at a slower speed Vi so that the maximum frequency
corresponds to the capabilities of the transmission system. The output
of the transmission system is fed to a recorder operating at a tape speed
of Vi and the signal is recorded. The record produced in this manner is
reproduced at a tape speed V. In this way the original signal is recovered.
The time of transmission is increased by the ratio of ~.
MISCELLANEOUS TRANSDUCERS 391
The amount of the signal on the original tape that is used is V ~ VI. The
392 ACOUSTICAL ENGINEERING
HEAD CAPSTAN
ROTATING
HEADS
LOUDSPEAKER
output of the rotating heads is fed to a recorder operating with a tape speed
of V - VI. In reproduction, the system of rotating heads rotated in a
direction opposite to the motion of the tape with a peripheral velocity of
VI. The original frequencies are restored.
The advantage of the system resides in the fact that the amount of tape
required for the storage system is reduced and the frequency band required
for transmission is reduced.
In an alternative reproducing system the tape is reproduced at a velocity V
as shown in Fig. 9.54. The time required for the reproduction is now
V - VI
reduced by a factor V
C. Time Compression System.-A time compression system similar to the
frequency compression system may be used for reducing the time of trans
mISSIOn.
The use of a magnetic tape for time compression is shown in Fig. 9.55.
The original signal is recorded at a tape speed V. Then this tape is repro
MISCELLANEOUS TRANSDUCERS 393
LOUDSPEAKER
MICROPHONE
LOUDSPEAKER
FIG. 9.55. The elements and processes in a magnetic
tape time compression system.
to the beam for recording signals derived from sound signals. A sound
motion picture reproducing system is a combination of a light source, optical
system, and photoelectric cell and a mechanism for moving a film carrying
an optical sound record by means of which the recorded variations may be
converted into electrical signals of approximately like form. The elements
of a complete motion picture recording and reproducing system are described
in Sec. 13.7. It is the purpose of this section to describe the film and sound
track, the modulators, and film transport used in the recording of sound on
photographic film and the optical-electronic reproducer and film transport
used in reproduction of sound from photographic film.
A. Film and Sound Track.-In the recording of sound motion pictures the
picture and sound are recorded on separate photographic films. See Sec.
13.7. Therefore, the camera and sound recorder must be synchronized.
This is accomplished by the use of an interlock system between the camera
and sound recorder and the use of perforated film in the form of sprocket
MISCELLANEOUS TRANSDUCERS 395
holes along the two edges of the film for both the camera and sound recorder.
Fig. 9.56.
The sound track on 35-millimeter film occupies a space about .1 inch
wide just inside the sprocket holes as shown in Fig. 9.56. There are two
types of sound track in general use today-namely, variable area and
variable density. The type of sound track shown in Fig. 9.56 is termed
bilateral variable area. There are also other types as, for example, unilateral,
duplex, class A pushpull, and class B pushpull variable area sound tracks.
SOUND SPROCKET SOUND
TRACK HOLE:S TRACK
0
ICYV'
0 0 o
0 PICTURE:
0 o PICTURE:
0 ARE:A
0 o
AREA
0 tj 0 Oc--_ _ _ _--'
0 0 o
VARIABLE AREA VARIABLE DENSITY
FIG. 9.56. The position of the picture and sound track in 35-millimeter sound
motion-picture film. Two types of sound track are shown-namely, variable area
and variable density.
OUTLINE or
MECHANICAL TRIANGULAR
S~I:'J
.-*.-
GALVANOMETER LIGHT
o
APERTURE
RECORDING SYSTEM
T" L
MECHANICAL
SLIT
rLn}-m2
ARMATURE~ r..c.. z
MAGNET
ELECTRICAL
CIRCUIT
MECHANICAL
NETWORK
OAMPER
FIG. 9.57. The elements of a variable area sound motion-picture film recording
system. The negative and positive sound tracks. Perspective and sectional views,
mechanical network, and electrical circuit of the galvanometer. In the mechanical
network, JM = the mechanical driving force. ml and C MI = the mass and com
pliance of the armature. m2, rM, and C,1f2 = the mass, mechanical resistance, and
compliance of the damper. In the electrical circuit, ZEM = the electrical motional
impedance. L and rEI = the damped inductance and electrical resistance. YEG
= the electrical resistance of the generator. e = the voltage of the generator. The
graph depicts the amplitude response frequency characteristic of the galvanometer.
Dotted and solid lines depict the amplitude response for the galvanometer alone and
with an electrical capacitance in shunt with the galvanometer, respectively.
71 Dimmick, G. L., Jour. Soc. Mot. Pic. Eng., Vol. 29, No.3, p. 258, 1937.
72 Kellogg, E. W., Jour. Soc. Mot. Pic. Eng., Vol. 36, No.2, p. 137, 1941.
74 Silent and Frayne, Jour. Soc. Mot. Pic. Eng., Vol. 18, No.5, p. SSt, 1932.
75 Scoville and Bell, Jour. Soc. Mot. Pic. Eng., Vol. 38, No.2, p. 125, 1942.
76 Kellogg, E. W., Jour. Soc. Mot. Pic. Eng., Vol. 15, No.5, p. 653, 1930.
77 Collins, M. E., Jour. Soc. Mot. Pic. Eng., Vol. 48, No.2, p. 148, 1947.
E:S i" C MT
llt~::~
ELECTRICAL MECHANICAL
CIRCUIT CIRCUIT
30
m
rM co2 5 .A
~
C M <>
..,20
LENS BRIDGE
VI
~ 15
J "
ASSEMBLY
PEG
RIBBON
0
::: 10
.....V
ASSEMBLY a:
_-,==,--",/,---FlLM
o
200 1000 10000
RECORDING SYSTEM FREQUENCY
TO TO
PAYOFF TAKEUP
REEL REEL
RECORDING DRUM
i 1IIIIIImi
OUTPUT
II
'~~~''''II~ :~I
PHOTO- 20
SLIT
4 810~ 2 4 8 103 2 4 '104 2
TRANSFORMER 2
FREQUENCY
REPRODUCING SYSTEM
FIG. 9.60. The elements of a motion picture film sound reproducing system and
the voltage response frequency characteristic with a constant amplitude film.
usually about .00075 inch. Under these conditions the amount of light
which impinges upon the photocell is proportional to the unexposed portion
of the sound track in variable area recording or to the inverse function of
the density in variable density recording. When the film is in motion the
light undulations which fall upon the photocell correspond to the voltage
variations applied to the recording galvanometer. The voltage output of the
photocell is proportional to the amount of light which falls upon the cathode.
The voltage output response frequency characteristic of a typical motion
picture film sound reproducing system using a constant amplitude film is
shown in Fig. 9.60. The falling off in response at the high-frequency portion
of the range is due to the finite dimensions of the slits in the recording and
reproducing systems. This reduction in response can be overcome by
compensations in the recording and reproducing systems.
400 ACOuSTICAL ENGINEERING
SPROCKET SPROCKET
HOLE HOLE
\
-...J ~
0 0 ~
........
0 0 0 0
0 0 0 PICTURE 0
AREA
0 0 0 0
0 0 0 0
In lr->
A
MAGNETIC
COATING B ~:AGNETIC TRACK/
.030 INCH WIDE
MAGNETIC TRACKS
.050 INCH WIDE
FIG. 9.62. A. Magnetic tape for original sound recording in sound motion
pictures. B. Magnetic strips on a motion-picture positive film.
that if desired the existing photographic recorder may be used by the addi
tion of magnetic heads. Special magnetic recorders have also been de
veloped, but these also use magnetic tape of the same dimensions. Another
reason for the use of the wide tape is that several tracks representing several
channels may be recorded on the tape shown in Fig. 9.62A because the total
width of the magnetic coating is approximately an inch. For example, a
soloist may be recorded on one track and the sections of the orchestra may be
recorded on other tracks. The three or more channels of stereophonic sound
may be recorded on the magnetic tape. See Sec. 13.7.
The information recorded on the magnetic tape is rerecorded on the
photographic film by means of the systems described in Sec. 9.8.
During the past three years wide screen motion picture systems 80 with
stereophonic sound have been introduced on a wide scale. The elements
80 Sponable, Brazz, and Grignon, Jour. Soc. Mot. Pic. Tel. Eng., Vol. 63, No. 1, p. I,
1954.
402 ACOUSTICAL ENGINEERING
of the system are described in Sec. 13.7. The three or four channels are
recorded on the tape shown in Fig. 9.62A. The information is rerecorded
on the magnetic strips cemented to the positive picture release print as
shown in Fig. 9.62B.
B. Recording Tape Transport.8l.82-The tape transport used in recording
sound on magnetic tape for sound motion pictures consists of a positive
sprocket drive of the perforated film and a constant speed drive where the
magnetic recording head is in contact with the tape. A magnetic tape
transport mechanism of this type is shown in Fig. 9.63. Positive drive of
the tape is obtained by means of the sprocket drive. The sprocket drive is
interlocked with the camera drive so that synchronism of the picture and
TO
PAYOFF
REEL RECORD
HEAD
MAGNETIC
TAPE
MONITOR
TA~~UP HEAD
REEL
sound will be obtained. See Sec. 13.7. When the tape passes over the
sprocket drive, variations in the motion of the film at the sprocket hole
frequency are produced. These variations in the film speed must be removed
at the recording point to eliminate spurious frequency modulation of the
magnetic recording on the magnetic tape. Uniform speed at the recording
point is provided by the filter between the sprocket drive and the recording
point consisting of the inertia elements of the two rollers and the two drums
and the compliance of the tape between the sprocket drive and the rollers
and the drums. The drums are equipped with flywheels to provide additional
inertia. Damping of the inertia and compliance system is provided by the
mechanical resistance of the dash pot. The spring system maintains a tight
loop for the magnetic tape. This design of mechanical system provides
uniform motion at the magnetic heads. Two magnetic heads are provided,
one for recording and the other for monitoring.
The film recorder of Fig. 9.59 may also be used for recording one track of
magnetic tape by placing a magnetic recording head inside the drum in
81 Hittle, C. E., Jour . Soc. Mot. Pic. Eng., Vol. 58, No.4. p. 323. 1952.
82 Davis and Manley. Jour. Soc. Mot. Pic. Tel. Eng.• Vol. 62. No.3. p. 208. 1954.
the filter system. Damping of the FIG. 9.64. Schematic view of the film
loop tension rollers provides the transport mechanism of a magnetic re
mechanical resistance element. The producer for sound motion pictures.
filter system removes the variations
introduced by the sprocket drive and provides uniform motion of the film at
the magnetic heads. .
9.11. Volume Limiters, Compressors, and Expanders. 84 ,85,86,87
A volume compressor is a system that reduces the amplification of an
amplifier when the signal being amplified is large and increases the amplifi
cation when the signal is small. Compressors are used to reduce the volume
range in sound motion picture and phonograph recording, sound broad
casting, public address, and sound reinforcing systems, etc.
A volume expander is a system that increases the amplification of an
amplifier when the signal is large and decreases the amplification when the
signal is small. In reproduction, a volume expander is used to counteract
the effect of the compressor in recording.
Volume compressors and expanders are amplifiers in which the amplifi
cation varies as a function of the general level of the signal. The elements
of a compressor, limiter, or expander are shown in Fig. 9.6SA. The input
83 Phyffe and Hittle, Jour. Soc. Mot. Pic. Eng., Vol. 62, No.3, p. 215, 1954.
85 Norman, N. C., Bell Labs. Record, Vol. 13, No.4, p. 98, 1934.
86 Mathes and Wright, Bell Syst. Tech. Jour., Vol. 13, No.3, p. 315, 1934.
87 Steinberg, J. C., Jour. Acous. Soc. Amer., Vol. 13, No.2, p. 107, 1941.
i[2Jil2]
INPUT IN DB
LIMITER
INPUT IN DB
COMPRESSOR
il21 INPUT IN DB
EXPANDER
A B
FIG. 9.65. Input vs. output characteristics of limiters. compressors. and expanders.
A. The elements of a limiter or a compressor. B. The elements of an expander.
both the reverberation time of the chambers and the coupling between the
chambers. The loudspeakers, microphones, and amplifiers used for these
systems should be of the highest quality. Mixers are provided so that any
ratio of the original sound to reverberant sound may be obtained.
Reverberation, in the chamber described above, consists of the multiple
reflection of a large number of pencils of sound. Each pencil of sound
suffers a decrease in intensity with each reflection. These conditions can be
simulated by the system shown in Fig. 9.66B. The amplified sound signal
is passed through a number of transducers with progressive delay and attenu
ation. These transducers may be a series of pipes with loudspeakers and
microphones terminating the ends. The transducer may be a recorder and
a series of pickups on a phonograph record or magnetic tape 88 or phosphor
wheel. 89 The reverberation time may be varied by varying the progressive
88 Wolf, S. K.. Proc. I.R.E., Vol. 27. No.7, p. 365. 1939.
89 Goldmark and Hendricks. Proc. I.R.E .• Vol. 27. No. 12. p. 747. 1939.
attenuation. Mixers are provided so that any ratio of the original sound
to the reverberant sound may be obtained.
9.13. Hearing Aids.-Test made upon representative cross-sections
of the people in this country show a very large percentage to be hard of
hearing. Practically all of these people may obtain satisfaction from the
use of a hearing aid. A hearing aid is a complete reproducing system which
increases the sound pressure over that normally received by the ear.
The first and simplest hearing aid 90 consisted of a carbon microphone, a
battery, an attenuator, and a telephone receiver (Fig. 9.67A). This hearing
aid gave satisfactory service where the hearing loss was about 20 db.
The hearing aid shown in Fig. 9.67B consisted of a carbon microphone, a
carbon amplifier, an attenuator, a battery, and a telephone receiver. This
hearing aid gave satisfactory service where the hearing loss was about 40 db.
CARBON
I.UCROPHONE ATTENUATOR
CARBON I
MICROPHONE
ATTENUATOR
CARBON \
AMPLIFIER,I MICROPHONE
,------,
01., ""~m~
AMPLIFIER
WITH VOLUME AND
TONE CONTROL
RECEIVER
BATTERY
A B c
FIG. 9.67. Hearing aids. A. Simple carbon microphone hearing aid. B. Carbon
microphone hearing aid with a mechanical carbon amplifier. C. Schematic diagram
of a vacuum tube or transistor hearing aid.
The quality of the carbon type hearing aids was usually very poor, due to
the frequency and the amplitude distortion produced by the carbon micro
phone and amplifier.
During the past two decades and prior to the advent of the transistor,
hearing aids 91 ,92,93,94,95,96,97,98 employing vacuum tube amplifiers almost
completely replaced the carbon types. This has been due to the develop
ment of small low-current drain vacuum tubes and small high-efficiency
batteries. The quality is far superior to that of the carbon type. Further
more, suitable compensation circuits may be introduced to complement the
ear characteristics. The schematic arrangement of the components of a
vacuum tube hearing aid is shown in Fig. 9.67C. The microphone used in
hearing aids today is a diaphragm crystal or ceramic type similar to that
described in Sec. 8.2C2 and depicted in Fig. 8.9.
90 TufInell, W. L., Bell Labs. Record, Vol. 18, No. 1. p . 8, 1939.
91 Ramanow, F. F., Jour. Acous. Soc. Amer., Vol. 13, No.3, p. 295, 1942.
92 Sabine, P. E., Jour. Acous. Soc. Amer., Vol. 16, No. 1. p. 38, 1944.
93 Carlisle and Mundel, Jour. Acous. Soc. Amer., Vol. 16, No. 1, p. 45,1944.
94 Grossman and Molloy, Jour. Acous. Soc. Amer., Vol. 16, No. 1. p. 52, 1944.
95 Hanson, W. W., Jour. Acous. Soc. Amer., Vol. 16, No. 1, p. 60, 1944.
98 LeBel, C. j., Jour. Acous. Soc. Amer., Vol. 16, No. 1. p. 63,1944.
97 Watson, N. A., Jour. Acous. Soc. Amer., Vol. 16, No.3, p. 194, 1945.
98 Strommen, E., Jour. Acous. Soc. Amer., Vol. 15, No.4, p. 211, 1944.
c::::=s
rEG L rEI
'''~
~ 15
COIL z
:r 10
C M• MAGNET '"
~ 5
o
m,rMI C M2 eM' 100 1000 10000
fREQUENCY
CROSS-SECTIONAL VIEW
~c~
~
m,
;M ~---1§-r.-"2~~
ELECTRICAL CIRCUIT MECHANICAL NE TWORK
30
'o" 20
~ 10
'\
Z
lr 0
Vr-...
'"
~ -10 - ./ \
V
SECTIONAL VIEW
-20
100 1000
1\ 10000
fREQUENCY
fi =:;,,, "'"
RECEIVER
103 Jones, Clark. Jour. Acous. Soc. Amer., Vol. 18. No.2, p. 371. 1946.
104 Silverman, Daniel.. Jour. A.I.E.E., Vol. 58, No. 11. 455,1939.
106 Frederick and Dodge, Bell Syst. Tech. Jour., Vol. 3, No.4, p. 531. 1924.
SYSTOLIC SOUNDS
DIASTOLIC SOUNDS
STOLIC MURMURS
PRESTOLIC MURMLIRS
PERISTOLIC SOUNDS
RESPIRATORY SOUNDSlRANDOMj
BINAURAL
TAPERED TUBES
c
FIG. 9.72. Sectional views of stethoscopes. A. Diaphragm type. B. Open bell
type. C. Wide range selective type.
delivers greater output in the frequency range from 250 to 1500 cycles. The
open bell has better low-frequency response but the general output level in
the mid-frequency range is lower than the diaphragm type.
There are two reasons for the use of a diaphragm instead of an open
bell-namely, to exclude or attenuate external noises, and to eliminate
leakage between the body and the stethoscope. The open bell stethoscope
actually amplifies air-borne noises in the manner of the ear trumpet. If
the effective slit between the body and the bell of the open bell stethoscope
is just a small fraction of a thousandth of an inch, the low-frequency response
is attenuated due to this leakage. If the bell is pressed against the body so
this leak is effectively eliminated, the body stiffness represented in the
acoustic impedance of the body is increased with a resultant attenuation of
low frequencies.
MISCELLANEOUS TRANSDUCERS 413
In the existing diaphragm type stethoscopes the investigators have
found that it is necessary to use a resonant diaphragm in order to obtain
good output They have placed these resonances in the mid-frequency
range where the ear is quite sensitive. As a consequence the stiffness of
the diaphragm is quite high and the result is very high attenuation of the
low-frequency response.
A wide-range acoustical stethoscope108 ,109 is shown in Fig. 9.72C. The
chest piece of radical design consists of a light polythene diaphragm sup
ported by a multipyramid resilient back plate. This structure provides
an efficient coupling means to the high acoustical impedance of the body.
The adequate resilience of the chest piece insures uniform response to low
tones. The light-weight diaphragm coupled directly to the body makes it
possible to obtain output beyond 4000 cycles. The acoustical impedance
of the chest piece is matched to the acoustical impedance of the tube or
line at the input end. The relatively high acoustical impedance at the
input end of the line is matched to the relatively low acoustical impedance
of the ear by the use of a tapered tube or line. The sensitivities in the low
and high-frequency ranges are much greater than those of existing stetho
scopes due to the matching of acoustical impedances. The high-frequency
response is maintained to 4000 cycles while most existing stethoscopes cut
off at 1500 cycles. There are certain instances in which the entire frequency
range is not desired. This is particularly true when the particular sounds in
question are confined to the low-, high-, or mid-frequency range. For
example, in listening to high-frequency prestolic murmurs, peristaltic, and
respiratory sounds, it may be desirable to eliminate the low frequencies.
In other instances, it may be desirable to attenuate the high-frequency
range. Therefore, to increase the usefulness of the stethoscope, an acoustical
filter has been added in which it is possible to attenuate either the low- or
high-frequency ranges, or both. The acoustical filter provides a system in
which frequency discrimination may be introduced at will, and thereby
increases the usefulness of the stethoscope by classification of the charac
teristic sounds in the body into frequency bands.
The electrical stethoscope consists of the combination of a microphone,
amplifier, and telephone receivers. In one type the pickup device consists
of a bell-shaped horn, coupled to the microphone diaphragm. The coupling
system is similar to that of Fig. 8.1. Condenser, magnetic, and crystal
type transducers have been used in the microphone for these applications.
The amplifier is equipped with low- and high-frequency tone controls for
attenuating the response in either or both the high- and low-frequency
ranges. The addition of a recording system similar to the electrocardio
graph may be used to obtain an oscillographic record depicting the sounds in
the body. Since the output of the electrical stethoscope is greater than that
of the acoustical stethoscope, noises generated by the clothing, movement
of the headpiece, etc., cause considerably more interference than in the
108 Olson, H. F., Electronics, Vol. 16, No.8, p. 184. 1943.
acoustical stethoscope. This is due to the fact that most of these noises in
the acoustical stethoscope fall below the threshold of hearing.
9.17. Ear Defenders. llO-Ear defender is a term used to designate a
device which introduces attenuation of sound between a point outside the
head and the eardrum. There are two types-namely, the cushion type
and the insert type. The cushion type is similar to a pair of headphones
with soft cushion ear pads. The cushion type is heavy, cumbersome, and
uncomfortable and for that reason it has not been used to any appreciable
extent. The insert type is some form of plug which is pushed into the ear
~
PLASTIC WA~PL:~~IC
.-;" •• ..f.. ~',~
·~;":i~r.~~:;~~?}
A B
SECTIONAL VIEWS
ACOUSTICAL NETWORK
FIG. 9.73. A. Sectional view of an ear defender in the ear canal and the
acoustical network of the system. In the acoustical network, M 1 = the
inertance due to the mass of the ear defender. CAl and r Al = the effective
acoustical capacitance and acoustical resistance of the ear defender with
respect to the wall of the ear canal. M 2 and r A2 = the inertance and
acoustical resistance of the leak between the ear defender and the wall of
the ear canal. C A2 = the acoustical capacitance of the entrapped volume
of the ear canal. PI = the sound pressure outside the ear. P2 = the
sound pressure in the ear canal. The separate sectional views show two
different designs of ear defenders with one and two sealing flanges.
respectively. B. Wax filled ear defender.
canal. One form, which was used extensively a number of years ago,
consisted of a wad of cotton. The attenuation of a wad of cotton decreases
with decrease of the frequency. The attenuation below 500 cycles is quite
small. In order to obtain high attenuation at the low frequencies, the seal
between the defender and the ear canal must be practically airtight, because
a very minute hole will reduce the attenuation to a negligible amount. This
fact can be deduced from a consideration of the acoustic network of the ear
defender of Fig. 9.73A. A successful insert type of ear defender must be
made of suitable material combined with a shape which will provide
adequate attenuation, comfort, easy insertion, and positive retention.
110 Watson and Knudsen. Jour. Acous Soc. Amer.• Vol. 15, No.3. p. 153. 1944.
MISCELLANEOUS TRANSDUCERS 415
Ear defenders have been developed which satisfy the above requirements.
The most successful ear defenders have been made of synthetic rubbers or
soft plastics, because these materials remain resilient over long periods of
time and are resistant to ear wax. The shape which appears to be most
successful is a skirt closed at the top and equipped with one or more thin
flounces which rest against the ear canal and thereby provide the seal
(Fig.9.73A.). A tab, fastened at the bottom of the skirt, is used for inserting
or removing the defender. A good ear defender will introduce an attenua
tion of between 30 to 35 db over the frequency range from 60 to 8000 cycles.
Another design ll1 of the ear defender employs a plastic case with soft
elastic walls and a viscous core of malleable wax. Fig. 9.73B. This design
is more comfortable because the body heat softens the wax with the result
that the defender corresponds to the ear canal in which it is placed without
distortion of the ear canal. Such distortion leads to discomfort.
9.18. Electronic Sound and Vibration Reducers and Absorbers.
Existing systems for the absorption of sound and the control of vibrations
are all of the passive type. Recently, active systems have been developed
for the control of sound, reverberation, and vibration. These systems are
in the form of combinations of electronic elements. It is the purpose of this
section to describe electronic sound and vibration reducers and absorbers.
A. Free-Field Zone-Type Sound Reducer.1 12-The free-field, zone-type
sound reducer consists of a microphone, amplifier, and loudspeaker con
nected in inverse fashion so as to reduce the sound pressure of any incident
sound wave in the vicinity of the microphone-loudspeaker combination.
A sectional view, schematic electrical diagram, and acoustical circuit of an
electronic sound reducer are shown in Fig. 9.74. The system is connected
and equalized for response and phase with respect to frequency, so that the
sound pressure is reduced at the microphone. The driving pressure P2,
of Fig. 9.74, is given by
9.28
The amplitude and phase relations of the sound pressures PI and P2 are
selected so as to make the sound pressure pa as small as possible. Under
these conditions the operation of the system is a sound pressure reducer.
The amount of sound pressure reduction is a function of the distance between
the microphone and loudspeaker, the wavelength of the sound wave, the
phase relation in the electronic system, and the distance from the
microphone-loudspeaker combination. Typical sound reduction frequency
111 Zwilocki, J., Jour. Acous. Soc. Amer., Vol. 27, No.3, p. 460,1955.
1120lson and May, Jour. Acous. Soc. Amer., Vol. 25, No.6, p. 1130, 1953.
416 ACOUSTICAL ENGINEERING
ACOUSTICAL NETWORK
SECTIONAL VIEW
FIG. 9.74. Sectional view. schematic electrical diagram. and acoustical network of an
electronic sound absorber. PI = the sound pressure in free space. MI = the inertance
of the air load. rAI = the acoustical resistance of the air load. M2 = the inertance
of the cone and voice coil of the loudspeaker. CAl = the acoustical capacitance of
the suspension system of the cone. r A2 = the acoustical resistance of the cloth over the
apertures in the back plate. CA2 the acoustical capacitance of the volume of the
cabinet. r A3 = the acoustical resistance of the sound absorbing material in the cabinet.
P2 = the driving sound pressure in the loudspeaker. Pa = the sound pressure at the
microphone.
characteristics for various distances from the reducer are shown in Fig. 9.75.
These characteristics show that the electronic sound reducer may be used to
reduce undesired sounds over a limited zone of operation.
One application for the electronic noise reducer is in the form of a noise
reducer in airplanes and automobiles where the noise level is relatively high
in the low-frequency range. With the practical use of conventional sound
absorbing materials in automobiles and airplanes, the reduction in noise
level in this frequency region is relatively small. For these applications, the
noise reducer may be installed on the back of the seat. There are also
many other applications for the zone-type noise reducer. Other applica..
tions include the reduction in noise from machines, ducts, etc.
R = 24"
/------ ..........
>' "
/ "
// R = 10" '\
~
I R 4" \ \
I
(
I
/
I
I
(-~,
'!;
\
...,-IM I
\ ILl
II:
:::>
~LSI ~-5~--4-----~~-+--~~~L-~~
I
\
\
\
\
\
\
/ 'I / I
ILl
II:
\
, " _",/
/
I
"'- / /
' ......... -- ,/'
SCHEMATIC VIEW
FREQUENCY IN CYCLES PER SECOND
ACOUSTICAL NETWORK
SECTIONAL VIEW
FIG. 9.76. Sectional view, schematic electrical diagram and acoustical network of an
electronic sound absorber. PI = the sound pressure in free space. Ml = the inertance
of the air load. rAI = the acoustical resistance of the air load. M2 = the inertance
of the cone and voice coil of the loudspeaker. r AS = the acoustical resistance of the
screen covering the microphone and cone. Ms = the inertance of the screen. CAl
= the acoustical capacitance of the suspension system of the cone. r A2 = the acoustical
resistance of cloth over the apertures in the back plate. C A2 = the acoustical capacitance
of the volume of the cabinet. P2 = the driving sound pressure in the loudspeaker.
Pa = the sound pressure at the microphone. YAa = the acoustical resistance of the
sound absorbing material in the cabinet.
RECEIVER
DRIVER
IAPHRAGM
LEVER
SYSTEM
r
ln, x,
VOICE COIL
MAGNET -1'4--001
SENSOR MECHANICAL NETWORK
SCHEMATIC VIEW
FIG. 9.78. Sectional view, schematic electrical diagram, and mechanical network of an
electronic vibration reducer. Xl = the input vibrational velocity. IM1 = the input
driving force. m 1 = the mass of the lever system and voice coil. 1M2 = the driving
force generated in the voice coil. eM = the compliance of the lever system and the
centering spider. i2 = the vibrational velocity of the voice coil. m2 = the mass of
the magnetic structure. ZMS = the mechanical impedance of the support. 1M3 = the
driving force at the support. i3 = the vibrational velocity of the support.
control and isolate the vibrations. In most problems involving the control
of vibrations, the amplitudes are relatively small and the mechanical im
pedance relatively large. Under these conditions, piezoelectric transducers
may be used. For example, an electronic vibration reducer may consist
of a piezoelectric driver and sensor with a suitable amplifier.
A dynamic system may also be used as a vibration reducer. Since the
amplitude of the vibrations produced by most machines is small compared
MISCELLANEOUS TRANSDUCERS 419
to the amplitude which may be obtained from a dynamic unit, a transformer
in the form of a lever may be used between the dynamic unit and the machine.
A schematic view of a dynamic vibration reducer is shown in Fig. 9.78. The
sensor is a piezoelectric transducer. The amplifier is similar to the amplifier
used for the electronic noise reducer. The performance of the system may
be obtained from the mechanical network. The mechanical impedances in
the mechanical network are all referred to the input and output terminal
impedances. Cognizance must be taken of the lever system, which is in
effect a mechanical transformer, in referring the force 1M2 and the masses
and compliances to the input and terminal mechanical impedances. The
problem in isolating the machine from the supports is to adjust the phase
and magnitude of the force 1M2 so that the resultant force 1M3 developed in
the support will be a minimum. The velocity X3 in the support may be
expressed as follows:
. 1MIZM2 - 1M2ZMI
X3 = -----''---,---"''------,---- 9.29
ZMIZM2 + ZMIZM3 + ZM2ZM3
where ZMl = mechanical impedance of the machine,
ZM2 = mechanical impedance of the driver and sensor, and
ZM3 = mechanical impedance of the support.
It may be that a part of the mechanical impedance of the driver and
sensor may be included in the mechanical impedance of the machine and/or
the mechanical impedance of the support.
From a consideration of equation 9.29, it will be seen that the magnitude
of the velocity X3 can be reduced by the application of the force 1M2 in the
proper magnitude and phase. For example, X3 = 0 when
1MIZM2 = 1M2ZMI 9.30
Under these conditions,1M3 is also zero. That is to say, no vibrations are
produced in the support. The machine is perfectly isolated from the
support.
Another problem is to reduce the vibration of the machine without regard
to the vibration transmitted to the support. The velocity Xl of the machine
may be expressed as follows:
.
Xl =
1MI (ZM2 + ZM3)
"---'------'----''----
- 1M2ZM3
9.31
ZMIZM2 + ZMIZM3 + ZM2ZM3
From a consideration of equation 9.31, it will be seen that the magnitude
of the velocity Xl can be reduced by the application of the force 1M2 in the
proper magnitude and phase. For example, the velocity of the machine
Xl will be zero if
9.32
If Xl = 0, there will be no motion of the machine. There will, however,
under these conditions be a larger velocity X3.
There are many applications for an electronic vibration reducer which
decreases the coupling between an offending vibration producer and a
420 ACOUSTICAL ENGINEERING
115 Olson, H. F., Electronics, Vol. 20, No. 12, p. 119, 1947.
I
~ 0
...rn~ -5 "\['\;\
INPUT
z
oQ.
gJ -10 1,\\
\
u:
-15
100 200 400 1000 2000 4000 10000
FREQUENCY
filters which pass frequencies over a range of an octave are used at the input
and output of the nonlinear element. The amplitude characteristic of the
nonlinear element is illustrated in Fig. 9.80. As will be described later,
this amplitude characteristic can be obtained with a properly biased diode
vacuum tube or crystal rectifier. By using this method the system will
exhibit high attenuation to signals of small amplitudes.
NOISE INPUT
NOISE OUTPUT
/\/\/\
VV\J SINE WAVE INPUT
f\ C\ !\
v V V
/\/\/\
V\JV
SINE WAVE OUTPUT
then, is a sine wave of the same frequency as the input sine wave. If two
sine waves of different frequencies are impressed upon the system, the two
frequencies must lie within the pass band octave in order to be admitted by
the input band-pass filter. The output of the nonlinear element contains
harmonics and subharmonics of the two fundamental frequencies, but these
are rejected by the output band-pass filter. The output of the nonlinear
element also contains the sum of the two frequencies and the difference of
the two frequencies . Since the input is confined to an octave, the band-pass
output filter will reject the sum and difference frequencies.
A system with an upper cutoff of 12,000 cycles and three channels of
noise reduction is depicted in the block diagram of Fig. 9.81. This system
INPUT
l":
1. ;. "'!1\ \
0
~
...~ -5 \
lr
"'
~-IO
1!
I\/\ \
-1540 100 200 400
l 1\ ~ '. \
1000 2000 4000 10KC 20KC
FREQUENCY IN CYCLES PER SECOND
uses the nonlinear elements of Fig. 9.80. Each nonlinear system is equipped
with a separate bias control so that the noise can be reduced in each band
without discrimination against the useful signal. The response-frequency
characteristic of the separate channels and the over-all response is shown in
Fig. 9.81. Conventional band-pass filters are used to confine the response
to octave bands. An amplifier overcomes the loss in the filters and nonlinear
elements. Noise reduction of up to 20 db can be obtained in each of the
channels.
10
MEASUREMENTS
7 Kaye, G. W. C., Jour. Acous. Soc. Amer., Vol. 7, No.3, p. 174, 1936.
8 Glover and Baurnzweiger, Jour. Acous. Soc. Amer., Vol. 10, No.3, p. 200, 1939.
MEASUREMENTS 425
b. Thermophone. 9,10,l1-The thermophone consists of one or more strips
of thin gold leaf mounted upon terminal blocks (Fig. 10.IB). In the usual
method the thermophone strip carries a known steady current upon which
a smaller sinusoidal current is superimposed. In this case, the variation of
the pressure in the chamber occurs primarily at the frequency of the alter
nating current. The cavity of the thermophone is usually filled with
hydrogen. The wavelength in hydrogen is considerably longer than in air
and, as a consequence, the standing waves are shifted to a higher frequency
beyond the useful response range.
The peak alternating pressure developed in the cavity is given by
P= . 96SioirE
wmCV ActDl/2 10.2
where D =
4KS2
( 1 - wCV A
)2 + (1 + V45Act + 4KSa 4KS2 )2
wC + wCV A
A = Tct - Y - - l
Tsy -1
m = .:..:..(y_---:-l'-)T...=.s
yPO
- JwCpp
a - 2K
C = total thermal capacity of the strip product of the mass in grams
and the specific heat,
io = steady current, in amperes,
i = peak value of the alternating current component, in amperes,
rE = total electrical resistance of the strip, in ohms,
Ts = mean temperature of the strip, in degrees Kelvin,
Ta = mean temperature of the gas in the enclosure, in degrees Kelvin,
K = thermal conductivity of the gas,
p = density of the gas, in grams per cubic centimeter,
y= CplC v ,
po = average pressure of the enclosure, in dynes per square centi
meter,
5 = total area of one side of the thermo phone foil, in square centi
meters,
w = 2"1Tj, and
11 Ballantine, S., Jour. Acous. Soc., Amer., Vol. III, No.3, p. 319, 1932.
The determination of the ratio e/p is carried out in the same manner as
the pistonphone.
c. Electrostatic Actuator.l2.-The electrostatic actuator consists of an
auxiliary electrode in the form of a grill mounted in front of the microphone
diaphragm, Fig. 10.1e. The actuator is perforated so that it does not
A B c
P = 8.85eoe
d2
X 10-7 10 3
•
':;S'~"
The acoustical impedance of the
microphone, M, and the reversible A
transducer, 51, should be the same
in order to obtain accurate results.
Furthermore, the acoustical im
pedance of the microphone at the
diaphragm should be high. Micro- B
phones satisfying these requirements
,:;&d[:~"
are, in general, condenser and piezo
electric microphones.
The first and second experiments c A-C
INPUT
in the reciprocity procedure are
shown in Fig. 1O.2A and Fig. 1O.2B.
An alternating current is fed to the FIG. 10.2. The three experiments of the
reciprocity procedure for obtaining the
loudspeaker, 52. A sound pressure, pressure calibration of a microphone.
PI, is produced in the volume having A. The open-circuit voltage, es, of the
an acoustical capacitance CA. Let reversible microphone loudspeaker, S1>
the open circuit voltage, in statvolts, when used as a microphone and actuated
by a sound pressure, Pl' B. The open
of 51 used as a microphone be desig circuit voltage, eM, of the microphone, M,
nated as es and the open circuit to be calibrated, when actuated by a sound
voltage output, in statvolts, of the pressure, Pl' C. The open-circuit voltage,
microphone M be designated as eM. eM', of the microphone, M, to be calibrated,
Let Ks = output, in statvolts per when actuated by a sound pressure, P2,
produced by the reversible microphone
dyne per square centimeter, of 51 loudspeaker, S1> used as a loudspeaker
and KM = output, in statvolts per with a current input, i, and a volume
dyne per square centimeter, of M. coupling, CA.
Since the sound pressure, PI, in dynes
per square centimeter, is the same for 51 and M, it is evident that
PI = ~ = eM lOA
Ks KM
In the experiments of Fig. 10.2, it is assumed that the acoustical imped
ance of the diaphragm of the units 51 and M are large compared to the
14 Rayleigh, "Theory of Sound," The Macmillan Company, Vol. 1, p. 145.
16 Cook, R. F., Jour. Research, Natl. Bur. Standards, Vol. 25, No.5, p. 489, 1940.
17 McLean, W. R., Jour. Acous. Soc. Amer., Vol. 12, No. 1, p. 140, 1940.
18 " American Standard Method for the Pressure Calibration of Laboratory Standard
KM = e'M 10.7
P2
where P2 = sound pressure, in dynes per square centimeter.
From equations to.5, 10.6, and to.7,
K M K s= jwCAe'M
. to.8
z
10.9
KM=J~~
esZ
10. to
where es, eM, e'M, and i are obtained from the three experiments A, B, and C
of Fig. to.2. The acoustical capacitance CA is obtained from the dimensions
of the coupling cavity, The voltages are in statvolts and the currents in
statamperes and the acoustical capacitance in (centimeter)5 per dyne.
Equation 10.10 was derived for electrostatic or condenser transducers.
However, equation 10.10 applies to other transducers provided the stipula
tions in the derivation are satisfied.
2. Field Response.-The field or free-wave response frequency charac
teristic of a microphone is the ratio elP as a function of the frequency, where
e is the open-circuit voltage generated by the microphone, in volts, and P
19 Olson, "Dynamical Analogies." D. Van Nostrand Company, Princeton, N.J.,
1943.
MEASUREMENTS 429
is the sound pressure, in dynes per square centimeter, in a free progressive
wave prior to the introduction of the microphone.
At the present time the Rayleigh disk and the reciprocity procedure are
the two most common methods in use today for obtaining the field response
frequency characteristic of a microphone. It is the purpose of this section
to describe the calibration of a microphone by means of the Rayleigh disk
and reciprocity methods.
a. Rayleigh Disk. 2o ,21,22,23-Rayleigh observed that when a disk was
suspended by a light fiber it would tend to turn at right angles to the im
pinging sound wave. Koenig24 developed the formula for the turning
moment of the disk as
M = ~3 pa 3u 2 sin 28 10.11
,p = ~ 10.12
5
where 5 = moment of torsion of the suspension, in dyne centimeter.
The moment of torsion of the suspension is given by
LAMP
OPTICAL
SYSTEM
Fig. 10.3 the deflection of the disk can be determined from the deflection of
the light beam on the scale.
b. Reciprocity.-The acoustical reciprocity theorem was originally
enunciated by Helmholtz and Rayleigh. 25. Ballantine 26 established
reciprocity theorems for mechanoacoustic, electromechano, and electro
mechanoacoustic systems. Ballantine also carried out a generalized dis
cussion to show that a microphone may be calibrated in terms of electrical
standards by the use of the extended reciprocity relations. Later other
investigators 27 ,28,29 extended the applications of reciprocity in both closed
and field systems. It is the purpose of this section to outline the reciprocity
procedure for the field calibration of microphones.
For the application of the reciprocity principle to the calibration of a
microphone, three transducers are used as follows: the microphone, M, to
be calibrated, a reversible microphone loudspeaker 51, and a loudspeaker 52.
For the reversible microphone loudspeaker it is convenient to use a small
back-enclosed loudspeaker.
The first and second experiments are shown schematically in Fig. lOAA
and Fig. lOAB. An alternating current is fed to the loudspeaker 52. A
sound pressure PI is produced at a distance d. Let the open-circuit voltage,
in abvolts, of 51 used as a microphone be designated as es and the output of
the microphone M be designated as eM. Let Ks = output, in abvolts per
25 Rayleigh, "Theory of Sound," The Macmillan Company, Vol. I., p. 145.
26 Ballantine, S., Proc. Inst. Rad. Engrs., Vol. 17, No.6, p. 929,1929.
27 Cook, R. K., Jour. Research, Natl. Bur. Standards, Vol. 25, No.5, p. 489, 1940.
28 McLean, W. R., Jour. Acous. Soc. Amer., Vol. 12, No.1, p. 140, 1940.
MEASUREMENTS 431
dyne per square centimeter of 51, and, KM = output, in abvolts per dyne
per square centimeter, of M. Since the sound pressure PI, in dynes per
square centimeter, is the same for 51 and M 1 , it is evident that
es eM
P1=-=- 10.14
Ks KM
The voltage output,30 m abvolts, of the microphone loudspeaker 51
used as a microphone is
es = ElX1 10.15
where B = flux density in the air gap, in gausses,
l = length of the conductor, in centimeters, and
Xl = velocity of the voice coil, in centimeters per second.
Sa
A-C ----r-i/1
B IN PUT -L----""J
. S,
~e~
A-C~
C I N PUT -----L---"'J
I· d -----;·~I
~ = ElA = Ks 10.17
PI ZM
P -_ pckiKs _ rAiKs
4-rrd - U>.. 10.21
where r A = pc.
The sound pressure, p, in dynes per square centimeter, at M in terms of
the constant KM and the open-circuit voltage e'M, in abvolts, is
10.22
KM = J2d)..e~e' M 10.25
r AleS
where es, eM, e'M, and i are obtained from the experiments of Fig. 10.4.
The units are as follows: Voltages in abvolts, currents in abamperes, dis
tances in centimeters, wavelengths in centimeters, and r A = pc = 41.5.
The calibration of microphones by the Rayleigh disk and reciprocity
methods should be made under free-field conditions, that is, in a large room
in which the reflections are negligible or outdoors at a great distance from
reflecting surfaces. A free-field sound room suitable for these measure
ments is described in Sec. 1O.3A4.
A high-quality microphone calibrated by any of the above methods may
be used as a secondary standard for the calibrations of other microphones.
3. Secondary Calibration of Microphones. 32-A secondary calibration of a
microphone is obtained by a comparison of the response of the microphone
to be calibrated with the response of a microphone that has been calibrated
by primary means described in the preceding sections. Response frequency
characteristics are obtained on the two microphones. The ratio response
of the microphone being calibrated to the response of the calibrated micro
phone yields the calibrated response of the microphone being calibrated.
4. Artificial Voice.-The proximity of the head in close talking speech
type microphones influences the response frequency characteristics. There
fore, in testing microphones of this type it is desirable to provide testing
means 33 which simulate actual operating conditions. The artificial voice
consisting of a small loudspeaker unit mounted in the head of a manikin,
as shown in Fig. 10.5, provides a means for obtaining the response frequency
characteristics of close talking microphones. Resonances in the tube
connecting the loudspeaker unit and the mouth are eliminated by the intro
duction of series and shunt mechanical resistances. The response frequency
characteristic shown in Fig. 10.5 can be obtained by a suitable choice of
constants of the mechanical system.
5. Artificial Throat.-Throat microphones have been described in Sec.
8.7. Throat microphones are actuated by sound waves transmitted through
the throat. An artificial throat 34 for testing throat microphones consists of
a mass-controlled system driven by a voice coil located in magnetic field.
Specifically the voice coil is coupled to a massive platform. The centering
system is made very compliant to insure mass control. In order to maintain
constant velocity with respect to frequency the driving oscillator and
amplifier are compensated so that the current through the voice coil is
proportional to the frequency. The platform system, which the voice coil
32 American Standards Association, "Secondary Calibration of Microphones,"
z-24.11- 1954.
33 Inglis, Gray, and Jenkins, Bell Syst. Tech. Jour., Vol. 11, No.2, p. 293,1932.
34 Greibach, E. H., Elec. Eng., Vol. 65, No.4, p. 184, 1946.
43+ ACOUSTICAL ENGINEERING
20
OJ f
015
_I--..
r--
"'Z
<II
10
1/
oQ.
<II
"'
a: 5
4 8 103 2 810 4
fREQUENCY
FIG. 10.5. Perspective view, sectional view, mechanical network, and response
frequency characteristic of an artificial voice. In the mechanical network, mD = the
mass of the diaphragm and suspension of the small loudspeaker unit. YMD and
C MD = the mechanical resistance and compliance of the suspension system of the
small loudspeaker unit. C MO = the compliance of the air chamber behind the
diaphragm. mI' YM • • • m5' YM5 = the mass and mechanical resistances of the series
elements in the pipe. YMI', C MI . . . YM4', C M4 = the mechanical resistances and
compliances of the shunt elements of the line. mu and YMU = the mass and mechan
ical resistance of the air load on the mouth. The response frequency characteristic
depicts the free-field sound pressure at a distance of 2 inches.
OSCILLATOR
when it is actuated by a pure tone. The plot of the total distortion, in per
cent of the fundamental, is termed the distortion characteristic. It is also
common practice to plot the individual components in per cent as the
distortion characteristics.
It is difficult to obtain a sound source which will generate an intense
sound wave of very low distortion in free space. The arrangement 35 shown
in Fig. 10.7 provides a simple means of obtaining a sound wave free from
35 Phelps, W. D., Jour. Acous. Soc. Amer.• Vol. 11, No.2. p. 219. 1939.
436 ACOUSTICAL ENGINEERING
INCIDENT
SOUND
~
CATHODE-RAY
OSCILLOGRAPH
used as the reference standard. The standard microphone and the micro
phone to be tested may be placed side by side in a plane progressive wave
in free space, Fig. 10.8. The outputs of the two microphones are amplified
by separate identical amplifiers and connected to the vertical and horizontal
plates of a cathode-ray oscillograph. The resultant Lissajou figure indicates
the phase relations between the output of the two microphones. The two
microphones are shifted relative to each other in a line parallel to the direction
of propagation until the outputs of the two microphones are in phase. The
phase angle, in degrees, between the output of the two microphones is
4> = ~A 360 0
10.26
outdoors under wind conditions. It is very difficult to test the wind response
under actual wind conditions because it is impossible to find constant wind
conditions. Wind consists of a steady flow of air with superimposed pulses
of air. A wind generator 36 which delivers a steady flow of air with super
imposed pulses is shown in Fig. 10.9. Almost any practical value of wind
velocity and pulses can be obtained by adjusting the speed of rotation and
36 Olson, Preston, and Bleazey, Unpublished Report.
438 ACOUSTICAL ENGINEERING
the angles of the paddles. Comparison of the results obtained with actual
wind with those of the wind generator shows very good correlation with
respect to the distribution of the frequency components. This indicates
that the wind components in the two cases are the same.
In another method 37 for obtaining the wind response of a microphone,
the microphone is suspended at the end of a pendulum about ten feet in
length. In this way maximum wind velocities up to 20 miles an hour can be
obtained. In addition, there is a variation in velocity from zero to the
maximum velocity.
10.3. Testing of Loudspeakers. 38 ,39-Many different measurements are
required to determine the performance of a loudspeaker. The most
important characteristics which depict the performance of a loudspeaker
are as follows:
1. Response frequency characteristic
2. Directional characteristic.
3. Nonlinear distortion characteristic
4. Efficiency frequency characteristic
5. Phase distortion characteristic
6. Electrical impedance characteristic
7. Transient response characteristic
VZE
MEASUREMENTS 439
characteristic represents the sound pressure as a function of the frequency
which would be obtained from the speaker if fed from the generator which
would automatically deliver constant apparent power, e2 jzE, to the voice
coil over the frequency range.
The response may be expressed by a value equal to the above ratio or
may be expressed in decibels relative to an arbitrary value of response
corresponding to 1 volt, 1 ohm, and 1 dyne per square centimeter.
p
Absolute response =
ejvzE
20 loglO - 1 - = 20
PVZE
10glO -e- 10.28
1/0
The apparatus and methods employed for obtaining the response fre
quency characteristics of loudspeakers will be described in the sections
which follow.
2. Apparatus for Measuring the Sound Pressure Frequency Relationship
of a Sound Source.-An arrangement for obtaining the sound pressure
frequency characteristic by the semiautomatic method 40 is shown in Fig.
10.10.
This method yields a response frequency curve on semilogarithmic paper
in about 3 minutes. Rotation of a condenser governing the beat frequency
HANDLE
41 Ballantine, S., Jour. Acous. Soc. Amer., Vol. 5, No.1, p. 10, 1933.
with white wax is moved under a stylus by a motor which changes the value
of the beat frequency generated at the same time. The loudspeaker under
test is connected to the output of the beat frequency generator and the
variations in response are recorded on the paper directly on a decibel scale
by a stylus which scratches through the wax coating on the recording
paper.
The rectifier output incorporates a control circuit which causes direct
current to flow through one circuit when the rectifier current is less than a
certain critical value and through a second circuit when it is greater than
a second critical value. In the first case, the control circuit operates a
magnetic clutch which causes a potentiometer to operate and increases
the voltage. In the second case, the voltage is decreased.
The output of the rectifier is kept balanced to within the voltage change
produced by a change in potentiometer corresponding to the smallest unit
43 Wente, Bedell, and Swartzel, Jour. Acous. Soc. Amer., Vol. 6, No.3, p. 121, 1935.
MEASUREMENTS 441
of the attenuator calibration. The motion of the potentiometer is com
municated to the stylus which gives a trace on the recording paper. The
same motor which drives the oscillator frequency control moves the poten
tiometer by means of the magnetic clutches.
The speed with which changes in sound level are recorded may be varied
from 10 to 560 db per second through alteration of the speed of rotation of
the clutches.
In another design 44 of high-speed level recorder a thyratron actuated
reversible motor drives a fountain pen and records directly on graph paper.
The speed is somewhat slower than the clutch system but the conventional
paper record is more convenient to use and file.
A high-speed level recorder45 with a dynamic drive is shown in Fig. 10.12.
The dynamic driving mechanism is used to drive the stylus and the contactor
POLE
SECTIONAL VIEW
MICRO
PHONE
LOW
PASS
AMPLIriER
F"lLTER
AMPLIF"IER
PERSISTENCE IMAGE
110 V.A.C. CATHODE -RAY OSCILLOGRAPH
trace. The maximum recording speed is about 900 db per second with a
potentiometer having a 60 db range.
A cathode-ray tube,46,47 with a long persistence screen, may be used as
a response indicator and recorder, Fig. 10.14. A motor drives the beat
frequency oscillator and a potentiometer. The potentiometer varies the
46 Hackley, R. A., Broadcast News, No. 28, p. 20, July, 1938.
MEASUREMENTS 443
voltage on the horizontal deflection plate of the cathode-ray tube and
thereby drives the cathode-ray beam across the tube in synchronism with
the oscillator. A reversing switch changes the direction of the motor travel
at the upper and lower limits of the audio-frequency range. The output of
the oscillator actuates the loudspeaker. The sound is picked up by the
microphone and amplified. The output of the amplifier is detected by a
linear or logarithmic detector and fed to a low-pass filter. The output of
the filter is amplified by a d-c amplifier, the output of which is connected
to the vertical plates of the cathode-ray tube. The cathode-ray beam traces
the response charactenstic upon the persistence image screen. The ordinates
are in decibels when the logarithmic detector is used. The ordinates are
proportional to the sound pressure when the linear detector is used. The
A~r" ~"'''''I
LOUD SPEAKER MICROPHONE
0:0 2 3 .4 5 6
"i!""j"cr"",<I
0:0 2 3 4 5 6
C~i
0:0 I 2 3 4 5 6
F"REQUENCY IN KC
position of the band-pass filter is varied with respect to frequency. The out
put of the filter is detected and measured by means of a meter. The
response characteristic of a loudspeaker is shown in Fig. 1O.15C.
Apparatus employing thermal noise for obtaining response characteristics
has not been developed to the stage where it may be used with the facility
of other methods. It appears, however, that this type of measurement
will become very important for all types of acoustical measurements when
suitable apparatus has been developed.
3. Calibration of the Sound Measuring Equipment. 49-The microphone
should be calibrated in terms of the pressure in a free progressive sound
wave. The microphone, amplifier, and detector should have a combined
characteristic which is substantially independent of the frequency over the
frequency range under consideration. If it is not substantially constant
over the frequency range the data must be adjusted for known variations.
MEASURING
SYSTEM
FIG. 10.16. Schematic arrangement for obtaining the factor Pie in the
formula for absolute response of a loudspeaker.
long, 36 feet wide, and 36 feet high. The next objective was to obtain an
absorption coefficient as near unity as possible. The high- and low-fre
quency ranges present the greatest difficulty in attaining this objective.
It is a comparatively simple matter to attain high absorption in the mid
frequency range. In the high-frequency range the principal difficulty is
reflection from grills, control boxes, and test apparatus. These reflections
can be eliminated by acoustical treatment of these reflecting surfaces. In
the case of the low-frequency range it appears to be an inexorable fact that
the ideal objective can be attained only in a relatively large room with
correspondingly thick absorption material. An examination of existing
rooms indicates that regardless of the form of treatment it appears that
absorption deviates quite rapidly from unity when the thickness of the
treatment is less than a quarter wavelength. In this statement, it is
assumed that thickness of the material is measured to an outside boundary
of relatively high acoustical impedance compared with the characteristic
acoustical impedance of air. It is also assumed that treatment does not
involve resonant systems.
The absorbing system employed in this room is of the baffle type, that
is, strips of absorbing material arranged normal to the walls of the room as
shown in Fig. 10.17. Several years ago a smaller room (22 feet long, 20
feet wide, and 13 feet high) was treated with baffles. The performance of
this room appeared to be comparable to rooms with equivalent thickness
of other types of absorbing material. The advantage of the baffle type of
treatment is the relatively simple construction and lower cost as compared
with more elaborate absorbing systems.
Plan and elevation views of the room are shown in Fig. 10.17. One
inch Ozite is spaced 1 foot from the walls, ceiling, and floor. One-inch Ozite
baffies, 7 feet in length and spaced 2 feet apart, are placed normal to the
walls, ceiling, and floor. Four-foot baffles of the same material are placed
between the 7-foot baffles. The total thickness of the absorbing material,
measured from the outside wall, is 8 feet. This leaves the inside dimensions
of the room 32 feet long, 20 feet wide, and 20 feet high. A special grill,
12 feet wide and 24 feet long, is supported on vibration-isolated feet. The
ratio of open to total area in the grill is 0.87. This is a relatively open grill
when it is considered that the grill platform will carry a load of 200 pounds
per square foot. The floor level of the grill is located 11 feet above the
floor level of the room. The floor level of the grill coincides with the first
floor level which makes it readily accessible to the adjoining laboratory.
The acoustical merit of the room can be expressed by the deviation in
sound pressure from an inverse distance characteristic. Pressure response
frequency characteristics were obtained at various distances from a small
loudspeaker. The maximum deviation in pressure from an inverse distance
characteristic for various frequencies is shown in Fig. 10.18. It will be
noted that the deviation in the mid-frequency ranges is negligible. The
deviation at the high frequencies is due to the grill, overhead trolley track,
power and signal outlet boxes. These units which if treated, will make
FREE- FIELD SOUND ROOM O~~'_~IO
I Itt t ~
II
~1=I=~~~WI+l+l-----i-~~ MONORAIL
L I i I
r ~ IIi i i Iii !iI
i iIH Ii Hi!!--: i
~ !! li i d!i H!!d HlHiii 1~ 1:
,~k
l!l!
~
~
SECTION A·Iit
---
1.00
I -I"'"
~ .99
~
U
..."
au .98
:3
Z .97
o
j:;
Q.
~ .96
<Il
ID
-<
.9~0 100 1000 10,000
fREQUENCY IN CYCLES PER SECOND
FACE VIEW
FIG. 10.20. Wall treatments for free-field rooms, X. Baffle type; Y. Pyramid
type; Z. Wedge type.
frequency range above 40 cycles for distances between the source and obser
vation up to 8 feet. This distance can be increased if either the source or
the microphone, or both, are directional.
Three common types of sound absorbing systems used for free-field or
anechoic rooms are shown in Fig. 10.20. Fig. 10.20, X depicts the baffle
type of sound absorbing system employed in Fig. 10.17 and described in this
section. In the system described in this section the spacing between the
baffies is 12 inches. Each baffle is made up of two layers of i-inch Ozite with
actual over-all thickness of 2 inches because the material was obtained in
the uncompacted form. Employing twice as many baffles would improve
450 ACOUSTICAL ENGINEERING
VACUUM
TUBE
VOLTMETER
HARMONIC
ANALVZER
FIG. 10.21. Schematic arrangement of the apparatus for measuring the non
linear distortion of a loudspeaker.
20
o
'" -20
o
~
v "
....
~-40
oDo
/ \
en
~-60
/ \
II r-...
-80
V
VACUUM
TU8E
VOLTMETER
SCHEMATIC DIAGRAM
-100
6420246
CYCLES OF RESONANCE
"
FIG. 10.22. Schematic arrangement of the elements in a heterodyne-type harmonic
analyzer.
HARMONIC
ANALYZER LEVEL
RECORDER
INPUI~Ti.~::==~~~t======2~C:~==~:::~
INPUT
WITH PERSISTANCE
IMAGE SCREEN
~~r~~~~~~ r-----1
AMPLIFIER
function of the frequency. The vertical scale may be either linear or decibel.
The frequency scale is logarithmic.
The balance bridge for measuring the total distortion is shown sche
matically in Fig. 10.25. A part of the output of the oscillator is fed to the
apparatus to be tested and another part to the analyzer. The amplitude
and phase relations of the fundamentals from the oscillator and apparatus
VACUUM
TUBE
VOLTMETER
VACUUM
TUBE
VOLTMETER~--~:::~~____,
CATHODE - RAY
OSCILLOGRAPH
INPUT SIGNAL
B~-------------------+~~----------------~,
OUTPUT SIGNAL
c~V
pf\Aflflflf\I\Q{)"~"""fl
----
D
REGTIFIED MODULATED HIGH fREQUENCY
ReSULTANT INTERMODULA'rION
FIG. 10.28. Typical wave shapes in the various stages of the inter
modulation system of distortion measurement. A. The input signal to
the loudspeaker. B. The ontput signal of the loudspeaker. C. The
modulated high-frequency output of the band-pass filter. D. The recti
fied modulated high frequency of the detector. E. The resultant
intermodulation output of the copper oxide rectifier.
58 Olson and Pennie, RCA Review, Vol. 12, No. 1. p. 35, 1951.
MEASUREMENTS 459
high-pass filters, sequentially interposed between the microphone pickup
and the recording equipment to attenuate the 40- to IS,OOO-cycle sweep
fundamental. The primary advantage of using this method for harmonic
distortion measurements is its dependability. The filters themselves may
be made very rugged. Furthermore, should the filter switching circuit fail
to function properly, the distortion readings will immediately go to 100 per
cent, thus reading fundamental rather than harmonic, and thereby providing
a positive check against a possible switching error.
The useful frequency range of each filter is determined by two frequencies,
namely, fe and f 00" The frequency at which the response is down one
r------------..,
I I
I I
I
RECORD
• 1()4
FIG. 10.29. Schematic arrangement of the apparatus employed in the automatic non
linear distortion recording system. The graph shows the response-frequency charac
teristic and the distortion frequency characteristic of a typical commercial 12-inch
loudspeaker.
decibel is fe. This frequency sets the lower limit of the useful pass band of
each high-pass filter when recording distortion to an accuracy of 10 per cent.
The frequency at which the response is down sixty decibels is fr;o This
frequency sets the upper limit of the useful rejection band of each filter
when recording distortion to an accuracy of 10 per cent for a 0.3 per cent,
2nd harmonic distortion value. The response frequency characteristics of
the filters are shown in Fig. 10.30. The fe and fx) overlap characteristic of
adjacent filters are very close at the lower frequencies; and it, therefore,
becomes very important that the frequency at which a filter is switched be
held to a close frequency tolerance if the full possible accuracy of the distor
tion analysis is to be realized. For this reason, an electronic rather than a
mechanical system for the detection of the switch frequency is used. A
460 ACOUSTICAL ENGINEERING
r ,-
2
(
3
{
4' 5 6{ 1/ 8
f 9 10' II 12 I. 14 15
{
-10
-20
""
~ -30
.
Z
o
III
"'" ·40
\J
-50
-60
~ I I
40 80 100 200 400 8001000 2000 4000 8000 10000 15000
FREQUENCY IN CYCLES PER SECOND
FIG. 10.30. Response frequency characteristics of the fourteen high-pass filters used in
the system of Fig. 10.29.
step switch solenoid power turned on and with the gain control of the pre
amplifier turned some 20 to 40 decibels higher. The resultant character
istics, with due consideration for the difference in preamplifier settings,
gives the relative distortion frequency characteristic of the loudspeaker
under test, see Fig. 10.29.
D. Efficiency Frequency Characteristic. 59-The efficiency of a loudspeaker
at any frequency is the ratio of the total useful acoustical power radiated
to the electrical power supplied to the load, the current wave of which
exercises a controlling influence on the wave shape of the sound pressure.
The plot of efficiency, in per cent, versus frequency, in cycles, is termed the
efficiency frequency characteristic.
The measurement of efficiency of a loudspeaker may be divided into two
methods, direct and indirect. One direct method depends on measuring
the total energy flow through a spherical surface without reflections. Several
59 Standards on Electroacoustics, Institute of Radio Engineers, 1938.
MEASUREMENTS 461
indirect methods have been developed. The most common of these consists
in measuring the electrical impedance under various conditions of diaphragm
loading. It has been found in practice that these two methods of deter
mining efficiency are those most widely used at the present time.
1. Direct Determination oj Radiated Power.-The sound power output
from a speaker at a particular frequency may be obtained by measuring
the total flow of sound power through a spherical surface of which the sound
source is the center. The surface of the sphere is divided into incremental
areas and the power transmitted through each area is determined from the
sound pressure and the particle velocity as well as the phase displacement
10
o ,.-
""\
CD -10
~
o
~-20
z
:r
(/)
.....-SWITCH
LEVEL
~-30
ELECTRICAL
NETWORK
-50
-6 -3 0 3 6
FREQUENCY
PA = :c JJ p2 dS X 10-7 10.31
where P A = total acoustical power, in watts,
p = density of the medium, in grams per cubic centimeter,
c= velocity of sound in medium, in centimeters per second,
p= root-mean-square pressure, in dynes per square centimeter,
over the element of areas dS, and
dS = element of area on spherical surface, in square centimeters.
462 ACOUSTICAL ENGINEERING
The input electrical power can be determined from the electrical current,
voltage, and phase angle, while operating under the above conditions.
The efficiency, fL, in per cent, is then
PA
fL = P E X 100 10.32
where P A = total acoustical output, in watts, and
PE = electrical input, in watts.
As previously mentioned, the loudspeaker should be located so that the
reflected energy reaching the measuring equipment is negligible. This
means that the measurements must be made either in a free-field sound room
SQUARE
AMPLIFIERS LAW
I
r-------------,I DETECTORS
: I
I
I
I
I
I
I
I
I LOUD
ISPEAKER
POWER
I I
AMPLIFIER I
IL ____________ ....JI D-C
AMPLIFIER
OSCILLATORr==~~~r===::~~~~J
SHAFT LEVEL LOGARITHMIC
DRUM
PEN RECORDER AMPLIF'lER
conditions. This method also assumes that there are no losses due to viscous
air friction. Since the amplitude of the vibration of a voice coil is normally
small at the higher frequencies, the problem of blocking the voice coil
against motion is not a simple matter. Obviously, any motion will intro
duce an error in the determination of the efficiency.
E. Phase Distortion Characteristic.-The phase distortion characteristic
of a loudspeaker is a plot of the phase angle between the sound output and
some reference sound as a function of the frequency.
Two microphones and separate amplifiers and a cathode oscillograph
may be used as outlined in Sec. lO.2D, Fig. 10.8. A reference sound may
be set up by a separate loudspeaker, in which the phase shift is small, and
SQUARE
WAVE
GENERATOR
APPARATUS
UNDER
TEST
S
CATHODE
RAY
OSCILLOGRAPH
-~L----+<>S
SQUARE
WAVE I
GENERATOR LOUD SPEAKER MICROPHONE CATHODE
RAY
OSCILLOGRAPH
SOUND ROOM
CATHODE RAY
OSCILLOSCOPE
FIG. 10.34. Schematic arrangement of the apparatus used in the tone burst
system for depicting the transient response of a loudspeaker.
63 Corrington. M. S.• Jour. Audio Eng. Soc .• Vol. 3. No. 1. p. 35. 1955.
fed to an electronic gate which intermpts the sine wave signal at regular
intervals, thereby producing a series of tone bursts. The tone bursts are
amplified and fed to the loudspeaker. The sound output from the loud
speaker is picked up on the microphone and amplified and applied to the
vertical deflection system of a cathode-ray oscilloscope. The acoustical
output from the loudspeaker may be compared with the electrical input to
the loudspeaker. Another gate may be employed so that the residue after
the electrical input has ceased may be displayed on the oscilloscope. A
delay network is required to compensate for the time required for the wave
to travel from the loudspeaker to microphone.
H. Subjective Measurements. 64 _A subjective test of a loudspeaker
involves a determination of some of the performance characteristics by
direct listening to the loudspeaker operating under controlled program and
environment conditions. Listening tests play an important part in research,
development, and commercialization of loudspeakers. Listening tests
range in scope from exceedingly simple comparison tests to elaborately
controlled and conducted tests. In many cases the apparatus for making
all the objective tests outlined in the preceding sections are not available.
Furthermore, there is always some difficulty in evaluating the objective
measurements. For this reason a sUbjective test of efficiency, frequency
response, directional characteristics, nonlinear distortion, and transient
response, in which two or more loudspeakers are compared with each other
or with a reference loudspeaker, is widely used. The subjective test
may be used to determine the lumped effects of the following factors:
loudness, frequency range, tone balance, spacial distribution, quality, and
hangover. It is the purpose of this section to describe listening tests of loud
speakers.
1. Loudspeaker Environment.-The listening test of a loudspeaker should
be conducted in the environment for which it was designed to {)perate.
Specifically, a loudspeaker designed for home-type radio receivers, phono
graphs, and television receivers should be tested in a room with dimensions
and acoustics similar to those of the average living room in the home. A
loudspeaker for an automobile radio receiver should be tested in an auto
mobile. A loudspeaker for a sound motion-picture system should be tested
in a typical theater. A loudspeaker for a public address, sound reinforcing,
or paging system should be tested under typical surroundings for these
systems.
2. Loudspeaker Housing, Placement, and Mounting.-The horn, baffle,
housing, or cabinet for loudspeaker listening tests should be similar to those
used under actual operating conditions. The placement and mounting
arrangement in the test invironment should correspond to those used in
actual installations.
3. Signal Sound Level.-The signal sound level produced by a loudspeaker
in a listening test should correspond to the sound level under actual operating
conditions in the field. The use of the proper level is very important in
64 Olson, H. F., Trans. IRE, Prof. Group on Audio, Vol. AD-l, No.5, p. 7, 1953.
MEASUREMENTS 4-67
detennining the balance of high, mid, and low frequencies, the distortion,
the transient response, etc., under actual operating conditions. The upper
sound levels in the description which follows do not necessarily represent the
upper power capabilities of the systems. The signal sound level will be
somewhere between 65 and 75 db for a radio receiver, phonograph, or
television receiver operating in a typical or average living room. The
signal sound level will be somewhere from 65 to 100 db for an automobile
radio receiver. The signal sound level will be between 76 to 85 db for speech
reproduction, and 75 to 95 db for music reproduction for a sound motion
picture reproducing system operating in a theater. The signal sound level
for a public address, sound reinforcing, or paging system will vary over wide
limits depending upon the application. To summarize, the signal sound
level of the test should correspond to the level under actual operating
conditions.
4. Ambient Noise Level.-The ambient noise under which the listening
test is conducted should correspond to the ambient noise encountered under
actual conditions. This involves two main factors, namely, the sound level
and spectrum of the ambient noise. For example, the average ambient noise
sound level in the average living room is 42 db. The average ambient
noise sound level in a theater is also 42 db. In an automobile, the ambient
noise sound level depends upon the speed, open or closed windows, the road,
etc. In public address, sound reinforcing or paging applications, the noise
sound level will vary over wide limits. This must be taken into account, and
the noise conditions under which the equipment will be operated must be
simulated in the listening tests. It is important that the spectrum of the
noise encountered under the actual operating conditions should be simulated
in the subjective tests as well as the noise level.
5. Signal or Program Material.-The signal or program material used in
listening tests should be similar to that encountered in the field. This is
not so for the material presented under" Frequency Range" and" Power
Handling Capacity." A radio or television receiver should be operated
from typical broadcast or television transmitters. Under certain condi
tions it may be necessary to use the equivalent of a radio or television
transmitter, as, for example, a modulated signal generator. A phonograph
should be operated from typical commercial records. A sound motion
picture reproducing system should be operated from typical sound motion
picture film. Sound reinforcing systems for use with music should be tested
with musical program material. Public address and paging systems should
be tested with speech as the program material.
6. Reference Systems.-Almost all listening tests on loudspeakers are con
ducted by comparing the loudspeaker under test with a reference loudspeaker.
The reference system is, in general, a loudspeaker which is similar to the
loudspeaker under test. The loudspeakers should be placed behind a light
opaque, sound-transparent curtain so that it is impossible to identify the
loudspeakers by sight. A suitable indicator should show which loudspeaker
is operating at any time. In general, the procedures in most listening
468 ACOUSTICAL ENGINEERING
tests are not formalized because the tests are conducted to determine the
engineering and commercial aspects of rather small changes in design. If a
jury-type procedure is used, secret ballots should be taken of the preference.
Statistical methods should be employed in planning and conducting such
jury tests.
7. Relative Loudness Efficiency.-The relative loudness efficiency of a
loudspeaker is determined from a loudness balance. High-quality trans
formers should be used to match each loudspeaker to the appropriate im
pedance. In some cases it is desirable to include the driving means in
determining the efficiency, because this is important in any practical design.
The input to the loudspeakers should be adjusted so that the loudness levels
of all loudspeakers are the same. The attenuation required to adjust to
the same loudness gives a measure of the relative loudness efficiency. In
these tests, the observers should move around to different locations to
insure that no advantages are given to any loudspeaker due to a better
listening location. For the same reason, the locations of the loudspeakers
should also be interchanged.
8. Relative Directivity.-The relative directivity of a loudspeaker is
determined by listening at observation points removed from the axis. In
order to reduce the effect of the difference in the angle during a comparison,
the following precautions should be observed: Only two loudspeakers should
be used at a time. The loudspeakers should be placed as close together as
possible. The position of the two loudspeakers should be interchanged
during the test. In determining the relative directivity, listening tests should
be conducted along different angles with respect to the axis. This test
indicates the loss in loudness. level and frequency discrimination for observa
tion points removed from the axis.
9. Frequency Range.-The approximate frequency ranges of loudspeakers
may be determined from listening tests by employing program material
which has a wider frequency range than the loudspeaker under test in
combination with calibrated high- and low-pass filters introduced between
the program source and the loudspeaker. It is very important that
the program material contain adequate frequency components in both the
high- and low-frequency ranges and thereby insure reliable results. The
approximate frequency range can be determined by noting the settings of
the filters for which there is no appreciable frequency discrimination, as
determined by the quality of reproduction. The filters should have at
least three cutoff steps per octave.
10. Power Handling Capacity.-The power handling capacity of a loud
speaker may be determined by employing a low distortion program source
capable of overloading the loudspeaker without introducing distortion in the
program source which is fed to the loudspeaker. The frequency range of
the system which feeds the loudspeaker should be restricted by means of
filters to correspond to that of the loudspeaker under test. The power level
at which the distortion becomes intolerable may be considered to be the
power handling capacity of the loudspeaker. In this connection intolerable
MEASUREMENTS 469
distortion depends upon the application in which the loudspeaker is to be
used. This requires a high order of judgment by the listener.
The test outlined above for determining the power handling capacity may
appear to be oversimplified in view of the many factors involved. For
example, the power handling capacity of a loudspeaker may be determined
by failure of the diaphragm, the suspension system, the voice coil structure,
and heating of the voice coil. Of course, all these forms of failure will be
manifested as intolerable distortion. Again it should be emphasized that
the crux of this test is the determination of what is considered intolerable
distortion.
11. Response Frequency Contour.-In most completely integrated systems,
such as radio and television receivers and phonographs, there are distinct
economic and technical advantages in employing components which indi
vidually do not exhibit a uniform response frequency characteristic but
taken collectively do exhibit a uniform response frequency characteristic.
In these applications, listening tests are very useful in checking the objective
measurements for a proper balance of the frequency characteristic. This
type of listening requires great skill obtained through practice. A reference
system which is known to be acceptable is almost a necessity in tests of this
type.
12. Nonlinear Distortion.-Loudspeakers are used with other components
in a sound reproducing system. Therefore, in a properly integrated system
the limitations upon the allowable nonlinear distortion of each element
depends upon the allowable distortion of the system as a whole. For
example, it would be technically and economically unsound to use a wide
range, high-quality loudspeaker in a reproducing system in which the
components in the remainder of the system were of much lower quality.
The quality of the loudspeaker required for the application can be deter
mined from listening tests of loudspeakers of various degrees of quality.
In this way it is possible to determine the loudspeaker which introduces
distortion of such magnitude as to be perceptible above the distortion of the
remainder of the system.
13. Transient Response.-Since all speech and music are of a transient
character, the transient response is another important characteristic which
depicts the performance of a loudspeaker. Poor transient response leds to
fuzzy reproduction with poor definition. As a result the character of speech
and music is destroyed. In objective measurements, a deviation in the
sound output from the rapid growth and decay characteristic of an applied
tone burst depicts the transient response of a loudspeaker. See Sec. 10.3G.
A loudspeaker with a nonuniform response frequency characteristic will
exhibit poor transient response. A subjective evaluation of the transient
response of a loudspeaker can be obtained by listening to the reproduction
of such sounds, as piano, wood blocks, and other sounds with rapid growth
or decay characteristics. The growth and decay will not be clean but fuzzy
when the transient response of the loudspeaker is poor. If there is a
peak in the response frequency characteristic in the low-frequency range,
470 ACOUSTICAL ENGINEERING
LOUD
SPEAKER
RECEIVER
FIG. 10.35. Schematic arrangement of the apparatus for obtaining the sub
jective response frequency characteristic of telephone receivers. In A, the
observer listens to the sound produced by the loudspeaker. In B, the observer
adjusts the attenuator until the sound level produced by the telephone
receivers appears to be equal to the sound level produced by the loudspeaker.
In C, a sound level meter is used to measure the sound pressure produced by
the loudspeaker.
be the same as the intensity produced in the ear when the head is immersed
in the original sound field. There are two types of measurements upon
telephone receivers, namely, SUbjective and objective.
A. Subjective Measurements.-The schematic arrangement for obtaining a
subjective response frequency characteristic65 ,66 of telephone receivers is
shown in Fig. 10.35. A free progressive sound wave is established by means
of a loudspeaker driven by an oscillator and power amplifier. The test
65 Olson and Massa, Jour. Acous. Soc. Amer.• Vol. 6. No.4. p. 250.1935.
66 Olson. H. F., Jour. Soc. Mot. Pic. Eng.• Vol. 27. No.5. p. 537. 1936.
MEASUREMENTS 471
should be conducted in a free-field sound room (see Sec. 10.3A4). With the
receivers removed the observer listens to the sound produced by the loud
speaker as shown in Fig. 10.35A. Next the observer places the receivers
on his ears and the output of the oscillator and amplifier is transferred from
the loudspeaker to the headphones by throwing the switch and reducing the
power by means of a suitable attenuator as shown in Fig. 1O.35B. The
voltage across the receivers is adjusted until the intensity in the observer's
ear seems to be the same as the free-wave intensity produced by the loud
speaker. The sound pressure produced by the loudspeaker at the distance
of the observer is measured by means of a calibrated microphone, amplifier,
and meter combination as shown in Fig. 1O.35C. This procedure is repeated
at several frequencies. The absolute response of the telephone receiver is
given by equation 10.27, where p is the free-wave sound pressure, e is the
voltage across the telephone receivers when a balance is obtained, and ZE
is electrical impedance of the telephone receivers.
B. Objective Measurements.-l. Artificial Ear. 67 ,68,69,70,71-The acoustical
impedance frequency characteristic looking through the ear cap of a tele
phone receiver as normally worn has been investigated by Inglis, Gray, and
Jenkins. This is shown in Fig. 10.36. An artificial ear and the acoustical
network which yields approximately the same acoustical impedance charac
teristic are shown in Fig. 10.36. A standard condenser microphone is used
to measure the pressure. A series of slits corresponding to the leak between
the ear cap and the ear are represented by the inertance, ME, and acoustical
resistance, rAE. The walls of the cavity, CAE (4 cubic centimeters), are
67 Inglis, Gray and Jenkins, Bell Syst. Tech. Jour., Vol. 11, No.2, p. 293,1932.
68 Olson and Massa, Jour. Acous. Soc. Amer., Vol. 6, No.2, p. 250, 1935.
69 Olson, H. F., Jour. Soc. Mot. Pic. Eng., Vo. 27, No.5, p. 537, 1936.
70 Romanow, F. F., Jour. Acous. SOL Amer., Vol. 13, No.3, p. 294- 1942.
lined with felt to reduce resonances at the high frequencies. The response
frequency characteristic obtained upon the artificial ear, in general, agrees
quite well with the sUbjective tests.
The tests outlined for loudspeakers may be performed upon telephone
receivers by employing the artificial ear. The same apparatus as described
in Sec. lO.3A2 may of course be used and will not be repeated here.
The pressure delivered by a telephone to a closed cavity as a function of
the frequency is sometimes used to depict the response of a telephone
receiver. The artificial ear shown in Fig. lO.36 may be used for this purpose
by closing the slits.
2. Artificial Mastoid.-The artificial mastoid is a system for objectively
measuring the response of a bone conduction telephone receiver. In one
form the artificial mastoid 72 ,73 consists of a rubber block having approxi
mately the same acoustical impedance as the human head at the mastoid
bone. The velocity which the bone conduction receiver delivers to this
acoustical impedance is measured by a vertical or hill-and-dale phonograph
pickup (see Sec. 9.3B2d). A strain-gage type 74 of artificial mastoid has
been developed. The translating element is a ceramic transducer..
10.5. Testing of Phonographs.-A. Measurement of the Response of a
Phonograph Record by the Optical Method.-The response frequency charac
teristic of a lateral cut phonograph record may be obtained by means of
the optical method 75,76,77 as shown in the schematic diagram of Fig. 10.37.
The point source of light is placed at a distance of at least lO feet, so that the
light which strikes the grooves is practically parallel. The incident light
is reflected from the sides of the groove of the record. In the case of a
groove without modulation the width of the reflected light image will be a
fine narrow line. If a sine wave is cut in the groove the width of the image
will be proportional to the amplitude and the frequency. To make an
accurate measurement of the width of the image the distance of the observer
from the record should be at least lO feet. The distance b of Fig. lO.37A
can be measured accurately at this distance by employing a telescope.
Under these conditions, the amplitude of the modulation, in inches, is given
by
bu
xp = 2] lO.34
73 Romanow, F. F., Jour. Acous. Soc. Amer., Vol. 13, No.3, p. 294,1942.
74 Carlisle and Pearson, Jour. Acous. Soc. Amer., Vol. 23, No.3, p. 300, 1951.
77 Bauer, B. B., Jour. Acous. Soc. Amer., Vol. 18, No.2, p. 387, 1946.
, r
MEASUREMENTS 473
The peak velocity, in inches per second, is given by
xp = Trbu 10.35
The rms velocity, in inches per second, is given by
. Tr b 10.36
x = v'2 u
The image shown in Fig. 10.37B depicts a record with constant velocity
and any frequency. This shows that the radius of the groove or the fre
quency does not influence the width of the reflected image when the velocity
A EYE
TELESCOPE
LIGHT
f
10 fEET +
RECORD ~==;:::;:::=====> ~
ARRANGEMENT Of APPARATUS
~bf-.
A B c
FIG. 10.37. Schematic arrangement of the apparatus for the measure
ment of the response of a phonograph record by the optical method.
The light patterns for various types of modulation are shown in A, B,
and C. A shows a record with a modulated and unmodulated groove.
B shows a record with constant velocity and any frequency. C shows
a record cut with sine wave modulation having the response frequency
characteristic of Fig. 9.14. The frequency increases as a logarithmic
function from the inside to the outside. Note: the frequency pattern
at 180 0 is not shown.
is constant. The image in Fig. 1O.37C depicts a frequency record cut with
a sine wave modulation having the velocity response frequency character
istic as shown in Fig. 9.14.
It has been shown 78 that conventional light patterns are in error due to the
diffraction of light. The error is proportional to the frequency. It is of the
order of It decibels less than the conventional method at 20,000 cycles.
B. Testing of Phonograph Pickups.-Phonograph pickups are generally
78 Bauer, B. B., Jour. Acous. Soc. Amer., Vol. 27, No.3, p . 586, 1955.
474 ACO"CSTICAL ENGINEERING
INPUTI1r;::=~=::; 40
"'30
o /"
..........
w
~20
./"
MAGNET
oa.
(J)
w
/
a:: 10
V
oV
20 100 1000 10000
SECTIONAL VIEW FREQUENCY
FIG. 10.38. Electrical circuit diagram, a sectional view and the response fre
quency characteristic of a feedback type calibrator for obtaining the response
frequency characteristic of phonograph pickups.
81 Kellogg and Morgan. Jour. Acous. Soc. Amer.• Vol. 7. No.4. p. 271. 1936.
frequency at, for example, 1000 cycles. In reproducing this record, the
output is fed to a frequency discriminating network and detector similar
to that used in radio frequency modulation detectors. The magnitude of the
wow is the difference between the highest and lowest speed, in percentage of
the average speed.
10.7. Measurement of Acoustical Impedance. 82 ,83,84,85-There are a
number of methods of measuring acoustical impedance. A purely acoustical
means for measuring acoustical impedance has been devised by Stewart.
This method measures the change in acoustical transmission through a long
uniform tube when the unknown acoustical impedance is inserted as a branch.
The acoustical impedance bridge is shown schematically in Fig. 10.39.
The loudspeakers are. connected to two pipes, one of which is variable in
length and the other equipped with a means for attaching the unknown
impedance. At some distance beyond this point the two pipes are joined
by a small pipe which, in turn, is connected to another pipe leading to a
microphone and amplifier. Standing waves in the pipes are reduced by the
introduction of small tufts of felt.
With the branch closed the voltage across the two loudspeaker units
and the length of the variable tube are adjusted until a minimum reading
is obtained in the output meter. The unknown acoustical impedance is
now attached and the process repeated.
The unknown acoustical impedance can be obtained from the following
equation
h~ (cos B + j sin B) 10.37
where d = distance between the first position of the pointer without the
branch to the second position with the branch in place, in
centimeters. The direction toward the loudspeaker units
is positive and
,\ = wavelength of sound in air, in centimeters.
The recent emphasis on acoustical impedance as a means of describing
the sound absorbing properties of materials has brought about a renewed
interest in tube methods of acoustical impedance measurement. The
TELESCOPIC TUBING
i!l'!!"''''''''''!''''!!'' LOUD-SPEAKER
SCALE UNIT
87 Hall, W. M., Jour. Acous. Soc. Amer., Vol. 11, No. 1, p. 140, 1939.
88 Bolt and Brown, Jour. Acous. Soc. Amer., Vol. 12, No. 1, p. 31, 1940.
89 Sabine, H. J., Jour. Acous. Soc. Amer., Vol. 14, No.2, p. 143, 1942.
2K sin (2~1)
XAI = pc 10.42
1 + K2 + 2K cos e;~l)
where p = density of air, in grams per cubic centimeter,
c = velocity of sound in centimeters per second,
Dl = distance between the material and the first pressure minimum,
in centimeters,
D2 = distance between two adjacent minima or one-half wavelength,
in centimeters, and
K = pressure reflection coefficient obtained from equation 10.40.
LOUD SPEAKER
POWER
OSCILLATOR
r------, AMPLIfiER
SCALE
PROBE TUBE
PIPE
BAND
LOGARITHMIC
PASS AMPLIfiER
fiLTER DETECTOR
PRESSURE:
MICROPHONE
FIG. 10.40. Schematic arrangement of the apparatus for measuring the
acoustical impedance by means of the standing wave pattern in a tube ter
minated by the unknown acoustical impedance.
the mixed signal of the two reeds or the signal from the top reed alone. The
other meter reads the signal from the bottom reed alone. A powerful
driving system capable of handling 24 watts at the high frequencies with
the driving coil mounted on a rubber support is used. Considerably more
power is needed at the high frequencies because of the low compliance of
the reeds used at these frequencies.
A reed whose natural frequency is somewhere above 1.5 times the fre
quency at which the measurement is to be made is generally acceptable.
FRONT VIEW
VECTOR DIAGRAM
SECTIONAL VIEW
FIG. 10.41. Front and sectional views, mechanical network, and vector
diagram of a mechanical impedance bridge. In the mechanical network:
m = the effective mass of the reed. . eM = the compliance of the reed.
ZMX = the mechanical impedance being measured. 1M = the total
driving force. 1Ml = the driving force of the unloaded reed. 1M2 = the
driving force of the loaded reed and the mechanical impedance, ZMX.
1M3 = the driving force of the loaded reed. 1MX = the driving force
ofthe mechanicalimpedance, ZMX. Vo = the total velocity. VOl = the
velocity of the compliance of the unloaded reed. VMl = the velocity
of the mass of the unloaded reed. V02 = the velocity of the compliance
of the loaded reed. VM2 = the velocity of the mass and mechanical
impedance, ZMX. The vector diagram shows the magnitudes and
phases of the forces and velocities for a typical condition.
IZMX 1 =
1M2 - 1MI ( XMM2)
XMM - - - 10.45
1M2 XMC
XMC = - - , and
WCM
eM = compliance of the reed, given by equation 10.44.
If a vector diagram is desired, 1M2 may be obtained by turning a switch,
so the meter reads the signal from the loaded reed alone. From the values
of these three meter readings, the three forces may be plotted in their proper
phase relations, and a vector diagram drawn from which the mechanical
impedance may be calculated. For most measurements, the quick and easy
method of calculating the mechan
ical impedance by the formula PICKUP
and the ratio of the two forces will STEEL STYLUS ELECTRODE
give sufficient information. REED
Another method 91 for measuring
the mechanical impedance of phono
COIL
graph pickups and other devices
consists of measuring the motion of
a calibrated vibrating reed with and
without the load imposed by the SPACERS
pickup at the free end of the reed.
The calibrated reed is made of steel
and forms the arnlature of the electro
magnetic vibrator, see Fig. 10.42.
A set of four similar but of different
dimensions covers the frequency FIG. 10.42. Schematic view of the driver
range of 30 to 10,000 cycles. and pickup of a system for the measure
The alternating driving force ment of mechanical impedance.
resulting from the flow of electrical
current through the coil surrounding the reed is applied to the reed at the
upper pole faces near the free end of the reed. A capacity pickup electrode is
situated between the reed and one of the pole pieces. This electrode is main
tained at a d-c potential of about 115 volts. Motion of the reed causes a
change in capacitance between the reed and the electrode and thus generates
an alternating voltage component having the frequency and phase of the
91 Woodward and Halter, Jour. Acous. Soc. Amer., Vol. 25. No.2. p. 302, 1953.
482 ACOUSTICAL ENGINEERING
10.55
elO A..
- '1"'0
eo
where 4> is the phase difference between et and e and 4>0 is the phase difference
between eto and eo. In practice, the driving current for each reed is held
at some constant magnitude. Therefore, et = eto, and equation (10) may
be simplified to
ZML= ZMO (~E4> - 4>0 - 1) 10.56
Equation 10.56 is the working equation for the system. The values of eo,
e, ,p, and 4>0 are measured at each frequency of interest and are substituted
in equation 10.56 to compute ZML in complex form. The value of ZMO must
be known at each frequency. The converse procedure of applying known
loads to the reed and computing ZMO serves to calibrate the reed.
A block diagram of the electrical components of the mechanical imped
ance measuring system is shown in Fig. 10.44. A variable-gain amplifier
receiving its input signal from an audio-oscillator supplies the driving current
to the vibrator. The voltage drop across a resistor in series with the vibrator
coil is amplified, rectified, and used to control the gain of the driving amplifier.
4-84- ACOUSTICAL ENGINEERING
A~:~~~~~ n~~,n
ranging from the earliest acoustic
types to the most recent designs.
h, ; hz 10.9. Measurement of Porosity.
UU
._:_;_=
92 Rettinger, M., Jour. Acous. Soc. Amer., Vol. 6, No.3, p. 188, 1935.
93 Gemant, A., Jour. App. Phys., Vol. 12, No. 10, p. 725,1941.
94 Morse, Bolt and Brown, Jour. Acous. Soc. Amer., Vol. 12, No.2, p. 217, 1940.
95 Brown and Bolt, Jour. Acous. Soc. Amer., Vol. 13, No.4, p. 337, 1942.
96 Beranek, L., Jour. Acous. Soc. Amer., Vol. 13, No.3, p. 248, 1942.
97 Morse and Bolt, Rev. Mod. Phys., Vol. 16, No.2, p. 69,1944.
MEASUREMENTS 485
valve is closed and the free side of the U tube is elevated until the levels,
in centimeters, have changed from h to h1 on one side and h2 on the other
side of the U tube. The pressure change tlpo, in dynes per square centi
meter, equals (h2 - h1 ) 980. The reduction of volume tl V 0, in cubic centi
meters, in the chamber is (h1 - h)S, where 5 is the cross-sectional area
of the tube, in square centimeters. The porosity is given by
10.57
rADG = IJ 10.58
THROTTL
VALVE
CHAMBER
L ___ __________ J
,
r
HIGH D-C
OSCILLATOR VOL TAGE
AM PLiFfER
SUPPLY
PERSI5TEN T IMAGE
CATHOOE-RAY OSCILLOGRAPH
impossible, at the present time, to build a noise meter which will show the
true loudness level. The discrepancies can be determined by actual use
and suitable weighting factors applied to the results. Objective measure
ments are almost indispensable in any scientific investigation. The noise
meter or sound level meter provides a system for measuring the sound level
of a sound.
A schematic diagram of a sound level or noise meter is shown in Fig. 10.48.
The microphone should be calibrated in terms of a free wave. The direc
tional characteristics of the microphone should be independent of the fre
quency. The attenuator and meter should be calibrated in decibels. A
sound meter reading 60 db means a sound level of 60 db above the reference
level. The reference point of the decibel scale incorporated in a sound
meter shall be the reference sound intensity at 1000 cycles in a free pro
gressive wave-namely, 10-16 watts per square centimeter. The response
frequency characteristic of the human ear shows less sensitivity for fre
quencies above and below 3000 cycles, Fig. 12.11. The over-all frequency
response of an ideal noise meter should be the reciprocal of the ear response
frequency characteristics. This would make the noise meter unduly com
MEASUREMENTS 489
plicated. The response frequency characteristics recommended for the
noise meters by the American Standards Association are 109,110,111,112 shown
in Fig. 10.49. Curve A is recommended for measurements at the lower
levels and curve B for measurements around 70 db above the threshold.
For very loud sounds (80 to 100 db), the flat characteristic of curve C should
be used.
The noise meter may be used for noise analysis in offices, factories,
restaurants, etc. In these measurements a large number of observations
TOLERANCES<:= .... OESIGN OBJECTIVE
----
+10
-- f - - -- .
-10 ::::-:: ~
- ~
f
A -- ....
~ -20
#"
'"
"
;-
~I"
w
/
'"
-10
.......
_f-
-- B
-'
+10 -20 "
-r --- - -I- --- -- - --
or-- - ---- 1--- - . I-c-- --
;-
-IOZ O 100 1000 10000
fREQUENCY IN CYCLES PER SECOND
should be made in various positions. The noise meter may also be used to
measure the transmission or attenuation by walls, floor, ceilings, and doors
(see Sees. 10.13 and 11.2U).
The noise meter is also a useful tool in work on the quieting of machinery.
Since the radiation pattern of machinery noise is very complex, a large
number of measurements should be made in various directions relative to
some axis of the system. For these investigations a frequency analyzer
of the heterodyne type (see Sec. 1O.3C) is a useful adjunct for determining
109 Amer. Tent. Standards for Sound Level Meters Z24.3, American Standards
Association, New York, N. Y., 1936; or Jour. Acous. Soc. Amer., Vol. 8, No.2, 1936.
110 American Standard for Noise Measurement Z24.2-1942, or Jour. Acous. Soc.
Amer., Vol. 14, No. 1. p. 102, 1942. Also American Standards Association, Z24.2
1942.
111 American Standard for Sound Level Meters for Measurement of Noise and other
Sounds, Z24.3, American Standards Association, New York, N.Y., 1944. Also
American Standards Association, Z24.3-1944.
112 American Standards Association, "Test Code for Apparatus Noise Measurement,"
Z24.7-1949.
490 ACOUSTICAL ENGINEERING
I I
I I
I I
I I OUTPUT
~:: - :::~
OUTPUT
I I
I I
I
I 600 1200 f. I
~ "" -",,0
~~2400- 4800~~
VACUUM
TUBE
VOLTMETER
5r-----r-r-----r-----r-----r-r-----r-----.-----~
-10 I---+-t--t\--+-+-t--+-+t--+-+t--t+-+--++-+--+l---\---H
BAND PASS
FILTERS
OUTPUT
VACUUM
TUBE
VOLTMETER
of the source, if the source impedance is small compared to the input imped
ance or constant over the audio-frequency range. The output of the filter
selecting switch is connected to a calibrated attenuator. The output of the
attenuator is connected to an amplifier. The output of the amplifier is
used to drive the indicating meter and the monitoring output. The response
frequency characteristics of the filters are shown in Fig. 10.51.
113 American Standards Association, "Octave-Band Filter Sets," Z24.10-1953.
114 Beranek, L., .. Acoustic Measurements," John Wiley and Sons, New York, N.Y.,
1949.
492 ACOUSTICAL ENGINEERING
o
{
\
...
(/)
2
a' -20
(/)
\
'"0:
-30 III \\
-40
20
I ,"; Y
40 80100
\,
200
\~ X~
400 1000
XX \
2000
FREQUENCY IN CYCLES PER SECOND
~ :~ X
4000 10000
\\
20000
FIG. 10.53. Response frequency characteristics of the band-pass filters used in the
audio spectrometer of Fig. 10.52.
CONNECTION DIAGRAM
MAGNETIC
TAPE
REEL
,.~
~
"'--
5 -
- RECORD
MOTOR
e: -----
'" - RECORDING
PAPER ON
TIME DRUM
AMPLIFIER
AND
DETECTOR
AMPLIfiER
AND
DETECTOR
FIG. 10.57. Schematic arrangement of the apparatus for obtaining the response
frequency characteristic of a hearing aid.
SPECIMEN
SECONDARY
WINDIIIIG
AC
Ar------.......- - - L_ _-----l
CASE
SECTIONAL VIEW
FIG. 10.59. Schematic arrangement of the apparatus for measuring the rough
ness of a surface, showing a sectional view of the mechano-electronic transducer,
the viscous coupler, and the electrical diagram.
Later work13 has shown that equation 11.2 is unsatisfactory for large
rooms or rooms with very large absorption. The equation developed by
Eyring is
T = .05V 11.3
- Slog. (1 -- aav)
where V = volume, in cubic feet,
S = total area, in square feet, and
aav = average absorption per square foot, in sabins.
Frequency
Thick- Mount-
Material ness
(in.)
ing 128 1 256 I 512 110241204814096 Author
Coefficient
-~ ~---
......
Acoustex 60R 1 2 .14 .28 .81 .94 .83 .80
Fiberglass Tile 1 2 .22 .46 .97 .90 .68 .52
Acoustone F. II 1 .16 .31 .87 .92 .83 .87
Acousti-Celotex C-4 .
Acousteel Pad with
16
Abbreviations in the above table are as follows: A.M.A., American Materials Asso
ciation; ·W.S., Wallace Sabine; P.S., P. E. Sabine; F. W., F. R. Watson; V.K., V. O.
Knudsen; B.R., Building Research Station, England; B.S., U.S. Bureau of Standards.
Mountings in the above table are as follows:
1. Cemented to plaster board.
2. Nailed to 1" X 2" furring 12" O.C.
3. Attached to metal supports applied to 1 H X 2" wood furring.
4. Laid on 24 gao sheet iron, nailed to I' X 2" wood furring 24H O.C.
ARCHITECTURAL ACOUSTICS 503
Frequency
Thick- Mount-
Material ness
(in.)
ing 128 I 256 I 512 110241204814096 Author
Coefficient
Brick Wall, Un
painted. .024 .025 .031 .042 .049 .070 W.S.
Brick Wall, Painted. .012 .013 .017 .020 .023 .025 W.S.
Concrete Porous
Block, Set in 1:3
Cement, Sand,
Mortar. 2 .15 .21 .43 .37 .39 .51 B.R.
Plaster, Lime on
Wood Lath on
Wood Studs, Rough
Finish. ! .039 .056 .061 .089 .054 .070 P.S.
Plaster, Gypsum on
Wood Lath on
Wood Studs, Rough
Finish. ! .023 .039 .039 .052 .037 .035 P.S.
Ozite t .09 .19 .28 .51 .56 .47 P.S.
_.
Individual Object Absorption Units in sq. ft. (Sabins)
Audience, per Person,
Man with Coat 2.3 3.2 4.8 6.2 7.6 7.0 B.S.
Auditorium, Chairs,
Solid Seat and
Back .15 .22 .25 .28 . 50 P.S.
Auditorium Chairs,
Upholstered 3.1 3.0 3.2 3.4 I F.W.
Abbreviations in the above table are as follows: A.M.A., American Materials Asso
ciation; W.S., Wallace Sabine; P.S., P. E. Sabine; F.W., F. R. Watson; V.K., V. O.
Knudsen; B.R., Building Research Station, England; B.S., U.S. Bureau of Standards.
Mountings in the above table are as follows:
1. Cemented to plaster board.
2. Nailed to 1 H X 2H furring 12H O.C.
3. Attached to metal supports applied to 1 # X 2# wood furring.
4. Laid on 24 gao sheet iron, nailed to 1 N X 2" wood furring 24' O.C.
15 Gemant, A., Jour. App. Phys., Vol. 12, No. 10, p. 725,1941.
16 Morse, Bolt, and Brown, Jour. Acous. Soc. Amer., Vol. 12, No.2, p. 217, 1940.
17 Brown and Bolt, Jour Acous. Soc. Amer., Vol. 13, No.4, p .. 337,1942.
18 Beranek, L., Jour. Acous. Soc. Amer., Vol. 13, No.3, p. 248,1942.
19 Morse and Bolt, Rev. Mod. Phys., Vol. 16, No.2, p. 69,1944.
where rAID = d-c acoustical resistance of the material per unit cube, in
acoustical ohms,
d = thickness of the material, in centimeters,
p = density of air, in grams per cubic centimeter,
m = ratio of the effective density of the air in the pores to its
density in the open,
P = porosity, the ratio of the volume of air in the pores to the
total volume,
c = velocity of sound, in centimeters per second,
w = 271"/, and
/ = frequency, in cycles per second.
In Fig. 11.1, the acoustical resistance, in acoustical ohms, per unit area is
rAlrft
rAI = -3- 11.5
11.6
fA M
~A
AT
ACOUSTICAL CIRCUIT
SECTION
300
-<...
a:
-<
I-
Z
200
100
1----- .!A!..
-- -- -- - --- -~
:>
a: 0
w
Il.
w -100
U ...... - ....i-"
.~ -- ~-
100
I
/
200 500 1000 2000 5000
fREQUENCY IN CYCLES PER SECOND
FIG. 11.2. The acoustical resistance, XAlo and acoustical reactance, r Alo
characteristics of Permacoustic 1 inch in thickness and backed by a rigid
wall. Dashed lines: experimentally measured characteristics. Solid lines:
computed characteristics. The ordinates represent the acoustical im
pedance per square centimeter.
506 ACOUSTICAL ENGINEERING
a=l- I
ZAI -
ZAI + pc
PCl2 1l.8
1.0
t
Z
-
w
g.a
....... ~
~
::5 .6
IJ
/ 1-----
z
Q:4
./ .... _I--f
~ ""
t
o.
a: ",'"
Sl·2 ,
m
« _.... ' " ..
100 200 500 1000 2000 5000
fREQUENCY IN CYCLES PER SECOND
region in which the dimensions of the absorber are small compared to the
wavelength.
The functional sound absorber is made in the form of a thin shell of
acoustical absorbing material. The magnitude of the acoustical resistance
of the shell is selected to yield the highest absorption of sound.
The absorption coefficient frequency characteristic of a typical functional
sound absorber, obtained from reverberation chamber measurements, is
shown in Fig. l1.4C. It will be seen that the absorption per unit area is
about two times that of conventional materials. The shell-type functional
~~~~~~~~A~B~SORBING
= ~ MATERIAL rAG
t.: 1,6
..."-
8 1•2 r --... i"'..
../
~
- rAG~::ZA z~ ~ .8
.
i=
~~~~~~ABSORBER
= rAG t.4c
--~GMG
ZA7L;J
ACOUSTICAL CIRCUIT
INCIDENT SOUND 8 D
PERSPECTIVE VIEW
sound absorber is very economical in the use of material, because the sound
absorption, in sabins per pound, is about twelve times that of conventional
absorbing materials.
The functional sound absorber is very easy to install because it is merely
suspended on wires. A typical installation is shown in Fig. lOAD. This
is a truss type roof where the installation of conventional materials would
be very costly. In the case of skylights, conventional materials, installed
as a false ceiling, of course, impair the lighting. On the other hand, the
functional sound absorber will not impair skylighting. It is particularly
useful in industrial applications where the principal objective is to absorb
sound and appearance is not a factor.
ARCHITECTURAL ACOUSTICS 509
D. Resonator Sound Absorber. 26 ,27,28-The resonator sound absorber
consists of a simple resonant acoustical system. One type consists of a
Helmholtz resonator with some absorbing material located in the cavity of
the resonator, or with a cloth over the mouth to provide additional acoustical
resistance. The acoustical performance is practically the same for all types
of simple resonators. Therefore, a consideration of the performance of the
Helmholtz type will indicate the characteristics of resonant sound absorbers.
A sectional view and acoustical network of a Helmholtz resonator are
shown in Fig. 11.5. The performance of the system can be determined from
the acoustical network and the constants
of the system. In one application, the
problem is to reduce the sound pressure
over a small volume. Under these
conditions, the acoustical resistance r A2
should be made as small as possible. A
typical measured sound-pressure frequency
SECTIONAL VIEW
characteristic at the mouth of the reso
nator is shown in Fig. 11.6. The charac
teristic shows that there is a reduction in
sound pressure in the frequency range
above 72 cycles and an increase in sound
pressure in the frequency range below 72 ACOUSTICAL CIRCUIT
cycles. Since there is an increase of pres FIG. 11.5. Sectional view and
sure over a certain portion of the frequency acoustical network of a Helmholtz
range, it is obvious that the simple reso resonator sound absorber. In the
acoustical network, p = the sound
nator is not suitable for a free-field, spot pressure in free space. M 1 and
type sound reducer. l' A1 = the inertance and acoustical
When the resonator is used as a sound resistance ofthe air load. M 2 = the
absorber, the action is somewhat different. inertance of the aperture. CAl
In order to obtain the maximum sound = the acoustical capacitance of the
volume. r.A2 = the acoustical resist
absorption, the acoustical resistance rAl ance of the cloth over the opening
should be equal to the acoustical resist of the resonator. P2 = the sound
ance, r A2. This can be accomplished by pressure at the mouth.
selecting the proper value of the acoustical
resistance r A2 provided by the cloth over the opening of the resonator. If a
bank of resonators is used, so that the ultimate acoustical resistance rAl is
obtained, the maximum efficiency of sound absorption will be obtained. A
typical sound-absorption frequency characteristic of a Helmholtz resonator
is shown in Fig. 11.7. An examination shows that a high value of sound
absorption is obtained over only a very narrow frequency range. At least
twenty resonators would be required to cover the frequency range from
26 Harris and Malloy, Jour. Acous. Soc. Amer.• Vol. 24, No. 1. p. 1. 1952. This paper
contains references to publications of other investigators on the subject of resonator
sound absorbers.
27 Ingard. Uno, Jour. Acous. Soc. Amer.• Vol. 25. No.6. p. 1037. 1953.
28 Olson and May. Jour. Acous. Soc. Amer.• Vol. 24. No.6, p. lBO, 1953.
10
.,o
'!;
'"'"z
o
Q.
'"'"
0:
o
.~
-L/ V
/
-10
·I~
20 30 40 60 80 100 200 300 400 600 800 10000
FREQUENCY IN CYCLES PER SECOND
30 to 200 cycles with tolerably good absorption. Since the average cubical
content of each resonator is two cubic feet, the total cubical content would
be 40 cubic feet. Since a bank of resonators is required for each frequency,
the entire assembly becomes an arrangement of tremendous bulk.
1.0
.8
(\
.
I
Z
U
j;:
\
~ ,6
o
\
u
z
o II
~ ~
0:
.,.,
o
C
.2
J \,
30 40 60 80 100 200 300 400 600 800 10000
FREQUENCY IN CYCLES PER SECOND
e~
Conventional sound absorbing mate
rials with high efficiency in the low
frequency range are extremely bulky. ELECTRONIC
SOUND
Therefore, conventional sound ab ABSORBER
sorbing systems are unsuitable where
compact high-efficiency sound absorbing
systems are required. When the elec
tronic sound absorber is commercialized
it will provide a compact system of
high-absorbing efficiency in the low
frequency range.
The electronic sound absorber may FIG. 11.8. Three electronic sound
be used as a diffraction sound absorber absorbers located in the corner of a
as shown in Fig. 11.8. In order to room.
obtain a high acoustical resistance load,
the absorbers should be mounted in the corner at the intersection of the
three surfaces, as shown in Fig. 11.8. As contrasted to the resonant sound
absorber with a narrow frequency range of sound absorption, the electronic
sound absorber exhibits high efficiency over several octaves in the low
frequency portion of the audio-frequency range.
F. Articulation and Reverberation Time.-The articulation 3o (see Sec.
10.16) of un amplified speech in auditoriums of various sizes as a function
of the reverberation time is shown in Fig. 11.9. The average power of
unamplified speech is much less than that required for distinct hearing.
A greater reverberation time increases the intensity of sound at the auditor.
However, increased reverberation decreases the intelligibility of speech.
These two factors oppose each other with the result that there is an optimum
reverberation for each auditorium which yields maximum articulation as
shown in Fig. 11.9. The obvious solution is the use of sound reinforcing
equipment. The articulation for a weak, average, and loud talker without
amplification as compared to amplified speech is shown in Fig. 11.9. By
proper selection and placement of the loudspeakers the articulation charac
teristic may be made considerably higher.
G. Sound Motion Picture Reproducing Systems. 31-The resultant sound
29 Olson and May, Jour. Acous. Soc. Amer., Vol. 25, No.6, p. 1130, 1953.
30 Knudsen, V. 0., Jour. Acous. Soc. Amer., Vol. 9, No.3, p. 175, 1938.
31 Olson and Massa. Jour. Soc. Mot. Pict. Eng., Vol. 23, No.2. p. 63, 1934.
energy density at the absorption of the auditor in a theater depends upon the
response and the directional characteristics of the loudspeaker and upon
the reverberation characteristics of the theater. From the standpoint of
the auditor, it may be said that there are two sources of sound energy
naII,lely, the direct sound, which travels directly from the loudspeaker to
the auditor; and the generally reflected sound, which is reflected from the
boundaries before reaching the auditor.
In a theater, free from acoustical difficulties, the energy density of the
generally reflected sound is practically the same for all parts of the theater.
100 100
..
Z 90
w
CURVE VOLUME
CU,FT. cu..... ..
Z 90 A A AMPLlriED SPEECH
w
u
~
A u B LOUD SPEAKER
~~
w C'-
a.
D
a. o WEAK SPEAKER
:!: 70 E :!: 70
Z z
r/ r--..... ~ ~
..
Q 60 06 0
-<
550
u
i=
'" 40
~
550
u
i=
a: 40
"'" ~ ~
-.......:::::
~
-< -<
300123467830012345678
X REVERBERATION TIME IN SECONDS Y REVERBERATION TIME IN SECONDS
P =Po~
r
11.10
>."~
~~60SSE
- ...I
=. :::4
D
POSITION ON
C
LISTENING
B
at this frequency. The direct radiation from the loudspeaker can then
be obtained for any point in the space.
The energy density, ergs per cubic centimeter, due to direct radiation
from the loudspeaker is
Po2Xo2Re2
ED = =--=-.......=-....,...:
2
11.11
r2pc
To analyze the distribution of the direct sound over the area, the plan
view of the theater and the directional characteristics of the reproducer in
the horizontal plane must be considered. The angle subtended at the loud
speaker by the area to be covered will determine the effective dispersion
angle of the reproducer.
The sound energy density due to the generally reflected sound is a func
tion of the absorption characteristics of the theater and the power output
of the reproducer. The sound energy density, ergs per cubic centimeter,
due to the generally reflected sound is given by
ER = -
4P [1 - (1 - a)
€(cS[lOg. (1-a)]t)(4V] 11.12
caS
where a = the average absorption per unit area, absorption coefficient,
S = the area of the absorbing materials, in square centimeters,
V = the volume of the room, in cubic centimeters,
t= time, in seconds,
c= the velocity of sound, in centimeters per second, and
P = the power output of the loudspeaker, in ergs per second.
The total sound energy density at any point in the theater will be the
sum of the direct and the generally reflected sound, and may be expressed by
Ep = ED + ER 11.13
A method has been outlined above, employing directional loudspeakers
for obtaining a uniform energy distribution of the direct sound. The
energy density of reflected sound, as shown by equation 11.12, is inde
pendent of the observation point. As a consequence, by employing direc..
tional loudspeakers, the total sound energy density will be the same in
all parts of the theater. Furthermore, the effective reverberation of the
reproduced sound (the ratio of generally reflected to direct sound) is the
same for all parts of the theater.
The distribution of a reproducing system in a theater is usually checked
by means of a response measuring system. The plan and elevation view
of a typical theater are shown in Fig. 11.11. The response characteristics
for positions A, B, C, and D'on the orchestra level and E and F on the
balcony level are shown in the respective graphs. These characteristics
show that it is possible to obtain uniform response in all parts of the theater
by following the procedures outlined in the preceding discussions.
In sound motion picture reproduction, the loudspeakers are usually
placed behind a perforated screen upon which the picture is projected. Fig.
11.11. Sound is transmitted through the screen by actual vibration of the
screen and by the perforations. In general the transmission by vibration,
is negligible. The perforations usually consist of small circular holes about
a millimeter in diameter. These holes form an inertance and acoustical
resistance (see Sees. 5.3 and 5.15). The acoustical resistance of the holes
introduces attenuation which is usually small. The acoustical reactance
ARCHITECTURAL ACOUSTICS SlS
due to the inertance increases with frequency, and therefore the attenuation
increases with frequency. The response frequency characteristic of the
screen shows more or less constant attenuation in the low- and mid-fre
quency ranges due to the acoustical resistance of the holes. However, the
attenuation in the high-frequency range increases with frequency due to
the acoustical reactance of the holes. The inertance increases with the
thickness of the screen and decreases as the ratio of the open to closed area
of the screen increases. For example, for 3 db attenuation at 10,000 cycles
~A
0 40
520
~
:g5B
the hole area is usually 15 to 20 per cent of the screen area. If the hole
""OEm 30
10
~30
~20
5;
0IE
40
10
~ ~o IOOOO~
D
100 1000 °50 100 1000 10000
FREQUENCY FREQUENCY
0
020 020
a. a.
eI) 10 en 10
'"D:: 050 100 1000
...
10000a:: °50 'DO 1000 10000
FREQUENCY FREQUENCY
"
0 40
30 '" c1)30
40
Z Z
020 020
a. a.
III 10 cI) 10
LOUDSPEAKERS
_________
SCREENf
ELEVATION VIEW
FIG. 11.12. A cross-sectional and plan view of a theater equipped with three loud
speakers behind the screen for stereophonic sound reproduction of sound motion
pictures. Loudspeakers for special effects are located on the sides of the theater.
L.OUD
-.-----
ORIGINAL SOUND
ORCHESTRA
r E D rLOOR
C
B
:tY RCE
STAGE
ORCHESTRA
I .L-J o_~~~t I
E D
POSITION ON
C
LISTENING
B
LEVEL
A 0
bination of the original source and the loudspeakers, is the same for all
parts of the audience.
The energy density, in ergs per cubic centimeter, at a distance r centi
meters due to direct radiation from the loudspeaker, from equation 11.11, is
P02xo2R82
ED2 = r2pc 2 11.15
where po = pressure, in dynes per square centimeter, obtained at a distance
Xo centimeters,
R8 = ratio of the pressure at the angle 0 to 0 = 0,
ER = 4(PDl +
P D2 ) [1 _ €(cS[log. (I-a)] t)14V] (1 - a) 11.16
caS
where a = average absorption per unit area, absorption coefficient,
S = area of absorbing materials, in square centimeters,
V = volume of the theater, in cubic centimeters,
t = time, in seconds,
c= velocity of sound, in centimeters per second,
PDl = power output of the original sound, in ergs per second, and
P D2 = power output of the loudspeaker, in ergs per second.
The aid obtained from reflected sound in a directional sound reinforcing
system is relatively small, ranging from 2 to 6 db.
The preceding considerations have been concerned with a sound rein
forcing system for a large theater or auditorium. In the system of Fig.
11.13 a single loudspeaker station is employed. This is sometimes termed
a high-level sound system. This is in contrast to a sound system employing
a large number of loudspeakers distributed over the area to be covered.
The latter system is termed a low-level sound system. Examples of these
two types of sound reinforcing systems 34 are depicted in Fig. 11.14. In
Fig. 11.14A the loudspeakers are located in a cluster above the stage. In
Fig. l1.14B the loudspeakers are distributed around the walls of the room.
Either system may be used with equally good results. The single loud
speaker station located above the original sound source produces somewhat
more realistic results since the original and augmented originate in the
same general direction. In the system of distributed loudspeakers, it is
somewhat easier to obtain good results because the feedback difficulties
are reduced. Furthermore, the problem of uniform distribution of the sound
34 Olson, "Musical Engineering," McGraw-Hill Book Company, New York, N.Y.,
1952.
ARCHITECTURAL ACOUSTICS 519
is simpler. The latter arrangement is almost universally used for school
rooms and churches.
In a sound reinforcing system of the type shown in Fig. 11.14B for most
locations in the room, the sound will not appear to be from the source on
the stage, but from the loudspeaker nearest to the listener. This is due to
the delay effect described in Sec. 12.18. This disconcerting condition can
MICROPHONE
AUDIENCE
In STAGE
MICROPHONE
AUDIENCE
I~
B
FIG. 11.14. Arrangements for a sound-reinforcing system in
a large room, as for example, a schoolroom, a church, an audi
torium, etc. A. A horn loudspeaker cluster located above the
stage. B. A large number of direct-radiator loudspeakers
distributed around the walls of the room.
6
LOUDSPEAKERS
7 5 3
SCHEMATIC DIAGRAM OF SYSTEM
2
3
MICROPHONE
I
LOUDSPEAKER F-'-'----I
ELEVATION VIEW
LOUOSPEAKERS
~
IT] IT]
6 4
MlcRoPHONE-l
LOUDSPEAKER I {J ~
LOUDSPEAKERS
rn7 rn
5
rn3
PLAN VIEW
/
/
LOUOSPEAKER
\
, LOUOSPEAKER
\ / /
~ ~
. . -l\"DIRECTIONAL
CHARACTERISTiC
"---'C DIRECTIONAL
CHARACTERISTiC
MICROPHONE
PLANE OF ZERO
RESPONSE
STAGE
DIRECTIONAL
CHARACTERISTIC
'"6 ~'"
OIRECTiONAL :J
MICROPHONE~
SHIELD
susceptible to the augmented sound and one which also furnishes a good
view of the action.
1. Theater Acoustics. 37 -Many theaters have major acoustical defects
which cause echoes and objectionable concentrations due to focusing of the
reflected sound. These reflections may be more important than the rever
beration time.
When a sound wave strikes a wall of a theater, a part is reflected, a part
absorbed, and a part transmitted. The reflection, for surfaces large com
pared to the wavelength, is analogous to specular reflection. The reflected
sound, in a poorly designed theater, produces highly concentrated zones of
reflected energy. For proper sound reflection control in an auditorium the
acoustical treatment and shape of the walls and ceiling must be such as to
thoroughly diffuse the reflected sound. The reflected sound energy received
in any auditorium location should not come from one particular reflecting
37 Standardization Committee, Jour. Soc. Mot. Pic. Eng., Vol. 36, No.3, p. 267,1941.
522 ACOeSTICAL ENGINEERING
, ,/
,,
,
,, ,,
,
,,
,
, , ,
,
,
"
\, \, '.J
'.J
:<
Z , , ;~
...'"0: :
., ,, ,, ,, ."
0
, , '0
,,
,,
!
..
,,
! WALL
! !
''''
:ID
,/
SIDE,'
.' ./
PLAN VIEW
CEtLING
FLOOR
ELEVATION VIEW
i~lrrffUl]
~ ~
I I Ililill]
~
FREQUENCY IN CYCLES PER SECOND
~~
2.0
Ul
~1.8
o
~1.6
Ul I-'
/"
1!: 1.4
V
~1.2
t.- ~
....
j: 1.0
z ~
g .8 i..---'
...10
15'" .6
ID
0:
~ .4
....
0: .2
o
1000 10000 100000 1000000
VOLUME IN CUBIC FEET
FIG. 11.18. Lower graph shows the optimum reverberation time for a
theater as a function of the volume for 1000 cycles. Upper graph shows
the relation between the reverberation time and the frequency, that is, the
reverberation time at other frequencies is obtained by multiplying by K.
frequency of 1000 cycles, is shown in the lower graph of Fig. 11.18. The
reverberation time for other frequencies can be obtained by multiplying by
the factor K. obtained from the upper graph of Fig. 11.18. The reverbera
tion time increases at the lower and higher frequencies so that the aural
rate of decay of pure tones will be approximately the same for all frequen
cies (see Secs. 12.6 and 12.7 and Figs. 12.11 and 12.13).
524 ACOUSTICAL ENGINEERING
100
,(/!
09-'
10
V- I/
Q~
0,0
V 17
I /
<:J~
/ ,,0 t'
V V"
I /
~
/ '\,1'
V V
.0I
1
/
V
I
10"
VOLUME IN CUBIC FEET
95
til / 'f V
~ 90 / J
~
u
3 80
/ V J
/
...070 / / /
!zw 60
U 50 / I /
a:
~40
/ II /
1/ J
/
~ 30
~
520 / / /
/
::;
::>
10 / I
/ /~
/
/35 40
J
45 50 55 60 65 70 75 80 85 90 95 100
SOUND LEVEL IN DECIBELS
40 Muller, W. A., Jour. Soc. Mot. Pic. Eng., Vol. 35, No.1, p. 49,1940.
I~.~~~M~IC:R:O~PH~O:N:E~~~~~
I/~)
LOUD!
PLAN VIEW
-SPEAKER
ELEVATION VIEW
FIG. 11.21. Two arrangements of sound systems for addressing assemblages in large
grandstands. For the stadium on the left a large number of loudspeakers are used, each
loudspeaker covering a small portion of the total area. For the ball park on the right,
a single loudspeaker cluster is used to supply the entire grandstand.
with characteristics which will deliver the required sound level over the
distances and areas considered. The steps in the selection of a system
may be as follows: First, the directional characteristics should be deter
mined, as outlined in the preceding discussion, so that uniform response is
obtained over the audience area. Second, either a single or a group of
loudspeakers having the desired directional characteristics should be selected.
Third, the response characteristic of the system on the axis at a specified
input and distance should be available to show the amplifier power required
to supply the desired intensity level. Fourth, the power handling capacity
of the loudspeakers and amplifiers should be adequate to supply the required
intensity level.
Two types of sound reinforcing installations for an outdoor theater are
shown in Fig. 11.22. The system on the left employs a single loudspeaker
LOUD-SPEAKERS
~AGE
r--~ AUDIENCE
AUDIENCE
lid
AREA lid lid
"'"
- '--
AREA
0 lid
LOUD "" "'" "'"
SPEAKER '-------:~-:-:-""C:'C::::c_-----'
• PLAN VIEW PLAN VIEW
station located either above or below the stage as shown in Fig. 11.22.
If the stage is quite low the logical position for the loudspeakers is at the
top of the stage. The same procedure for obtaining uniform sound cover
age and adequate intensity level of the sound from the loudspeakers as
used in the preceding considerations is applicable in this case. If the stage
is very high the separation between the loudspeakers and the action on
the stage will be very large. As a consequence, the wide difference in the
direction of the direct and reinforced sound will be particularly discon
certing to listeners in the front portion of the seating area. Under these
conditions, it may be desirable to locate the loudspeaker under the stage
as shown in Fig. 11.22. The system depicted on the right employs a large
number of loudspeakers, each one supplying a small portion of the audience.
ARCHITECTURAL ACOUSTICS 529
The directional characteristics of the loudspeakers should be selected so
that each individual area is adequately supplied. Cognizance must be
taken of the energy supplied from adjacent loudspeakers.
There are certain advantages in each system. In the case of the single
loudspeaker system, better illusion is obtained because the augmented
sound appears to come from the stage. On the other hand, the intensity
level outside the audience area in a backward direction falls off very slowly.
At a distance equal to the length of the audience area the level is only 6 db
lower than that existing in the audience area. In certain locations the
sound levels produced by such systems will cause considerable annoyance
to those located in the vicinity of the theater. By dividing the theater
area into small plots, each supplied by a loudspeaker, and by directing the
loudspeakers downward, the sound intensity level outside the audience
area will be considerably lower than in the case of the single loudspeaker
station and usually eliminates any annoyance difficulties. The short
sound projection distance is another advantage of the multiple loudspeaker
system.
The above typical examples of outdoor public address and sound re
inforcing systems illustrate the principal factors involved in this field of
sound reproduction.
N. Sound Motion-Picture "Drive-In" Theater.-A "drive-in" motion
picture theater is an outdoor system in which the audience attends the
show while seated in an automobile. A type of sound motion-picture
"drive-in" theater system is shown in Fig. 11.23. Sound motion-picture
LOUD SPtAKE:R
CABLE MECHANISM
BOOTH
FIG. 11.23. A perspective view showing the elements of a "drive in" sound
motion-picture theater. Perspective and sectional views of the loudspeaker used
in the" drive in" theater are shown in the lower right. (After Olson, .. Musical
Engineering," McGraw-Hill Book Company, New York, N.Y., 1952.)
530 ACOUSTICAL ENGINEERING
reproducing systems are described in Sec. 13.7. The photoelectric cell
and optical system are housed in the picture projector of Fig. 11.23.
The picture projector amplifier and master volume controls are located in
the projection booth. In the drive-in theater, the customers are seated
in their automobiles to view the picture and hear the sound. Individual
loudspeakers are provided to supply the sound to each automobile. Per
spective and sectional views of the loudspeaker are shown in Fig. 11.23.
In use, the loudspeaker is hung inside the automobile. The volume control
incorporated with the loudspeaker makes it possible for the listener to
adjust the level of the reproduced sound. When not in use the loudspeakers
are hung on posts provided for the purpose. Signal lines connect the indi
vidual loudspeakers to the amplifier in the booth.
O. Orchestra and Stage Shell.-When orchestra and stage productions are
conducted in outdoor theaters it is desirable to provide a shell to augment
III Iii
iill!! 'II
---------
FIG. 11.24. Perspective and elevation views of an orchestra shell with poly
cylindrical surfaces.
and direct the sound to the audience, to surround the orchestra with
reflecting surfaces and to protect the performers and instruments against
wind, dew, and other undesirable atmospherics. Most of the outdoor
orchestra shells have been of the concave type which produce intense and
sharp concentrations d reflected sound in both the shell and audience area.
These acoustical effects are particularly undesirable when the sound is
picked up by microphones on the stage for sound reinforcing and broad
casting. Under these conditions the intensifications and discriminations
make it appear that the orchestra is unbalanced with relation to the variou,>
instruments. Furthermore, it is impossible for the conductor to obtain
a true balance because these undesirable acoustical effects also exist at
the conductor's platform on the stage. The undesirable acoustical effects
can be overcome by means of an orchestra shell in which the boundaries
are polycylindrical surfaces as shown in Fig. 11.24. These surfaces reflect
the sound in a diffuse manner and thereby obviate concentrations of sound
energy on the stage and in the audience area. The acoustics of this type
ARCHITECTURAL ACOUSTICS 531
of orchestra shell make it possible for the conductor to obtain an improved
balance of the orchestra. This type of shell produces a uniform distribution
of sound in the audience area. The polycylindrical shell provides good
acoustics for microphone pickup for sound reinforcing or broadcasting.
P. General Announce and Paging Systems. 42 -General announce systems
are useful in factories, warehouses, railroad stations, airport terminals, etc.
A typical installation is depicted on the left portion of Fig. 11.25. For this
type of work intelligibility is more important than quality. The deleterious
effect of reverberation upon articulation can be reduced, and a better
control of sound distribution can be obtained, by reducing the low-frequency
~li]liIJliJJ
---a;i
I I I I I I
I
I I I I I I
--Q::ll)
iLiJJlITI
~[I]rn
(J:lD
LOUDSPEAKER
I I I I I I
MICROPHONE MICROPHONE
IQI
Jr;~""
PHONOGRAPH
I~ !. LOUDSPEAKER
~NSOLE
.. AND ~ AND
RADIO
~ "CONTROLS
CONTROLS
FIG. 11.25. Two uses of call, general announce, and sound distributing systems.
On the left, a high efficiency horn loudspeaker is used to obtain a high sound level
over a large floor area as in a factory or warehouse. On the right, small, direct
radiator loudspeakers are used to supply the small rooms at a relatively low level
as in paging, announcing, and centralized radio installations used in hospitals,
hotels, or schools.
response of the system. Furthermore, the cost of the amplifiers and loud
speakers is also reduced by limiting the frequency range. To find the power
required, the sound intensity level under actual operating conditions should
be determined. The system should be designed to produce an intensity
level 20 to 40 db above the general noise level. Under no conditions should
the system be designed to deliver an intensity level of less than 80 db. The
loudspeakers should be selected and arranged following an analysis similar
to that outlined in the preceding sections, so that uniform sound distribution
and adequate intensity levels are obtained.
For certain types of general announce, paging, and sound distributing
installations used in schools, hospitals, department stores, hotels, etc.,
the intensity level required is relatively low and the volume of the average
room is usually small. For most installations of this type, save in noisy
42 Olson, H. F., RCA Review, Vol. 1, No.!, p. 49,1936.
532 ACOUSTICAL ENGINEERING
rl=l--::::-L---~I~S::.;WITCH MICROPHONE I
LOUD-SPEAKER I
I
I
I
1- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ I
SUBSTATIONS MASTER STATION
more elaborate systems are similar to the general announce system described
in Sec. 11.2P with the addition of microphone positions in more than one
room. The simplest system consists of two units for use between two
stations. The master unit contains an amplifier, microphone, loudspeaker,
and a talk-listen switch. The remote unit consists of a microphone, loud
speaker, and talk-listen switch. In the simplest systems the loudspeaker
with suitable electrical compensation is also used as a microphone as shown
in Fig. 11.26. Additional stations and appropriate switching systems may
be added for communicating between a number of rooms. The voice
currents are carried in two ways: in one by direct wire, and in the other
by using a high-frequency carrier on the power mains. The latter system
does not require wiring but has the disadvantage that, in large buildings
having several separate systems, cross-talk may occur.
ARCHITECTURAL ACOUSTICS 533
R. Radio Receiver Operating in a Living Room.-The radio receiver and
phonograph represent by far the largest number of complete reproducing
systems. For this reason, the performance of a radio receiver in a room
is an extremely important problem. Equations 11.10, 11.11, and 11.12 for
the direct and reflected sound are applicable to a radio receiver in a room.
In the case of a theater it is possible to adjust the loudspeakers so that the
direct sound is the same in all parts of the auditorium. It is not practical
to arrange the loudspeakers in a radio receiver so that there will be no
variation of the direct sound with distance. In view of the rather small
distances and relatively small volume of the room this is not very important.
It is important, however, that the directional characteristic be independent
of the frequency and sufficiently broad to send direct sound into all listening
areas.
The response frequency characteristics of a good radio receiver taken
Ui1
at various listening positions in a typical living room are shown in Fig.
~sogA
W
<n30
<40
:gw40
cn30
sO B
~~ ~~
",som
~IO ~IO
~ °50 100 1000 10000 ~ °50 100 1000 10000
FREQUENCY FREQUENCY
d' POSITlONS\A
MICROPHONE
RADIO
RECEIVER
o~
w
~M
Z
C
o~
"
w 'SOleD
~~
Z
o~ o~
~ ~
(/)10 cniO
w w
PLAN VIEW ,,°50 100 1000 10000 a:: °.50 100 1000 10000
FREQUENCY FREQUENCY
FIG. 11.27. A plan view of a living room with a radio receiver. The graphs show the
response frequency characteristics for various positions in the room.
WINDSHIELD
ENGINE
HOOD
0:.0007
'"
r~
;.0006
0:.0005
I
/ \
\
w
0
~.0004
'"
V
/v\ ~ ~OOON
---
E·0003 .......
8'"
.0002 V \. r---. I--
f(
"'
z 6000",
o
~.OOOI
'. r ~ """
'
............ 3000N
I/)
III
1500",
-< 00
10 20 30 40 50 60 70 80 90 100
PER CENT RELATIVE HUMIDITY AT 20' CENTIGRADE
per centimeter for 1500, 3000, 6000, and 10,000 cycles as a function of the
humidity is shown in Fig. 11.29.
U. Sound Transmission Through Partitions. 45 ,46,47,48,49,50,51,52,53_The
problem of sound transmission through partitions and walls is complicated
because of the many factors involved. The problem of the mass-controlled
single-wall partition is very simple. The sound insulation of this type of
partition is proportional to the mass and frequency. For the usual building
materials and walls of ordinary dimensions supported at the edges, the
problem is that of the clamped rectangular plate with distributed resistance
throughout the plate and lumped damping at the edges. Obviously, the
performance of this system depends upon the size, the ratio of the two linear
dimensions, the weight of the material, the damping in the material, and the
edge supports. This type of problem is not amenable to an analytical
solution.
The transmittivity of a partition is defined as the ratio of the intensity
in the sound transmitted by the partition to the intensity in the sound
incident upon the partition. The transmission loss, in decibels, introduced
by the partition is given by
Ii 1
T.L. = 10 lOglO 1- = 10 lOglO- 11.17
i T
i" 21 36 I 38 1 40 45 41 40 .0001
Tile, 2" Gypsum 20 25 34 44 51 63 48 .000016 P.E.S.
Cinder Block 4" Un
finished 32 12 18 22 23 30 22 .0063
Cinder Block 4"
Plastered Both Sides 37 36 38 40 44 42 40 .0001
Wood Studs, 2" x 4",
Staggered, Metal
Lath, Plaster, i", 8"
Spacing 21 44 46 48 52 51 48 .000016
Tile Clay 6" X 12" x
12" Plastered both
sides 37 41 35 45 52 40 .0001 B.S.
Brick, 8" Plastered
both sides 87 50 48 55 63 50 .00001 B.S.
Brick, 16" Plastered
both sides 174 50 52 55 58 60 55 .0000032
Double Cinder Block,
Each 8", Air
Space 4", Outside
Plastered 132 55 56 58 67 60 60 .000001
Door, Light 4 Panel 13 16 20 23 22 22 .0063 P.E.S.
Door, Oak 15 18 23 26 25 25 .0032 P.E.S.
Door, Steel t" 25 27 31 36 31 35 .00032 P.E.S.
Window Glass, Plate i" 3.5 33 31 33 35 30 .001 B.S.
Window Glass, Small
Panes -/1..e-" 19 20 24 31 28 29 .0013 P.E.S.
Double Window Glass
t", Air Space 1" 7.0 42 43 44 48 45 45 .000032
Finish and Rough
Flooring, Plaster
Ceiling .
Concrete Slab 4" 48
34
38
1 35
39
I 37
42
39
46
39
48
37
42
.0002
.000063 .
I
Concrete Slab 4" Sus
pended Plaster Ceil
ing 51 46 150 55 5S 58 I 52 ,.00000631
The abbreviations in the above table are as follows: B.S., Bureau of Standards;
P.E.S., P. E. Sabine.
538 ACOUSTICAL ENGINEERING
Equation 11.18 shows that the reduction is due to both the loss introduced
by the partition and the absorption in the receiving room.
The choice of a partition for insulating a room against sound involves a
number of considerations. Some of the factors are the frequency distri
bution and intensity level of the components of the objectionable sound,
the transmission frequency characteristics of the partition, the ambient
noise or sound level in the receiving room which will mask the objection
able sound, and the response frequency characteristics of the ear.
Measurements have been made by various investigators upon the trans
mission by single partitions. The results of these measurements are shown
in Table 11.3.
The mass-controlled partition with air between the partition elements
is a low-pass filter in which the mass of the wall is the series element and
the volume between the partitions is the shunt element. The partitions
in this case are mounted in edge supports which allow freedom of motion
without cracks which would pass air-borne sound.
11.3. Collection of Sound.-A. Sound Collecting System. 54-When
a source of sound is caused to act in a rtlom, the first sound that strikes a
collecting system placed in the room is the sound that comes directly from
the source without reflection from the boundaries. Following that comes
sound that has been reflected once, twice, and so on, meaning that the
energy density of the sound increases with the time, as the number of
reflections increase. ultimately, the absorption of energy by the bound
aries equals the output of the source and the energy density at the col
lecting system no longer increases; this is called the steady state condition.
Therefore, at a given point in a room there are two distinct sources of sound,
namely: first, the direct and, second, the generally reflected sound. For
rooms that do not exhibit abnormal acoustical characteristics it may be
assumed that the ratio of the reflected to the direct sound represents the
effective reverberation of the collected sound.
Consider a sound collecting system, Fig. 11.30, the efficiency of reception
of which may be characterized as a function of the direction with respect
to some reference axis of the system. (The nondirectional collecting system
is a special case of the directional system in which the efficiency of reception
is the same in all directions.) The output of the microphone may be
expressed as
e = flPh(t/I) 11.19
54 Olson, H. F., Proc. Insf. Rad. Eng., Vol. 21. No.5, p. 655, 1933.
ARCHITECTURAL ACOUSTICS 539
If the distance, in centimeters, between the source of the sound and the
collecting system is D, the energy density at the microphone due to the
direct sound is
Eo
ED = D 2 47TC 11.20
where Eo = power output of the sound source, in ergs per second, and
c = velocity of sound, in centimeters.
To simplify the discussion, assume that the effective response angle of
the microphone is the solid angle Q steradians. The direction and phase of
the reflected sound are assumed to be random. Therefore, the reflected
sounds available for actuating the directional microphone are the pencils
of sound within the angle Q. The energy response of the directional micro
phone to generally reflected sound will be Qj47T, that of a nondirectional
microphone. The generally reflected sound to which the directional micro
phone is responsive is, therefore, given by
•
0
--
100
- OIR
RUl.E: TED
1000
FREQUENCY
T ....
10000
2 II
~'~~~~+r--~-H I
DIR£1~~
~ R£FLibTED
'" ~o 100 10000 ~ 100 toaD 10000
FREQUENCY
FIG. 11.30. Sound collecting system in a studio. Graph A shows the reverberation
time and the absorption coefficient of the boundaries of a typical studio. Graphs B, C,
and D show the energy response for the direct and reflected sounds for various micro
phones as follows: B. Nondirectional microphone. C. Bidirectional velocity or uni
directional microphone. D. Ultradirectional microphone.
SOUND
SOURCES 2
A B
~::~:"'-gz"'-UN~~~:;<D~~~~'::
CAMERA SOUND
SOUND SOURCES •
so~u
ELEVATION VIEW
II II
ELEVATION VIEW PLAN VIEW
C D
ULTRACIRE:CTIONAL
r
MICROPHONE
-0":CA"ERA
TSOUND
~OURC[
E F
FIG. 11.31. Examples illustrating the use of directional
microphones.
57 Olson. H. F .• Jour. Soc. Mot. Pic. Eng.• Vol. 16. No.6. p. 695. 1931.
58 Olson. H. F .• Jour. Soc. Mot. Pic. Eng.• Vol. 27. No.3. p. 284.1936.
69 Volkmann, J. E., Jour. Acous. Soc. Amer., Vol. 13, No.3, p. 234, 1942.
Fig. 11.32 are used to break up discrete reflections and thereby obtain a uni
form distribution of reflected sound energy in the studio. In general, studios
of this type are rectangular parallelopipeds.
Various expedients, in addition to the wall structures of Fig. 11.32, are
used to break up flutters and echoes. In some designs the walls are inclined
to eliminate parallelism between opposite walls. In other designs the
ceiling and/or walls are broken up into two or more nonplanar surfaces.
The ratio of the dimensions 72 of the studio is important in distributing
the characteristic resonant frequencies uniformly over the frequency range.
The graph of Fig. 11.33 shows the ratio of the dimensions for small, medium,
200
..>- ~ ~~
~
- -
100
t- 80
o.J
o.J 60 V
.........:i
'" 50
40
~
........
~ ~ ... 1-"
,........
~ i--- ,........1-'"
~
..,....
,........
,....
-
30
Z
Q
en 20
~ ::::: ~ f--", V ............
......'"
~
~ ~ ~~
i-"~
~ ~
",,"'"
Z
o.J ............
~
15 10
~
"" ,........ V
5 ............
3 45678e Z 3456789 2 3 4 5 789
2
104 105
VOLUME IN CUBIC rEET
and large studios. The most desirable ratio of the dimensions would be
in the ratio of the cube root of 2. This separates the dimensions by one
third octave. This ratio is possible for small studios but is not practical
for large studios, in that the ceiling height becomes too great. The dimen
sions of the small rooms are given by the lines C, D, and E of Fig. 11.33.
For medium studios the ratio of the dimensions is near the cube root of 4.
The dimensions of medium studios are given by the lines B, D, and F of
Fig. 11.33. This is approximately the ratio 2 : 3 : 5 which has been fre
quently used in the design of broadcast studios. For very large studios
the dimensions are given by the lines B, C, and G of Fig. 11.33.
In the second class of studio, termed the auditorium type, the performers
occupy one end of the room and the audience the other end. A plan and
sectional view of an auditorium type broadcast studio is shown in Fig.
11.34. The wall and ceiling surfaces of the stage are arranged to provide
72 Volkmann, J. E., Jour. Acous. Soc. Amer., Vol. 13, No.3, p. 234, 1942.
ARCHITECTURAL ACOUSTICS 545
sound diffusion so that the reflected sounds are properly mixed and the tonal
quality of the performer or performing group is enhanced. The stage
ceiling is broken in a saw-tooth fashion to provide a sound diffusing con
dition and to conceal the border lights and spotlights from the eyes of the
audience. The rear wall of the stage is constructed of a series of plaster
polycylindrical surfaces to provide diffuse reflection of the sound from this
boundary. The side walls of the stage are provided with spherical surfaces
for diffusing the reflected sound. The acoustical treatment of the stage,
I- :I
'" 0
" a:
::; 0
AUDIENCE
STAGE
AREA
CONTROl-
AOO~
ELEVATION VIEW
save for the diffusospheres on the side walls and polycylinders at the rear
of the stage, is rockwool, 2 inches in thickness, covered with perforated
sheet metal or perforated sheet asbestos. The diffusospheres throughout
the studio and polycylinders at the rear of the stage are made of plaster and
backed by rockwool. The ceiling and the side walls in the rear two thirds of
the auditorium section are untreated. The walls and ceiling in the front
of the auditorium section are treated with 2 inches of rockwool covered by
perforated asbestos. Heavy upholstered chairs in the audience area pro
vide substantially the same acoustical conditions with and without an
audience present in the studio. The control room is located so that the
occupants have an unobstructed view of the stage and studio seating section.
The clients' room is located so that the sponsors may watch and listen to
546 ACOUSTICAL ENGINEERING
the progress of the program. Lighting booths are also provided in the
rear for lighting the stage.
C. Sound Pickup Arrangement for a Radio Broadcast.-A broadcast studio
with a setup for the sound pickup for a sound broadcast is shown in Fig.
11.35. The monitor room is located next to the studio. A large window
between the monitor room and the studio makes it possible for the monitor
man to view the action in the studio. Separate microphones are used for
the orchestra, actors, and announcer. A transcription turntable supplies
RECORDIST TRANSCRIPTION
TURNTABLE
74 Volkmann, J. E., Jour. A cous. Soc. A mer., Vol. 13, No.3, p. 234, 1942.
75 Rettinger, M., Jour. Soc. Mot. Pic. Eng., Vol. 39, No.3, p. 186, 1942.
76 Livadary and Rettinger, Jour. Soc. Mot. Pic. Eng., Vol. 42, No.6, p. 361. 1944.
77 Slyfield, C. 0., Jour. Soc. Mot. Pic. Eng., Vol. 42, No.6, p. 367, 1944.
78 Ryder, L. L., Jour. Soc. Mot. Pic. Eng., Vol. 42, No.6, p. 369, 1944.
o ANNOUNCER
SO
MICROPHONE
o
MICROPHONE cp ACTORS
AUDIENCE
o 0
STAGE
cp MICROPHONE
000
SOUND
EFFECTS
D
0
CJo
ORCHESTRA
0
One fourth of the convex surface is made absorbent as shown in Fig. 11.37.
The ceiling construction is similar to the wall surface save that one fourth
of the convex surface is equipped with ventilating grills instead of absorbing
material. Wood polycylindrical surfaces comprise the rear wall of the shell.
The treatment on the side walls and rear wall consists of rockwool packed
between 2 by 4 inch vertical studs. Wood strips, 1 by 2 inches, were applied
to vertical studs graduated in spacing from 27 inches near the wainscoting
to 12 inches near the ceiling. Fiberboard i inch thick and plywood i inch
thick were applied to the studs between the stripping to produce a series of
horizontal rockwool, fiberboard and plywood panels. The construction of
the ceiling is similar to the walls save that the plywood panels are omitted
and the fiberboard panels made narrower due to the reflective floor parallel
to it. Since the wood strips were thicker than the plywood and fiberboard
panels, a sheet of muslin is stretched over the walls and ceiling to form a
monolithic surface broken only by a narrow decorative molding fastened
to the furring strips. The wall construction provides a uniform absorption
548 ACOUSTICAL ENGINEERING
and eliminates concentrations of the reflected sound. The live shell with
convex surfaces provides an ideal environment for the orchestra as well as
providing a means for eliminating sound concentrations and for directing
the flow of sound toward the absorbing part of the studio.
REFLECTING
SURFACE
ABSORBING
SURfACE
PLAN VIEW
IR
FIG.
.1
ELEVATION VIEW
cp MICROPHONE
fast tempo, the ratio of direct to reflected sound must be kept large or the
music will be blurred. In order to accomplish this objective, it is necessary
to use a very small distance between the microphone and the instruments.
This in turn requires a number of microphones, as shown in Fig. 11.40, in
which each microphone covers a single instrument or group of instruments.
In this arrangement, the balance of the orchestra is controlled by the
monitoring engineer. Sometimes dance bands are picked up for broad
casting from very noisy surroundings. This is another reason for the use
of a large number of microphones placed close to the instruments because a
larger ratio of desired sound to noise is obtained by this type of pickup.
The preceding considerations of the pickup of sound have been concerned
with single-channel sound reproduction.
Within the past three years stereophonic sound reproduction has been
developed and commercialized in sound motion pictures on a wide scale.
550 ACOUSTICAL ENGINEERING
o o o
TRUMPETS TROMBONE
o
o
o
0
0
0
BASS~N 0 K
DWMS
V
HORNS
o o 0 0 0
CELLOS VIOLAS 0
o 0
I ST
0
VIOLINS
0
o
CONDUCTOR
MICROPHONE
See Secs. 9.9 and l3.7B. The system has been of the three-channel type.
Stereophonic sound recordings for reproduction in the home in the form of
both disk and magnetic tape have been developed and commercialized
during the past three years. See Sec. l3.SE. The system has been of the
0 0 0 -& 0
TRUMPETS \ DRUMS
0 MICROPHONES
."'~
0 0
TROMI;IONES
¢
0
MICROPHONE
0
SAXOPHONES
0 0
~'(, " ,.. ..
rp MICROPHONES
I
MICROPHONE
.g.. 0
@ VOCALIST
CONDUCTOR
A
FIG. 11.41. Arrangement of the microphones with respect to area of action for
two- and three-channel stereophonic sound pickup.
channel system. Each microphone covers a portion of the action. The use
of three channels gives a smoother distribution. For this reason in some of
the recent motion-picture systems of the road-show type, five channels are
used.
F. Vocal Studios.80,81-In sound motion pictures when scoring an
orchestra and one or more vocalists, it has been the practice to record the
orchestra on one film channel and the vocalists on the second or separate
film channel. This permits great latitude in musical balance when the two
sound tracks are dubbed together. Frequency discrimination or accentua
tion of various portions of the frequency ranges in either or both the vocal
and orchestra recording may be made without any relation between the two.
Compression may be carried out in either or both channels. Synthetic
reverberation may be added in either or both channels. It is quite evident
that the use of two separate channels permits a wide range of artistic effects
which would be impossible if a single original record were made.
The vocal studio should be located adjacent to the orchestra studio. A
window between the two studios should be placed so that the vocalists or
80 Mounce, Portman, and Rettinger, Jour. Soc. Mot. Pic. Eng., Vol. 42, No.6, p. 375,
1944.
81 Ryder, L. L., Jour. Soc. Mot. Pic. Eng., Vol. 42, No.6, p. 379, 1944.
552 ACOUSTICAL ENGINEERING
vocal group can see the conductor. The vocalists hear the orchestra by
means of telephone receivers which reproduce the orchestra. In general,
the number in the vocal studio will not exceed thirty.
The acoustics of the vocal room should be similar to that of a small
standard broadcast studio (see Sec. l1.3B). One of the most important
considerations in the design of a vocal studio is the sound isolation between
the vocal studio and the orchestra. The sound level of the orchestra in the
vocal studio must be sufficiently low so that it will not be recorded in
the output of the microphones in the vocal studio. The type of wall and
window construction for the vocal studio to obtain the desired value of
sound isolation can be determined as outlined in Sec. 11.2R.
G. Reverberation Time of Broadcasting, Recording, and Scoring Studios. 82
-The optimum reverberation time of broadcasting, recording, and scoring
~:EfUlmll
I.(~o 100
FREQUENCY IN CYCLES
1000
PER
Ilmll
SECOND
10000
<111.4
o
z
81.2
v
w
on V
;!;I.o
/
~~
~.8 ,,/'"
1=
./
z .6
o
t-
« 4 -~
"'.
W
In
"'.2
'">
'"0: {i;oo 10000 100000 1000000
VOLUME IN CUBIC FEET
FIG. 11.42. Lower graph shows the reverberation time for a recording or
broadcasting studio as a function of the volume for 1000 cycles. Upper
graph shows the relation between the reverberation time and the frequency,
that is, the reverberation time at other frequencies is obtained by multiply
ing by K. (After Morris and Nixon.)
CAMERA
MAN
FIG. 11.43. Perspective view of a sound stage showing the arrangement of the elements
used in recording a sound motion picture.
MAN MEN
FIG. 11.44. Perspective view of a sound stage and monitor room showing the arrange
ment of the elements used in television broadcasting.
For the broadcasting of speech, the distance from the speaker to the micro
phone can be made very small so that the received reverberation is negligible.
However, for sound stages with large pickup distances, the received rever
beration is kept low by making the reverberation time of the sound stage as
low as possible and by the use of directional microphones. When the
reverberation time of the stage is low, the setting determines the acoustics
of the sound pickup by the microphone. In the case of sets consisting of
small rooms, the acoustics of the set mask the acoustics of the sound stage.
In the early days of sound motion-picture recording it was customary to
make the sets of acoustical materials having good transmission at the low
frequencies and high absorption at the high frequencies. In this way it
was possible to keep the reverberation time of the set very low. With the
advent of directional microphones it has been possible to use conventional
materials for the construction of sets.
In the television show described above and depicted in Fig. 11.44, there
ARCHITECTURAL ACOUSTICS 557
is no audience. In another type of television show, the action takes place
before an audience. For this type of show, conventional theaters have been
used. In some of these theaters the stage has been increased in size. The
broadcast studio of the type shown in Fig. 11.45 may be used for a television
show played to an audience. In the special television studio designed for
a show played to an audience, an amphitheater is combined with a sound
AUDIENCE
Q:l -
r
CAMERAS
1
STAGE
BOOM
o
ACTORS
o
MICROPHONE
(Y'
ANNOUNCER
MICROPHONE EB
cp ~ ~ MICROPHONE
0 0 0 ( /SPECIAL
cJ o 0 EFFECTS
ORCHESTRA
of Fig. 12.3 shows that the effective volume is somewhat greater than the
actual volume which is due to the resilient walls and eardrum.
~ l I l
0
0
0 .
0
0
~ ~
~
~
T
p MOrAO M, CAl TAl I I I MH
CAKM;:; I
M; I I I
10 I. 20
•• 30MM
FIG. 12.2. Schematic sectional view and acoustical network of the hearing mechanism.
In the acoustical network: p = the actuating sound pressure. Mo and rAo = the inert
ance and acoustical resistance of the air load upon the opening to the ear canal. M v
CAl' and r Al = the distributed inertance, acoustical capacitance, and acoustical resistance
of the ear canal. M2 = the inertance of the eardrum and hammer. CA2 = the
acoustical capacitance of the eardrum and tensor tympani. C A3 = the acoustical
capacitance of the hammer handle. M ... = the effective inertance of the ossicles and
hammer-anvil joint. CA4 = the acoustical capacitance of the hammer-anvil joint and
anvil arm. M4 = the inertance of the stirrup and oval window. CAo = the acoustical
capacitance of the oval window and the tensor stapedius. Ml" M 2', Ma' ... MK'
represent inertances of the liquid in the scala tympani. Ml", M 2", Ma" ... MK"
represent inertances of the liquid in the scala vestibula. Ml"" M 2'" ... MK"', and
CAl" CA2', and CAK' represent the inertances and acoustical capacitances of the basilar
membrane which separate!; the upper from the lower liquid. MH = the inertance of
the liquid in the heliotrema. The nerve terminals are represented by the volume
current meters Xv X 2 ••• XK' C A6 , M o' and rAa = the acoustical capacitance,
inertance, and acoustical resistance of the round window. (After Steinberg.)
60
400
a" r-..r-..
I'--~
w
U
200
a
-- :
V
V
Z
~ -200 ~
w
0. ,......1-
~ -400
...J
<t
u
-600 V
t
~ -800
k(
o
<t -1000
/
-1200
1I
V
-1400
-1600
V
50 100 500 1000 5000
FREQUENCY IN CYCLES PER SECOND
governed by all the elements of the vibrating system, that is, the acoustical
capacitance, CAl, of the vocal cords incurred by tension, the inertance, MI,
and acoustical resistance, r AI, of the vocal cords, the inertance, M 2, and
acoustical resistance, r A2, of the aperture and the load acoustical impedance,
ZAV, due to the vocal cavities. A schematic view and the acoustical net
work of the vocal cavities can be seen in Fig. 12.6. This shows that the
nature of the input acoustical impedance, ZAV, to the acoustical cavities is
extremely complex. The inertances M I , M2, and M3 and the acoustical
capacitances CAl and C A2 can be varied by changing the sizes of the apertures
and the volumes of cavities.
The oscillation of the vocal cords is of the relaxation type rather than
the conventional sinusoidal variation. This is borne outS by the rapid
starting and stopping in the case of some sounds. The oscillator shown
in Fig. 12.5 will produce waves of the relaxation type providing the circuit
8 Drew and Kellogg, Jour. Acous. Soc. Amer., Vol. 12, No. 1, p. 95, 1940.
562 ACOUSTICAL ENGINEERING
constants and the nonlinear elements are suitable. The wave shape of a
relaxation oscillator corresponds to the general wave shape of the output
of the vocal cords 9 shown in Fig. 12.7. The output of the vocal chords
was measured with a pressure microphone in the pharynx with the mouth
and nose cavities damped. The microphone was terminated in a long
damped pipe one inch in diameter. Referring to Fig. 12.7, it will be noted
that the bottom of the wave corresponding to the closing of the vocal chords
is flattened. This means that the vocal chords are closed or practically
HARD
PALATE
NOSE
TEETH
closed during a portion of the cycle. Nevertheless, the general shape of the
wave is that of a saw tooth. A saw-tooth wave contains the fundamental
and all the harmonics. Therefore, the generator, PG, should produce the
fundamental frequency and all the harmonics of the fundamental frequency.
When the vocal cords are set into vibration as outlined above, the output
of the larynx consists of a steady stream with superimposed impulses
(Fig. 12.7). This pulsating air stream passes through the air cavities of
the head. The harmonic content of the output is modified due to the dis
crimination introduced by the acoustical network of Fig. 12.6. The effect
of the vocal cavities is illustrated in Fig. 12.7, which shows the wave shape
9 Olson, H. F., Unpublished Report.
SPEECH, MUSIC, AND HEARING 563
of the sound output of the mouth and nose corresponding to the wave shape
of the output of the vocal cords. When the shape of the vocal cavities is
altered the acoustical elements of the acoustical network of Fig. 12.6 are
altered which in turn alters the output harmonic content. These changes
together with a change in the fundamental frequency of the vocal cords
CA'
TO
VOCAl.
t",
ZAV CAV'TIES
M2 M,
fA2 M,
C",f",
El.ECTRICAL ANAl.OGY
ZAL
1<
XOC CA,
Poe TO
l.LlNGS
SCHEMATIC VIEW
ACOUSTICAL NETWORK
M, M2
M2
M4
M,
NETWORK
CAl
ZAV
.SCHEMATIC VIEW
vocal cords are not used in the production of these sounds. The voice
constants are produced by a combination of the two systems.
The voice mechanism then consists of a number of acoustical elements
which can be varied by the person at will to produce a wide variation of
tones differing in frequency, quality, loudness, duration, growth, and decay.
~~ M f'I f'I K
~ V-V~\J'V'V'\J
Q. TIME OUTPUT OF MOUTH AND NOSE
....
e
FIG. 12.8. Sectional views of the head showing the
configuration of the mouth cavity for different vowel
sounds.
13Harris, Cyril M., Jour. Acous. Soc. Amer., Vol. 25, No.5, p. 970, 1953.
14Stevens, Kasowski, and Fant. Jour. Acous. Soc. Amer., Vol. 25, No.4. p. 734,1954.
15 Cooper, Delattre, Liberman, Borst, and Gerstman, Jour. Acous. Soc. Amer., Vol. 24,
No.6. p. 597, 1952.
16 Stevens and Bastide. Jour. Acous. Soc. Amer., Vol. 27. No. l.p. 207.1955.
568 ACOUSTICAL ENGINEERING
the resonances of the vocal tract, and actuated by saw tooth and noise
generators. Coded information from a punched paper tape controls the
resonant frequency of the tuned circuits, the frequency of the saw-tooth
generators, and the amplitude of the noise generator. Sounds, words, and
sentences have been synthesized by means of this system.
12.5. Visible Speech,l7-Visible speech is an electronic method of
changing spoken words into visible patterns that someone may learn to read.
A schematic arrangement of the apparatus for depicting speech in visible
patterns is shown in Fig. 12.10. Speech is picked up by the microphone
and converted into the corresponding electrical variations. These varia
tions are amplified and limited in amplitude so that the amplitude range is
BAND
PASS
f'ILTERS
o MICROPHONE
== "'" aa "
55.E. AMPLIFIER
~
!Ii!. '5;&; -.
!e _ J!J 5a=
_=-.. AND
- 5:.- ==
e:.
-.
&'i!!11iO.. :it:E LIMITER
;;;:;''!i!
"
FIG. 12.10. The elements of a system for producing visible speech. (After Potter.)
/
--------
120
120
11;;-- r-.. vI-" ~
100 ......1/ .-'~
100
- Vv ~J
-- -
III
I-- 90
~ c:::
Q
§ 80 I"'-
I-"
.-/
:;: .... -
~ r---
-- ---- J.
.....: l"' I/V
..J
'"
> , t" ~:::: L-----...--
70
V ~.
'"
..J60
...<ii>- 1'-' ~
-.....
60
50
IJ /
I'-~ ...........
.....
z
...
"'40 ........
-I- 40 V ../
i'......~
~
20 '"
.............
............ ........
.........
........
.......r-.
.......'"
.......
30
20
10
-- -----
IJI-"
..... v
./ I-"
/
/ .
/
o
20 ICO
,REQUENCY
FIG. 12.11. Contour lines of equal loudness for normal ears. Numbers on curves
...........
IN CYCLES
1000
0
PER
-
SECOND
V
i-'
10000
I0S
./
,/
r:! /
v
0' /
..,..,
'"z
a /
:>
o
...11 v
V
I
10
I
I
I
o 10 co 30 40 50 50 70 80 90 100
LOUDNESS LEVEL IN PHONS
reference tone until it sounds equally loud. The loudness level, in phons,
of It sound is numerically equal to the intensity level, in decibels, of the
1000-cycle pure tone which is judged by the listeners to be equivalent in
loudness. The phon is the unit of loudness level as specified in the preceding
sentence. A scale 19 showing the relation between loudness level, in phons,
and the loudness, in loudness units, is shown in Fig. 12.12. Recently a new
unit of loudness, termed the sone, has been adopted. I t is defined as follows:
a pure tone of 1000 cycles per second, 40 db above the normal threshold of
hearing, produces a loudness of one sone. One sone is equivalent to 1000
loudness units.
19 Fletcher H., Jour. Acous. Soc. Amer., Vol. 9, No.4, p. 275, 1938.
SPEECH, MUSIC, AND HEARING 571
The 10udness 20 ,21 of pure tones of various frequencies is shown in Fig.
12.13. For tones between 800 and 2000 cycles the loudness is the same for
the same pressure. The difference is small up to 8000 cycles. For higher
frequencies than this the loudness decreases as the frequency increases.
RELATIVE PRESSURE
I 10 100 1000 10000 100000
100000
.Y
1/ /.~~
'b1°-;
~ ,,/ ~
10000 ,1°_
~~ ~
/"
~ v,; 7" .j
::l
...J IOOO A ~ t/ ~1 /
~
// ~ 'b /
'"'"w // ) / V
Z
o
100
::l
-
/ / ~ '( / I
o r--o
...J
f-- c-,{/
/// Ij / II
10
Ifl
I
'11
I
r
II
- - ~
/ / I
I /I I
o
J' J201 40
f
60 80 100
INTENSITY LEVEL IN DB
FIG. 12.13. The relation between the intensity level and the
loudness of pure tones of the frequencies indicated. 0 db
= 0.000204 dyne per square centimeter. (After Fletcher.)
Further, it will be seen that for a 50-cycle tone the intensity required to
reach the threshold of hearing is 250,000 times that required for a reference
1000-cycle tone.
12.8. Pitch.-Frequency of a sound wave is the number of cycles per
second executed by the particles of the medium in which a sound is being
propagated. Pitch is that subjective quality of sound which determines
its position in a musical scale. Pitch may be measured as the frequency
of a pure tone having a specified sound pressure which seems to the average
ear to occupy the same position in a musical scale. Thus it will be seen
that there is definite distinction between frequency and pitch. For example,
a tone of a fixed frequency of a few hundred cycles decreases in pitch as the
20Munson and Gardner, Jour. Acous. Soc. Amer., Vol. 22, No.2, p. 177, 1950.
21 Beranek, MarshoII, Cudworth, and Peterson, Jour. Acous. Soc. Amer., Vol. 23,
No.3, p. 261, 1951.
572 ACOUSTICAL ENGINEERING
o
60
I-
~ -5
u
,r-. .......-~
f-""" ~
-/ ' t:/'/ ' '/
..............
~ t::;:::::;
/
V ./
I-""" ~
~
..,
~-15
, r-..~
110/
/
< \ If
o ~
120
0-20
I- LOUDNES~ ~
0::
LEVEL
-25
50 100 200 500 1000 200o
F"REQUENCY IN CYCLES PER SECOND
pitch of 1000 mels. The pitch of any sound that is judged by the listener
to be n times that of a I-mel tone is n mels.
12.9. Masking. 23 ,24-The reduction of the ability of a listener to hear
one sound in the presence of other sounds is known as masking. In testing
the masking properties of a sound, pure tones are generally used as the
masked sound. The number of decibels that the threshold level of a pure
tone is shifted, due to the presence of noise, is called the masking in decibels
at the frequency corresponding to that of the pure tone.
The masking effect of a pure tone, a narrow band of thermal noise and a
wide band of thermal noise is shown in Fig. 12.15. The figures on each of
the curves show the intensity level of the masking tone or noise. The
ordinates, in each of the charts, show the decibels above the threshold
that the various frequencies must be raised in order to be just heard in the
presence of the particular masking tone or noise. For example, referring
to the 4000-cycle tone having an intensity of 93 db, it is only necessary to
raise a 2000 cycle tone 20 db to be heard. On the other hand, a 10,000
cycle tone must be raised 66 db to be heard.
22 Snow, W. B., Jour. Acous. Soc. Amer., Vol. 8, No. 1, p. 14, 1936.
23 Wegel and Lane, Phys. Rev., Vol. 23. No.2. p. 266. 1924.
24 Fletcher and Munson, Jour. Acous. Soc. Amer.• Vol. 9, No. 1. p. 1, 1937.
200 CYCLE TONE NOISE 1140 TO 2110 CYCLES 100 NOISE 140 TO 1200 CYCLES
10 I
1 I 1m ~02
!Ie 80 80
99
I~ 1/ ~})
!60
~40
"
/ ..
lit, . .
4 '\
110
40 liff?3.
~/~9
"-
60
40
jl/ 1'\71 •
~3\\
~20
,,,}. r-... l!~/,
< tt,..~ ~ ~~.~
::-:i:~
20
~;; . ...•'...
40"
...
~ 0 o. • o
1000 CYCLE TONE 100 NOISE 10~TOI250CYCLEs 10ONOISE 530 1800 CYCLES
100
r\ I ~
eo 80 80
60
te~l\ 80
1\\ 03
60
Igh
40
.~
40
~\\ 831\ 40
~
\6C '\ .' ~l
I, \ 1- \~2
0
,it,)
20
o
\~
~
20
o
'~ . .-,: . ~-{ ::
4000 CYCLE TO NE 10 NOISE 3000 TO 32110 CYCLES /0ONOISE 3000T05~::VC LES
0
J, II
0 80 eo
II0
I~
110
I~I/
\ ~~~ /I. I
\
0 40 40
100
t:i3 i. ' ~\I
1000
FREQUENCY
20
0
10000 100 1000
~
FREQUENCY
'll
20
10000 qoo
- .. ..i5
.'
1000
FREQUENCY
10000
FIG. 12.15. Masking audiograms for single frequency tones, narrow bands of thermal
noise, and wide bands of thermal noise. The curves are labeled in db above the threshold.
(After Fletcher and Munsen.)
25
28Wegel and Lane. Phys. Rev.• Vol. 23, No.2, p. 266,1924.
1929.
28 Chapin and Firestone, Jour. Acous. Soc. Amer., Vol. 5, No.3. p. 173. 1934.
29 Lewis and Larsen, Nat. Acad. Sci., Vol. 23. p. 415, 1937.
30 Stevens and Davis, "Hearing," John Wiley and Sons, New York, N.Y.• 1938.
574 ACOUSTICAL ENGINEERING
V
,/
, / ."
.......
i-'
~ --
5!,!:i ~TH
/ ~V
,/ 2ND
~V
V
..-::
:::::
FIG. 12.16. The level above threshold at which harmonics are generated in
the ear at the various frequencies. (After Wegel and Lane.)
.0 7
.06
.05
\5
Af·04
~f\ \
T .03
E~~
~~
~
.02
.0 I
~ "" "" " -
~ I'-. .......
.........
r-= :::: ,
FIG. 12.17. The variation of Jflf with frequency for various sensation
levels. Jf is the change in frequency. Sensation level is level above
threshold. (After Shower and Biddulph.)
8
\
1\
1\
~
" r--.. ~
~
r-." '-- ..............1 VV
r-...
-
~
--
- --
-I- ./
v/
---
..........
~
r--. . . . r- ..l0
~,...
I
:::-- 40
so
100 1000 10000
F"REQUENCY IN CYCLES PER SECOND
The minimum time required for a pure tone to excite the ear in order to
be sensed as a pure tone is of interest in sound reproduction from the stand
point of the transient response and hangover. It appears that the time is
independent of the frequency and is about one twentieth of a second.
The minimum perceptible change in intensity 34 which the ear can detect
is of interest in certain types of sound reproduction where the level may
change as, for example, fluxations in the voltage gain of an amplifier. The
minimum perceptible change in intensity level of pure tones as a function
34 Fletcher, "Speech and Hearing," D. Van Nostrand Company, Princeton, N.J.,
1929.
576 ACOUSTICAL ENGINEERING
of the frequency for various sensation levels is shown in Fig. 12.18. These
characteristics show that the ear is most sensitive to intensity level changes
at the higher sensation levels.
12.14. Timbre (Tone quality).35,36,37-In general, it is said that the
three characteristics which describe a tone are loudness, pitch, and timbre
or quality. These quantities are not sufficient to describe a tone. Three
more are required, as follows: vibrato, duration, and growth and decay.
Loudness, pitch, vibrato, duration, and growth and decay are defined in
other sections. It is the purpose of this section to describe timbre.
Timbre is that characteristic of a tone which depends upon its harmonic
structure as modified by the other physical factors that describe a tone.
The harmonic structure of a tone is expressed in the number, intensity,
distribution, and phase relations of its components. Timbre, then, may
be said to be the instantaneous cross section of the tone. It ranges from
a pure tone through an infinite number of variations in complexity up to
a pitchless sound such as thermal noise. Work has been carried out on the
subjective measurements of timbre. However, the subject of timbre is
more complex than that of loudness and pitch, because it is an interrelated
function of the intensity, pitch, duration, growth, and decay.
12.15. Duration.-The duration of a note in music is indicated by the
kind of a note, as, for example, a whole, a half, a quarter, or an eighth note.
It is quite evident that the duration of a tone influences the aspect as per
ceived by the ear. A fine musical ear may detect a difference in the length
of two tones as small as 0.01 second. The duration is one of the important
means that the artist has for the interpretation of music.
12.16. Growth and Decay.3s-In the case of a certain instrument the
instantaneous cross section of the tone may be exactly similar to that of
another instrument but to the ear the sound appears entirely different.
The difference is due to the growth and decay of tone. In the case of an
organ pipe, time is required for the tone to build up and die down. In
the piano the build up time is very fast and decay time is very long when
the key is depressed. It is quite evident that growth and decay play an
important part in the nature of musical tones.
There is a special case of growth and decay, namely, portamento. Porta
mento is the passage from a tone of one frequency to a tone of a different
frequency in one continuous glide through all the intervening frequencies.
12.17. Auditory Localization. 39 ,40,41-The human hearing mechanism
36 Seashore, "Psychology of Music," McGraw-Hill Book Company, New York,
N.Y., 1938.
36 Fletcher, H., Amer. Jour. Phys., Vol. 14, No.4, p. 215,1946.
37 Olson, "Musical Engineering," McGraw-Hill Book Company, New York, N.Y.,
1952.
38 Olson, "Musical Engineering," McGraw-Hill Book Company, New York, N.Y.,
1952.
39 Stewart, G. W., Phys. Rev., Vol. IS, No.5, p. 425, 1920.
40 Stevens and Davis, "Hearing," John Wiley and Sons, New York, N.Y., 1938.
u Steinberg and Snow, Bell Syst. Tech. Jour., Vol. 13, No.2, p. 245, 1934.
SPEECH, MUSIC, AND HEARING 577
can localize sounds with great accuracy. This property is due to two effects,
namely, the difference in intensity and the difference in phase between the
sound at the two ears. The difference in phase between the sounds at
the two ears is due to the difference in time arrival at the two ears. The
difference in intensity at the two ears is due to diffraction. The pressure
at the two ears may be obtained from Fig. 1.5 by assuming the head to be
a rigid sphere. This assumption has been verified experimentally.
The binaural phase effect is confined principally to the lower frequencies,
namely, below 1000 cycles. The binaural phase effect has been utilized
for the location of airplanes, submarines, etc.
If attachments can be made to the ears which will virtually separate
them further (for example, spread the microphones on the dummy of Fig.
13.7), then a small rotation of the apparatus will mean a larger difference
in phase at the receivers than at the unaided ear. Of course, there may
be ambiguity as to whether the sound comes from in front or behind, but
this does not vitiate the method.
The difference in intensity at the two ears due to diffraction is very small,
below 1000 cycles. However, at the higher frequencies the difference in
intensity may be 20 to 30 db.
The human hearing mechanism is also a directional collecting system.
Using the system of Fig. 13.7, the reverberation in a room appears to be
normal. However, if only one microphone is used, the apparent reverbera
tion will be increased. This indicates that the human hearing mechanism
is very directional and discriminates against reverberation and other
undesirable sounds. For this reason it has been found necessary to use
. either very small pickup distances or directional collecting systems in the
monaural collection of sound.
12.18. Delay Effect.42-It has been established that if there are several
separated sources of sound, identical in content and amplitude but displaced
with respect to time, that the sound will appear to come from the source
which leads the other in time. The experiment which illustrates this
phenomena is shown in Fig. 12.19. The same signal is reproduced from
loudspeakers 1 and 2. The signal from loudspeaker 1 can be delayed by
means of the delay system. For each value of delay, the ratio of the voltage
input to the two loudspeakers is varied until it is impossible to distinguish
which loudspeaker appears to be the source. The results of this test are
shown by the graph of Fig. 12.19. This experiment shows that there can
be considerable unbalance before the sound ceases to appear to come from
undelayed source. With the same intensity of sound emanating from both
loudspeakers, the source always appears to be the undelayed loudspeaker.
The delay phenomena plays an important role in stereophonic systems.
12.19. Hearing Acuity in the United States Population. 43-Hearing
tests at New York and San Francisco World's Fairs in 1939 were conducted
42 Haas. H .• Akustika. Vol. 1. No.2. p. 49. 1951.
43 Steinberg. Montgomery, and Gardner. Jour. Acous. Soc. Amer.• Vol. 12. No.2.
p. 291. 1940.
578 ACOUSTICAL ENGINEERING
VOLTMETER
OBSERVER
12
C/)
~IO
5
e
08
.,...,6
';;-4
o
I
lL
/ -.........
..........
-----
0:2
o0 10 20 30 40 50
DELAY IN MILLISECONDS
MEN WOMEN
AGE5 AGES
-
5 5 10- 19
10-19
0 20-29
20-29
'o" 'o"
5 "- 5
---..
30-39
<Il
<Il
o
.J
,0
15
.......
,"" 40-49"":;
:ll'
o
v
..J
15
r--- -
50 -59_
"z
~20
~'\ >-- "~2
Z
0
~
'" 2 5
I: \ Y
50-59.,...< '" 2 5
I:
0
3440 880 1760 3520 7040 880 1760 3520 7040
FREQUENCY IN CYCLES PER SECOND FREQUENCY IN CYCLES PER SECOND
FIG. 12.20. Hearing loss frequency characteristic of men and women for different age
groups. (After Steinberg, Montgomery, and Gardner.)
SPEECH, MUSIC, AND HEARING 579
as a part of the Bell System Exhibits. About one-half million records were
obtained. The tests were divided into five age groups, 10-19, 20-29, 30-39,
40-49, and 50-59 years. Many cross checks were made with laboratory
tests to insure accuracy as, for example, the effect of background noise,
calibration of the equipment, estimation of age, etc.
The results of the tests for men and women are shown in Fig. 12.20. It
will be seen that the hearing acuity falls off with age, particularly in the
high-frequency ranges.
The composite results of these tests have been depicted in another manner
in Fig. 12.21. The upper curve is the threshold of feeling level. The lower
--
140
120
r- --.
\ \ '-- ~) I
\ '\ I
:g 100
\ \ , , I
80
\\ , / I
\, "
oJ
~ ,, ',,- ..... V I
I
~ 60 PE
"
~
iii 40
"
"
',,- ............. .....
-
CENT
5 V )
..... ~ .............
....... /
Z
~ 20
.~,,/
L
r- .... SO
o
" ............. 5
.... V
./
,.,V
-20~ 4 • 102 2
rREQUENCY
4
IN
103
CYCLES
• 2
PER
4
SECOND
• 104 2
curve of Fig. 12.11 is the threshold of hearing level for standard normal
hearing. The lower curve labeled 95 indicates that 95 out of 100 persons
in a typical group cannot hear pure tones whose frequency and intensity
level lie below this curve. The curve labeled 50 indicates that 50 out of
100 persons cannot hear these tones until they exceed the intensity level
indicated by this curve. The curve labeled 5 indicates that 5 out of 100
cannot hear these tones until they exceed the intensity levels indicated by
this curve.
12.20. The Frequency and Volume Ranges of Speech and Music.
The frequency range of the average normal ear is from 20 to 20,000 cycles.
The frequency range of most reproducing channels such as the radio, the
telephone, the phonograph, and the sound motion picture is considerably
less than that of the hearing range of the ear. It is interesting to note the
effect of the frequency range upon the intelligibility of speech and the
quality of music.
580 ACOUSTICAL ENGINEERING
,,,nnn
!II
...
oJ
.~,
" r
-- 200
~
u
a<II
..
III
20
...o
U
\ ~ ~~
~
2
~ ,~~ ~ I lalSl % ~
...> "
~
.J .....1"-
... """"<:r4:; /; V/hZ
W0;::W
102
~J
"
.J
Inn:;>
...
II:
> .......
::>
I-
V
-
iii .......
...... !II
Innn" ...
<II
...z
I .... II:
Q.
Z
_'>n
20 100 '00 )00
fRtQUtNCY IN CYCLtS PtR SECOND
FIG. 12.22. Frequency and volume ranges of speech and music. The solid
lines depict the boundaries of normal hearing. that is, the upper and lower
limits of intensity and frequency. (From Bell Laboratories Record, June, 1934.)
45 Snow, W. B., Jour. Acous. Soc. Amer., VoL 3, No. 1, Part 1, p. 155, 1931.
46 Fletcher and Galt, J01,1.r. Acous. Soc. Amer., Vol. 22, No.2, p. 89, 1950.
47 Hirsh, Reynolds, and Joseph, Jour. Acous. Soc. Amer., Vol. 26, No.4, p. 530, 1954.
TIMPANI
BASS DRUM
SNARE DRUM
'4" CYMBALS
BASS VIO~
CE~~O
PIANO I
VIOLIN
BASS TUBA
TROMBONE:
rRENCH HORN
TRUMPET
"ASS SAXOPHONE
BASSOON
BASS CLARINET
CLARINET
SOPRANO SAXOPHONE
OBOE:
ACCORDION
HARMONICA
rLUT[
PICCOLO
PIPE ORGAN
MALE SPEECH
rEMALE SPEECH
FOOT STEPS
HAND CLAPPING
ROOU
I JIL
KEY JINGLING
4 8 102 2 4
I
• 103 2 4 8 104 2
FREQUENCY IN CYCLES PER SECOND
The effect of reducing the high- and low-frequency range upon speech
articulation is shown in Fig. 12.24. These data were obtained at a talking
level into the system of 70 decibels. The reproduction level was also 70
decibels.
The syllable articulation in a sound reproducing system equipped with
low-pass filters with cutoff frequencies at 750, 1000, 1500, 1950, 3700, and
7000 cycles is shown in Fig. 12.25, as a function of the gain in the system.
The talking level is an over-all measure of the acoustic level and is defined
582 ACOUSTICAL ENGINEERING
.9
- ""
....- I,...--'
.8
"'"
'\.
J
/
/
LP/ \HP
/ \
:3
/ \
V
2
o
I
8001000
FREQUENCY IN CYCLES PER SECOND
2000
1\
4000
, 10000
FIG. 12.24. The effect of the frequency range upon the syllable articulation
of speech. HP High-pass filter-all frequencies below the frequency given
by the abscissa removed. LP Low-pass filter-3.ll frequencies above the
frequency given by the abscissa removed. (After French and Steinberg.)
'" -10
/ ~ ~~ ,\.,. 9 /
/"
I
2
V
J .8
~
U-20 z .7
II V
o
"' / 1/ ....... 3
Q
ti·
J
6
IV I 4
!-:30 ::>. 5
.,"' u
Ii V
~-40
.,
Q.
6 5 4 3 2 I
~q .4
'I / /
/ V
5
:t!-50
-60
100 200 400 1000 2000 4000 10000
.2
.1
0
-60
"
III/, ./ v
'.I:
-40 -GAIN
20
~
0 20
-6
40 60
FREQUENCY IN CYCLES PER SECOND IN DECIBELS
FIG. 12.25. The syllable articulation as a function of the reproduced sound level for
various low-pass transmission characteristics. (After Fletcher, "Speech and Hearing
in Communication," D. Van Nostrand Company.)
...~ -10 .8
II)
~-20
~ z
Q
.7
I
"'-2
~
o ~ .6
V
;; -30 :5<J .5
... I 2 3 4 ;:: .4 II
'~-40
"
a:
<! .3 I /-- 4
I I 1/
gj .2
Ir -50
I Ii ,/
o ~ ,/
-6900 200 400 1000 2000 4000 10000 -60 -40 -20 0 20 40 60
FREQUENCY IN CYCLES PER SECOND GAIN IN DECIBELS
FIG. 12.26. The syllable articulation as a function of the reproduced sound level for
various high-pass transmission characteristics. (After Fletcher, "Speech and Hearing
in Communication," D. Van Nostrand Company.)
II)
U z .7
~-20r---r---~-r~~~H-~~+--+~ Q
~.6
ill L IL
z 3 2 III '/4 5
~-~r---~~~~~~~~~~~-+~ :5<J
.5
.h I I
'"~-40r---~~~-.~+---~~~~-+~
z
..
III
;::.4
a:
<! .3
.2
'"rA 'J
/ ~ /
a:-50r---+-~~---r+---~+-~~-+~ /, ~ V
I
o /.~ V'
-60 -40 GAIN
-20 IN DECIBELS
0 20 40 60
FIG. 12.27. The syllable articulation as a function of the reproduced sound for limited
frequency range transmission characteristics. (After Fletcher, "Speech and Hearing in
Communication," D. Van Nostrand Company.)
For the optimum level of 69 decibels above the threshold level, the data of
Fig. 12.28 show the level in decibels of each critical frequency band. It is
to be noticed that response level falls off in the low- and high-frequency
ranges.
The data of Figs. 12.24, 12.25, and 12.26 indicate that some frequency
regions are more important than others in contributing to articulation or
70
..
j;J50
60
./
\-"" "n...
ID
~40
o
/ '-I
;!;30 /
-'
'"
~ 20
V
-'
10
o
100 200 400 600 800 1000 2000 4000 8000 10000
FREQUENCY IN CYCLES PER SECOND
intelligibility of speech. The importance function D(f), that is, the im
portance for articulation or intelligibility per cycle, is shown in Fig. 12.29.
The importance for articulation or intelligibility is the integral J~ D(f) df
which is depicted in Fig. 12.29.
1.0
.9 ./
/~
--°6
.B
V
~.7
~.
/ j
f
D(f)df
a: .5 / 0
/-
o
=.4 /
o
"'2.3
~
/V" / 103 D (f}
.2
V ,/' ""'" ""'k
---
I
.-'
o
100
/ 200
I--
400 BOO 1000 2000 4000
r-...
10000
FREQUENCY IN CYCLES PER SECOND
FIG. 12.29. The importance function per cycle, D(f) and the importance
f:
integral D(f) df of each frequency region for articulation or intelligibility.
(After Fletcher, .. Speech and Hearing in Communication," D. Van Nostrand
Company.)
SPEECH, MUSIC, AND HEARING 585
The preceding considerations of articulation have been concerned with
the effects of a nonuniform response frequency characteristic and the level
of the reproduced sound. There are other types of distortion which influence
the articulation and intelligibility. One interesting experiment is that of
multiplying all frequencies by the same number. This experiment can be
performed by recording a phonograph record at one turntable speed and
reproducing it at a different turntable speed. The syllable articulation as
a function of the ratio of the speed of rotation of a phonograph record in
reproducing to the speed of rotation in recording is shown in Fig. 12.30.
1.0
.9
.8
/ .......... 1'0..
I
z
2.6
.7
/ '" i'..
'" "'
~ /
:5u .5
~ .4
/
<I
.3
.2 /
/
I
The interesting fact is that there can be a large ratio of frequency shift
. before the articulation is impaired.
The reverberation of a room influences the articulation and intelligibility
of speech. The effect of reverberation upon the articulation is described
in Sec. 11.2F and depicted in Fig. 11.9.
Nonlinear distortion is produced when there is a deviation from linearity
in the input-output characteristic of the reproducing system. See Secs.
12.26 and 12.27. One of the most common examples of nonlinear distortion
is the overloading of the amplifier in a sound reproducing system. The
effect 49 of nonlinear distortion produced by the overloading of an amplifier
in a sound reproducing system upon articulation is depicted in Fig. 12.31.
The syllable articulation characteristic labeled 1 in Fig. 12.31 was obtained
on a wide frequency range, high quality feedback amplifier with no over
loading. The talking level was 70 decibels above threshold reference level.
49 Pollack, Irwin, Jour. Acous. Soc. Amer;, Vol. 24, No.5, p. 538, 1952.
586 ACOUSTICAL ENGINEERING
1.0 1.0
I
.9
V ...... .9
......
.B " .......
z .B
/ /
........
~ .7
II \ o 7 /
~.
1
~ .6 ~.6
3.5 I \2 :3
u· 5
1/ 12
u
i= .4 II ~\ ~.4 / 1
II: \ <I I
<I .3 3
.2
j i. 2 I
I
I
V \
I
I I
o /
-20 L-40 l/-20
-60 -40 0 20 40 60 EO 0 20 40 6o
GAIN IN DECIBELS GAIN IN DECIBELS
FIG. 12.31. The syllable articulation of FIG. 12.32. The syllable articulation of
sound reproducing system for two con producing system as a function of the
ditions of operation as a function of the reproduced sound level for two conditions
the absence of noise. The talking level is 70 decibels above threshold level.
The reproduction level of 0 gain corresponds to the talking level of 70
decibels. The measured noise level was very near the threshold of hearing,
when room noise having the spectrum distribution shown in Fig. 12.38
for a residence was introduced. This corresponds to a total noise level of
43 decibels. It will be seen that the articulation is reduced by a large factor
at the lower levels of reproduction. However, as would be expected, at the
higher levels of reproduction the articulation is not reduced.
Another form of distortion which may be applied to speech is clipping. 50
It is interesting to note that when the level of clipped speech is increased
the articulation approaches that of unclipped speech.
From the foregoing data it will be seen that a reasonably good articulation
and intelligibility of speech is obtained under conditions of severe inter
ference and distortion, as for example, a narrow frequency band, in the
50 Martin, D. W., Jour. Acous. Soc. Amer., Vol. 22, No.5, p. 614, 1950.
SPEECH, MUSIC, AND HEARING 587
presence of noise and large nonlinear distortion. However, the quality of
the reproduced speech is very. much impaired by these distortions. On the
other hand the intelligibility of a limited frequency range may be actually
superior to a wider frequency range due to the introduction of additional
noises and distortions in a wider frequency band unless particular precautions
are observed. In the case of speeches, plays and songs a limited frequency
range impairs the quality and artistic value of the reproduced sound.
12.22. The Effect of Frequency Discrimination upon the Quality of
Reproduced Music.-The effect of the frequency range 51 upon the quality
of reproduction of orchestral music is shown in Fig. 12.33. It will be seen
that the frequency range required for no appreciable loss in quality is
10
!za \..
... 0
v
II:
~6 0
\ /
40 K ~
>
r-
oJ
« \ /
V
=>20
a ,/
~o 100
"" 1000 10000
CUTOff fREQUENCY IN CYCLES PER SECOND
FIG. 12.33. The effect of the frequency range upon the quality of
orchestra music. H. High-pass filter-all frequencies below the
frequency given by the abscissa removed. L. Low-pass filter-all
frequencies above the frequency given by the abscissa removed. (After
Snow.)
52 Sivian, Dunn, and White, Jour. Acous. Soc. Amer., Vol. 2, No.3, p. 330, 1931.
53 Sivian, L. J., Jour. Acous. Soc. Amer., Vol. 1, No.2, Part 2, p. 1. 1930.
55 Drew and Kellogg, Jour. ACGus. Soc. Amer., Vol. 12, No. 1. p. 95,1940.
In
o L
-f-IL I"- I~ ~
::IE
M. h I-.
;:)
ct:-IOO
l.
.,;;; -100
h -100
'"r
~
U
'-L
!oJ
0.-140 140
~ 140
II)
41 BASS SAXOPHONEJ 5FT _2042 BASS VIOL 5FT -2 o 23
. TRIANGLE 3FT
0
[t::
r= -60 -60
,.1
"i: IL I
"-
0-100
h
100
n: - 100
IJ
!oJ
ct:-140 140 140
:;)
=:
!oJ
_20 86 TRUMPET 3FT -2
0
35 CLARINET .
35fT _20180 CYMBALS 3FT
ct:
0.
!oJ
CI
~-IOO
IoJ
- " l
It
-100
II IU
h
-,;; -100
-h I~ I-.. I
>
<-140 140 -140
0 -60
1- M -6 0 If II
~h LU It II
--.;;
I,...J 1.
-100 -10 0
L
-140 140 -140 h.
_20 16 FLUTE 3.5f"T -2 77 MALE SPEECH 21N _2063 fEMALE SPEECH 21N
-140 140
20 100 1000 10000 20 100 1000 10000-14~0 100 1000 10000
FREQUENCY IN CYCLES PER SECOND
FIG. 12.34. Ratio of the average pressure per cycle to the average total pressure of the
entire spectrum for speech, various musical instruments, and orchestras. The distance
and average total pressure, in dynes per square centimeter, are shown above each graph.
(After Sivian, Dunn, and White.)
SPEECH, MUSIC, AND HEARING 589
o I Li1- o r ~
h- 0 r "'"h
f-t-' f-j.J L.
1 i- I- II II
I o lL
~ 0 LJ 0
"1.1- r
f- ....
w
l
~~O
fI)
~o -30
::: 30 41 BASS SAXOPHONE35fT 30 4 2 BASS VIOL 5fT 3023 TRIANGLE 3fT
a: Ir
II.
,~~ r 11 r
f..-U 1'- h~
o o
j L I
'1. -"L I I
L... t-'
-30 1 -30
TRUMPET 3FT 3035 CLARINET 3 5fT 30180 CYMBALS 3f1:
"1..
o 0 Ir ~
o u r o
~
I- "L 1/ h r
u 1/ -~
'"
~-30 -30
- -30
II. 22 PICCOLO 3 SfT 3065 TROMBONE 3fT 3038 fRENCH HORN 3fT
0
...o 3
rL In 1/
t-'
0 o Ir o I... ~
h- 1/ rr P'h
I- I LI II
-3 I -30 -30
'"h
30 I 6 fLUTE 35F'1: 3077 MALE SPEECH 21N :3 o 63
• fEMALE SPEECH 21N.
o 1..- M t-rr-,
r-,
0 oI--'-h
I--f LU
-30 -30 -30
20 100 1000 10000 20 100 1000 10000 20 100 1000 10000
fREQUENCY IN CYCLES PER SECOND
FIG. 12.35. Ratio of the peak pressure to the average total pressure of the entire spec
trum for speech, various muscial instruments, and orchestras. The distance and average
total pressure, in dynes per square centimeter, are shown above each graph. (After
Sivian, Dunn, and White.)
590 ACOUSTICAL ENGINEERING
III
o
20
10
~f'~~
1040 ~ ~
oeo ~ A IA t'o
o ~L~,:!&-::
~ 10
~~ t.iji i"O:
080 . . ~
R
'"
bJ ~ 'I'tl ~ j..o.~t< f'o
It:
~ -10 /090
0/. l?'.1 "'"\ l'q ~),~1,,1'<> I'c;J'b
x ~ ~ f.blfo [p. 1'01"\
'"
bJ
~ -2 0
u
..'"''"" d po- F.a:.:: ~b'1oI..o 1:'t:1"
0/ 10t<" ~~
:.: "'" <> I'b.~ ro.~ ~j...;:
< -30 0:
I'b. 10 J.P-I-o.
!:i'"
'"'"
bJ
Q.
0;: "'\
-40 z
" I'"
-50
6.2.5 125 250 500 1000 2000 4000 8000 -50L-~62L.5~~12~5--2~~~5~0-0~10~O~OL2~0~~~40~0~0~8~~~
fREQUENCY fREQUENCY
FIG. 12.36. Peak pressures in one-eighth-second intervals of conversational speech, at
30 centimeters from the mouth; composite from the voices of 6 men and 5 women.
Measurements were made in the bands indicated by divisions of the frequency scale,
and the percentages are those of intervals having peak pressures greater than the indicated
ordinates. At the left in the graphs, peak measurements on speech as a whole are
given. The rms pressure over a long-time interval is also given. 0 db = 1 dyne per
square centimeter. (After Dunn and White.)
nl'lC.
MEN WOMEN
III 10 10
I '~"16 1
°l
P- III
o
l~
o
40
\;0l'~ ./ :&:: 1%
Z 0 2°0 ~ONC ,...... v;--. I.-. 1
~
~ -10
m I/) -I
111
P80
~~ W h:l:~ W
II:
°pso 0.- ~ ~r-,., )-0. 1<>-1-<,.
:l P90 I~ \~ P>ro :l 1'90 70n- ~ ~~ ~ th l.n.
~ "20 :lj-2 0
~ ra~ ti:\1'0 1-0
bJ
a:
Q. -3 o
"
~~ 7~
~o.f..n
\'q
1'0 bJ
II:
a. -3 o
"
wx
"u
w ...
00--
" ~
~ I:.i;!~to.
I'-..
I\,
... ') ~~ 1'0
........
~
j7~~
f~
~
ctl
~ -4 o
.....
_G.
Z
:0
~~ ~
ctl
~ -4
-Q.
Z
~
II: o "
It: 1 I
62.5 125 250 500 1000 20~ 4000 8000 62.$ 12.5 250 500 1000 2~ 4000 s~
fREQUENCY fREQUENCY
FIG. 12.37. Rms pressures in one-eighth intervals of conversational speech, at 30 centi
meters from the mouth; composite of 6 men and 5 women. Measurements were made
in the bands indicated by the divisions of the frequency scale, and the percentages are
those of intervals having peak pressures greater than the indicated ordinates. At the
left in the graphs, rms measurements on speech as a whole are given. The rms pressure
over a long-time interval is also given. 0 db = 1 dyne per square centimeter. (After
Dunn and White.)
SPEECH, MUSIC, AND HEARING 591
and orchestras is shown in Fig. 12.34. The ratio of the peak pressure to
the average pressure of the entire spectrum for speech, various musical
instruments, and orchestras is shown in Fig. 12.35. The peak or total
power output can be computed from the pressure and the distances following
the procedure is outlined in Sec. 10.3D.
The peak sound pressures 56 in speech for various frequency bands at
a distance of 30 centimeters are shown in Fig. 12.36. The percentages
on the graphs are those of intervals having peak pressures greater than
the indicated ordinates. For example, the characteristic labeled 20 per
cent means that 20 per cent of the peaks in the particular frequency interval
exceeded the ordinate value of the characteristic. At the left of the graphs,
peak measurements of the entire frequency band are given.
The rms sound pressures in speech for various frequency bands at a
distance of 30 centimeters are shown in Fig. 12.37.
12.24. Noise in Reproducing Systems.-Static, hiss, scratch, surface
noise, and hum are terms which have been introduced to describe various
kinds of noises produced by reproducing systems. A few of the types of
noises in reproducing systems will be discussed in this section.
In radio broadcasting systems random noise is produced by thermal
agitation 57 in conductors, and the small shot effect58 and fluctuating noise
in vacuum tubes are atmospherics and man-made interference. The
energy produced by thermal noise and the small shot effect is proportional
to the width of the frequency band. Of course, the effects of atmospherics
and man-made static can be reduced by increased power. Accentuating
the high-frequency response in transmission and introducing comple
mentary equalization in the receiver will reduce the noise, since the noise
is proportional to the band width. For the same reason, frequency modula
tion also reduces noise in radio reproducing systems.
Hum 59 is another source of noise in radio, phonograph, and sound motion
picture reproducing systems because practically all of these systems are
operated from the alternating current mains. Hum is due to inadequate
filtering in the high-voltage supply and inductive and capacitive coupling
between the power source and some part of the audio system.
Surface noise or record scratch in phonograph records is due to dirt or
foreign particles in the groove and to a granular characteristic of the record
material. The record noise generated in a pickup which produces constant
output for constant velocity is proportional to the frequency for a narrow
frequency band of constant width. It is for this reason that scratch is such
a troublesome problem at the high frequencies in record reproduction.
Surface noise in film motion picture reproduction is caused by the modula
tion of the light falling on the photocell, by dirt, scratches, and the natural
grain of the film. In the case of film the noise is proportional to the
56 Dunn and White, Jour. Acnus. Soc. Amer., Vol. 11. No.3, p. 278, 1940.
frequency band width. Noise reduction systems increase the ratio of signal
to noise in film reproduction by reducing the light to the least possible for
the instantaneous modulation.
Superimposed on the average atmospheric pressure are fluctuations caused
by thermal. velocities of air molecules. 6o This noise places a lower limit
upon the air as a transmitting medium. The ear exhibits the greatest
sensitivity between 1000 and 6000 cycles. The rms sound pressure due to
thermal noise in this frequency range is about .00005 dyne per square
centimeter. The maximum threshold sensitivity of the ear from Fig. 12.11
is .00008 dyne per square centimeter. In very sensitive ears the threshold
at 1000 cycles may be .0002 to .0001 dyne per square centimeter-that is,
of the same order of magnitude as thermal noise. For exceptionally good
ears a further increase in physiological sensitivity would be useless in the
presence of thermal noise. It is interesting to note in passing that the
thermal noise of the air molecules acting upon the diaphragm of a very
sensitive microphone may be of the same order as the thermal noise generated
in the electrical system (see Sec. 8.15).
The noises referred to above occur in the reproducing system. The
room noise in the studio, home, office, factory, and theater also influences
the reproduction of sound. Room noise and the reproduction of sound
will be considered in the next section.
12.25. Room Noise and the Reproduction of Sound. 61-The hearing
curves of Figs. 12.11 and 12.21 set the limits for an ideal transmission system
with the listener in a quiet place. The ideal of no noise is seldom realized
by listeners. Therefore, the lower limit is determined by the ambient
room noise. I t is the purpose of this section' to show the effect of room
noise upon the reproduction of sound.
The average noise spectrum 62 may be obtained from the measurements
of room noise. In general, the noise spectrum is the same for all types of
rooms. From Fig. 11.20, the total noise level for an average living room
is 43 decibels. The spectrum for average room noise having a total level
of 43 decibels is shown in Fig. 12.38. The ordinates, depicting the spectrum
level, are given by
I
B = 10 log W10 12.1
62 Hath, D. F., Jour. Acous. Soc. Amer., Vol. 12, No.4, p. 499,1941.
63 Fletcher and Munson, Jour. Acous. Soc. Amer., Vol. 9, No. L p. L 1937.
8or----,-.----,----r----~,_--_.----~--_r~
10
a
=<
:::J
~ Ol~---+~----~---+----+-~---+~~~---+~
IJ
W
Q.
~-2~~---8~IO~2.-~2~---4~---8~IO~3~~2~---4~---.~,~
FREQUENCY IN CYCLES PER SECOND
0
...
1'-
"
',
... ...
""-
i'... "
.........
t-....
)'-.
_ .....9 0...
.....~
80 '
" - , .
...
,
50
4 0",
I"-
-" i'... f"-. ,.......
i'... ")'-.
N;--- 1'...... '1--1--'
"
~o""" 1'......
.z 0
'" C"'.:: i'... " ~ t-- r"-- I-L.
L.
0
0
~ ~ "- r--..
I'-........ I"--t-....
'" ~
~
1'--.. __
r----.. "-
~
i--'
0
"-
I"- t-.. 10 t---...
......
-
-10
......... l V
-30
........
I-
100 500 1000
" 5000 10000
FREQUENCY IN CYCLES PER SECOND
and the masking contours of Fig. 12.39. For example, for a spectrum level
of 9 db at 1000 cycles, the masking level is 25 db. The masking curve for
average room noise obtained by these means is shown in Fig. 12.40. The
curves for an average business office and an average factory are also shown
in Fig. 12.40.
594 ACOUSTICAL ENGINEERING
From the masking curve of Fig. 12.40 it is possible to determine the per
missible hum and noise level in a radio receiver or phonograph used in the
home. For example, the 60-cycle hum must be kept below a 57-db level.
The 120-cycle component must be kept below a 41-db level.
120
:g 100 \
~
~~
..J
w
>
w
80
~
. . . r- :-- FAC~ORIES
I...Y
..J 60
,I'--- '---- J
>-
f-
iii 40 " r-...... - OF"f':ICES
l..I
-
..........
z
w
f-
r-- RESIDENCES
V
/
~ 20
02 .. 8 10 2 2 . 8 103 2 . 8 10 .
FREQUENCY IN CYCLES PER SECOND
FIG. 12.40. Hearing limits for pure tones. The above characteristics are
for a typical listener in typical residence, office, and factory noise. (After
Fletcher.)
The masking curve also shows that the volume range of the reproduced
sound is reduced by the ambient noise. For example, in the case of factory
noise the volume range is quite limited. Masking by the ambient noise is
one of the limitations for high-quality reproduction in noisy locations.
12.26. Combination Tones and Nonlinear Transducers.-In most of
the discussions in this book the elastic restoring force of the elements of a
vibrating system have been considered to be proportional to the first power
of the displacement. If a second power term is included the element is
asymmetrical, the restoring force being different in magnitude for positive
and negative displacements. According to most investigators the structure
of the ear is of such an asymmetrical character.
Assume that the displacement of the nonlinear element may be expressed
as follows
y = ap + bp2 12.2
where p = the actuating force,
a = a constant, and
b = a constant.
Suppose two harmonic forces h = po cos Wit and P2 = Po cos w2t are
impressed upon the system. The total force on the nonlinear element is
P= h + P2 = Po cos Wit + Po cos w2t 12.3
SPEECH, MUSIC, AND HEARING 595
Substituting equation 12.3 in 12.2 the resulting equation may be put in
the form
y = apo cos wIt + apo cos w2t + ibP02 cos 2wIt + tbpo2 cos 2w2t
12.4
+ bP02 cos (WI + W2)t + bP02 cos (WI - W2)t + bPo 2
If the element under consideration is the ear, it will be seen that six dif
ferent frequencies will be heard as follows: the first primary frequency
wI/27T, the second primary frequency W2/27T, the second harmonic of the
first primary frequency WI/7T, the second harmonic of the second primary
frequency W2/7T, the summation frequency (WI +
w2)/27T, the difference
frequency (WI - w2)/27T. The last term bp2 represents a steady force and
produces no sound.
Combination or sum and difference tones may be produced in any non
linear system as, for example, an intense sound in the air, in the throat of
a horn, by overloaded vacuum tube amplifiers, by diaphragms, and by
suspension systems.
12.27. Effect of Nonlinear Distortion upon the Quality of Reproduced
Speech and Music.-In an ideal reproducing system the elements are
invariant with respect to the time. However, in practical systems the
elements are nonlinear. These elements introduce nonlinear distortion.
LIVING ROOM
FREE-FIELD
SOUND ROOM
Some idea of the effect of nonlinear distortion can be obtained from a study
of the masking curves. From these data it will be seen that the high order
harmonics are noticeable at much lower levels than the lower order harmonics.
Furthermore, as the high-frequency range is increased the effect of the
harmonics is more noticeable. In the complex waves of speech and music,
sum and difference tones are also an important phase of the problem of
nonlinear distortion.
The effect of various types of nonlinear distortion 64 on the reproduction
of speech and music has been determined through the system shown in Fig.
12.4l. The over-all response frequency characteristic of the microphone,
64 Olson, H. F., Unpublished Report.
596 ACOUSTICAL ENGINEERING
I~ I11I I".
!oJ 0.5 WATT
IIII 0.5 WATT
I
~IIIII'
FIG.
I Z
12.42.
3 • !t 6 7 a 8 10 II 12 13 14
HARMONIC
IIII
I 2 3 4
COMPONENTS
amplifiers.
The sound was reproduced in a room with acoustics similar to a typical living
room. The noise level at the pickup point was about 0 db in the absence of
any performers. The noise level in the listening room was about 25 db.
Two types of distorting amplifiers were used-namely, a single-ended
triode and a single-ended pentode. The distortion components for these
two power output systems for various power levels are shown in Fig. 12.42.
These tests were limited to three subjective gradations of nonlinear
distortion-namely, perceptible, tolerable, and objectionable. Perceptible
is the amount of distortion in the distorting system required to be just
SPEECH, MUSIC, AND HEARING 597
discernible when compared to the reference system. Tolerable and objec
tionable are not as definite and are a matter of opinion. By tolerable
distortion is meant the amount of distortion which could be allowed in
20
III
o 0 -
:;!: -10
I ~r®"ti ~~
\oJ
<I)
~ -20
1\ \ 1\ \
n.
<I)
~ -30
\\ \1
4 8 10~ 2 4 8 10' 2 4
\ \\\
8104
rREQUENCY IN CYCLES PER SECOND
SPEECH MUSIC
TRIODE TRIODE
OBJECTIONABLE
TOLERABLE
PERCEPTIBLE
.5 2 4 8 16 4 8 16
DISTORTION PER
SPEECH
PENTODE
OBJECTIONABLE
TOLERABLE
PERCEPTIBLE
.5 2 4 8 16 .5 2 4 8 16
DISTORTION IN PER CENT
TELrrONE
TRANSCRIPTION
PHONr RAPH ,
OR I HIGH
rCORD
iUjLITY PH10NOGRAPH
I
HIGH QUjLlTY MjGNETIC TiPE REjRODUCER
SI UND
HIGH QUALITY
Ti
SrUjD
EM
SYSTEM
HEARINi ('0
D,'CTATINj rCHINE
NORMAL
HEY'r
I
2 4 e 10 I 2 4 8 10' 2 4 8 10·
FREQUENCY IN CYCLES PER SECOND
FIG. 12.44. The frequency ranges of sound reproducing systems.
high-quality sound systems are also in use in sound reinforcing systems and
other applications.
The frequency range of the hearing aid shown in Fig. 12.44 represents
the average response of high-quality vacuum tube hearing aids in use today.
The low-frequency range may be somewhat greater but in general this
added range cannot be used due to rumble and other low-frequency noises.
An examination of Fig. 12.44 shows that the general run of commercial
receivers, phonographs, television receivers, sound motion-picture systems,
magnetic tape reproducers, and sound systems sold during the past decade
600 ACOUSTICAL ENGINEERING
exhibit response frequency ranges between 100 and 5000 cycles. On the
other hand, the frequency range of radio and television transmitters, sound
motion-picture records, phonograph records, and magnetic tape is practically
that of the human ear. In other words, the potential frequency ranges of
radio and television receivers, phonographs, sound motion-picture and
magnetic tape reproducers are not limited by the source material or the
system. In this connection it is interesting to note that the sale of wide
frequency range high-quality radio and television receivers, phonographs,
sound motion·picture, and magnetic tape reproducers has increased at a
high rate during the past five years. In the case of some of these systems
great strides have been made to higher quality. There can be no doubt but
that there is a definite trend toward higher quality in reproduced sound.
One of the principal reasons for this state of affairs is due to the reduction
in nonlinear distortion.
12.29. Frequency Range Preference for Reproduced Speech and
Music. 65-In the preceding sections data have been presented depicting the
response frequency characteristics of the human ear, the manner in which
these characteristics vary with age, the loudness range, the effect of masking
sounds, the effect of nonlinear distortion, the effect of the frequency range
upon the articulation of speech and upon the quality of music, the frequency
ranges of speech and music, and the frequency ranges of reproducing systems.
A study has also been made of the frequency range preference of a representa
tive cross section of broadcast listeners.
As contrasted to the other data presented, the purpose of this investiga
tion was the determination of the frequency range of reproduced speech
and music that is most pleasant to the average listener. The investigation
was made with a variety of musical and voice passages. The tests were made
in a room with acoustics similar to those of a large living room. Both high
quality records and direct wire transmission from the studio were used with
very little difference in the results.
The frequency ranges employed for the tests are shown in Fig. 12.45
and were designated as wide, medium, and narrow frequency ranges.
The results of the tests are shown in Fig. 12.45. The general conclusion
of these tests is that listeners prefer either a narrow or medium frequency
range to a wide one. However, the exact choice of band width varies to
some extent within these limits, for different types of program content.
Listeners prefer a narrow to a wide tonal range even when informed that
one condition is low fidelity and the other high fidelity. Listeners prefer
a slightly wider band for female speech, piano, and popular orchestra selec
tiQns than for male speech, mixed dramatic speech, and classical orchestra
selections.
It is interesting to note that the frequency range preference of a repre
sentative cross section of broadcast listeners is essentially the same as the
frequency range of commercial radio receivers, phonographs, sound motion
picture systems, and sound systems.
65 Chinn and Eisenberg, Proc. Inst. Rad. Eng., Vol. 33, No.9, p. 571, 1945.
SPEECH, MUSIC, AND HEARING 601
fREQUENCY RANGES
I. NARROW
2.MEDIUM
3.WIDE
10
m
o
o
~ ..... r;. pf -- ~ ~~
... -10
I
III
Z
fi> -20 I \ 1\
...a:
III
:1 I \
-30
I' I
43'"
2 21"" 24'"
2
3 NP
56'!'.
27'1.
The frequency range preference for reproduced speech and music was con
sidered in the preceding section. These tests indicate that listeners prefer
66 Olson. H. F., Jour. Acous. Soc. Amer., Vol. 19, No.4, Part 1, p. 549, 1947.
602 ACOUSTICAL ENGINEERING
0
r\
, • ~g~o ~:S~LE
69;:!
o
o ;60 "
ORCHE STRA""
"0
..
ACOUSTICAL
FILTER
o ...." . . o-tNDICATOR
0
0
100 tooo
FREQUENCY
"
10000
lli
MUSIC
FIG. 12.46. Plan and elevation views of the schematic arrangement of the apparatus for
direct frequency preference testing of speech and music. A sectional view, acoustical
network and response frequency characteristic of the acoustical filter used in the tests
and the results are also depicted.
The reason for the difference between the results of the two tests is without
doubt due to the distortions listed in the first paragraph of this section.
The subjective tests of nonlinear distortion, described in Sec. 12.27, indicated
that the amount of tolerable distortion decreases as the frequency range is
increased. These tests also indicated that a very small amount of nonlinear
distortion can be detected when employing the full frequency range.
12.31. Frequency Ran~e Preference for Stereophonic Reproduced
Speech and Music. 67-Subjective tests of frequency range preference of
live speech and music comparing a restricted frequency range with the full
frequency range have shown that the average listener prefers the full fre
quency range. See Sec. 12.30. These tests were all acoustic. The sound
was not reproduced. Therefore, there were no electroacoustic transducers
67 Olson, Preston, Woodward, May, Morgan, and Bleazey, Unpublished Report.
604 ACOUSTICAL ENGINEERING
_..===~"" ORCHESTRA
f\ ----....j..ia..MICROPHONES
FIL"rERS
......" i'fo--+lI--LOUOSPEAKERS
CURTAIN
INDICATOR
, ...._ _ LISTENERS
LIVING ROOM
--- --
10
V-
A
-~..., ~A
o
~- '""- oJ
,, '\
\
\
\
~ -10
~
\
~
I \ B
I \
'"z -20
Ul
\
oQ. I \ \
Ul B/
::! -30 \
/
I 1\
-40
\
I
-50 I
of the reflected sound at the microphones for steady state sound conditions
is about 50 db below the direct sound. Therefore, it is impossible to detect
any acoustics of the studio in the reproduced sound.
The listening room is the same as that used in the all-acoustic test. The
listening room was designed to be the acoustical equivalent of an average
living room.
The reproducing system used in these tests employed two channels.
Each channel consisted of a RCA 44BX Velocity Microphone, a RCA OP-6
Amplifier, a laboratory developed triode-type push-pull power amplifier,
laboratory developed high- and low-pass electrical filters, and a RCA LCIA
Loudspeaker.
The over-all response frequency characteristics depicting the ratio of the
sound pressure output from the loudspeaker in free space to the sound pressure
606 ACOUSTICAL ENGINEERING
at the microphone in free space, with and without the electrical filters, are
shown in Fig. 12.48. In the restricted range condition there is attenuation
in both the low- and high-frequency ranges. The product of the low and
high cutoff frequencies is 500,000 (cyc1es)2. This value has been recom
mended by many investigators. 68 In this we have deviated from the original
all-acoustic frequency preference test in which a high frequency cutoff alone
was used. An argument in favor of the combination of high- and low
frequency cutoffs is that it approximates conventional radio and phonograph
response frequency characteristics.
The directivity pattern of the loudspeakers is very important where the
listeners are located at relatively large angles with respect to the loudspeaker.
In the loudspeakers used in these tests the variation in response at any
frequency over a total angle of 90° is less than ±2 db.
The nonlinear distortion is another important factor in reproduced sound.
The over-all nonlinear distortion was measured by supplying a distortionless
signal to the input of the chain consisting of the voltage amplifier, power
amplifier, and loudspeaker. The sound output of the loudspeaker was
picked up by the microphone and fed to a harmonic analyzer. This method
of measurement provides an overall distortion characteristic from sound
input to the microphone to sound output of the loudspeaker. The total
nonlinear distortion measured at the peak level of the reproduced sound
was less than 0.3 per cent. From the results reported in Sec. 12.27, it will be
seen that this value of nonlinear distortion is sufficiently low to be practically
imperceptible.
The level of the reproduced sound in the listening room is important in any
subjective test. Tests have shown that a peak level of about 70 db to 80
db is most pleasing for serious listening in a small room. The average peak
sound intensity level on a standard level indicator was 75 decibels.
The same six-piece band was used in these tests as in the case of the all
acoustic frequency range tests. The change from full frequency range to
restricted frequency range was made every 30 seconds. The results of these
tests indicate a preference for the full frequency range. Similar tests were
made for speech. The frequency preference tests for speech also indicate a
preference for the full frequency range.
12.32. Comparison of Live and Reproduced Symphony Orchestra. 69
-A demonstration was conducted on July 29, 1947, at the Berkshire
Festival in Tanglewood, Massachusetts, in which the full Boston Symphony
Orchestra was compared with a phonograph record of the orchestra. The
demonstration was made before an overflow audience in the Music Shed at
Tanglewood, Massachusetts.
Plan and elevation views of the Music Shed are shown in Fig. 12.49.
The main audience area with a ceiling height of 40 feet contains over 6000
seats. The colonnade, twenty-seven feet in width, along the outside edge,
68 Hanson, Rackey, and Nixon, "Down to Earth on High Fidelity," Eng. Dept.
N.B.C., New York, N.Y., March 27, 1944.
69 Olson, Preston, Riefsteck, Reiskind, and Pulley, Unpublished Report.
SPEECH, MUSIC, AND HEARING 607
with a ceiling height of twenty feet will accommodate an additional 3000
persons. From a consideration of the dimensions and the seating capacity,
it will be seen that the Music Shed is a very large structure.
The reverberation time of the audience area, with all the seats occupied,
ranges from 1 to l.5 seconds in the mid-frequency range. The relatively
f+---'<-*~------\----23B·-O· ---~
ELEVATION VIEW
low reverberation time for a structure of this size is due to the openings to
free space on all sides which represent 100 per cent absorption. However,
the reverberation time of the stage and shell is much greater. Therefore,
the general impression of perspective of the orchestra is lost save for the
first few rows of seats.
Plan and elevation views of the stage showing the location of the micro
phones used in recording and the loudspeakers used in reproducing are shown
in Fig. 12.50.
A schematic diagram of the recording system is shown in Fig. 12.50.
608 ACOUSTICAL ENGINEERING
Three RCA Type 44BX Velocity Microphones were used. The gain in each
of the three microphone channels was the same.
A schematic diagram of the sound reproducing system is shown in Fig.
12.51. Tests indicated that 12 RCA Type LC1A Loudspeakers driven by
3-40 watt amplifiers would reproduce the peak level of the orchestra. In
reproducing the record, twelve Duo Cone Loudspeakers were placed at the
front of the stage as shown in Fig. 12.50. In this way, the acoustics of the
reproduced sound matched the acoustics of the orchestra because the loud
speakers carried on, as it were, from where the microphones picked up the
sound. Furthermore, due to the relatively reverberant stage and relatively
dead listening area, it was not necessary to use perspectively reproduced
~
~--o
-.,
'" STAGE
69~0"~~ f!J
•
~3-, STAGE FLOOR
10'-0"
~
3'~r-1--.:........;.:..::...--"-'-'----D 4'-0"
,---- ---,
ELEVATION VIEW
sound to maintain the illusion that the reproduced sound was identical with
the original sound.
The peak sound level attained by the full orchestra at the box seats, which
are located about 85 feet from the stage, was 95 db. This value agrees with
measurements made on many other orchestras in many other locations.
It was found that this level could be matchp.d with 120 watts input to the
twelve loudspeakers in reproducing the sound. The noise level with the full
audience and when the audience was quiet was about 40 db.
The over-all response frequency characteristic depicting the ratio of the
sound pressure output from the loudspeaker in free space to the sound
pressure at the microphone in free space is shown in Fig. 12.52.
SPEECH, MUSIC, AND HEARING 609
The last four minutes of Beethoven's overture to "Egmont" was recorded
by the Boston Symphony Orchestra with Serge Koussevitsky conducting.
In the demonstration the Boston Symphony Orchestra, with Serge Kousse
vitsky conducting, played the preceding portion of the selection and the
MICROPHON ES MIXERS
RECORDING
~OUDSPEAKERS
VOLUME
CONTRO~
REPRODUCING
10
-
CD
o
-J r--v. '"
;;
- A
""'"
-.~
'" 0 V
o'"
z
:);-10
'"
II:
/ ""
-20
40 100 200 400 1000 2000 4000 10000 20000
FREQUENCY IN CYCLES PER SECOND
switch was made from original sound to reproduced sound. Many rehearsals
were carried out to insure continuity of the music as well as the amplitude
during the switch from live to reproduced music.
Comments from both music critics and layman were that tone color,
dynamic range, and general fidelity of the reproduced sound matched the
610 ACOUSTICAL ENGINEERING
J BANJO
UKULELE
PIANO
I
TRUMPET
TROMBONE
BASS TROMBONE
FRENCH HORN
BAas TUBA
PICCOLO
FLUTE
SOPRANO Cl.ARII\IET
ALTO CLARINET
BASS CLARIN E T J .1
I,SOPRANO SAXOPHONE
ALTO SAXOPHONE
TENOR SAXOPHONE
BARITONE SAXOPHONE
BAsa SAXOPHONE
I I
.1 OBOE
I
eNGLISH HORN
BASSOON
REED ORGAN
ACCORDION
HARMONICA
PIPE ORGAN
EL.ECTRIC ORGAN
rf=Ui
XYL~PHONE
10 II .. • 102. II .. 8 103 II 8 10"
FREQUENCY IN CYCLES PER SECOND
FIG. 12.53. Frequency ranges of the fundamental frequencies of voices and various
musical instruments.
SPEECH, MUSIC, AND HEARING 611
formances ... is little short of revolutionary in fidelity of reproduction."
Miles Kastendieck, Journal American.
" ... in an almost imperceptible switch. It was scarcely apparent where
the live orchestra left off and the canned music began, so alike in volume
and tone quality was the music which came from the machine.... It is the
closest I have yet heard mechanical reproduction come to the real thing."
Robert Hague, N. Y. PM.
"For a moment, even the listeners who anticipated the daring gesture
were not quite sure which was performing.... But it (the reproduced sound)
carried on with a veracity of tone color and dynamic range that a verdict in
its favor could not be denied."-Irving Kolodin, N. Y. Sun.
12.33. Fundamental Frequency Ran~es of Voices and Musical
Instruments.-Musical instruments and the voice produce fundamental
frequencies and overtones of fundamental frequencies. The overtone
structure is one of the characteristics which distinguish various instruments
and voices. If musical instruments produced the fundamental without
overtones, each instrument would produce a pure sine wave and would
therefore be the same as the output of an oscillator and loudspeaker combina
tion. The fundamental frequency is the lowest frequency component in the
complex sound wave. When a musician speaks of the "range" of a voice
or musical instrument, he means the frequency range of the fundamental
frequency. The fundamental frequency ranges of voices and various musical
instruments are shown in Fig. 12.53. There may be some variation from
these frequency ranges among various instruments and voices but in general
the ranges are typical. Comparing the frequency ranges of the fundamentals
of Fig. 12.53 with the entire frequency spectrum of musical instruments of
Fig. 12.23, it will be seen that the overtones of the instruments extend the
frequency ranges by a factor of two or more octaves.
12.34. Musical Scale. 7o-An octave is the interval between any two
tones whose frequency is 2 : 1.
A tone is a sound giving a definite sensation of pitch.
A scale is a series of tones ascending or descending in frequency by definite
intervals suitable for musical purposes.
A cent is the interval between any two tones whose frequency ratio is
the twelve-hundredth root of two.
For the practical production of music a so-called equally tempered scale
has been found to be most satisfactory. The equally tempered scale is a
division of the octave into twelve equal intervals called equally tempered
half tones.
The standard pitch for America is based on the frequency 440 cycles per
second for the tone A on the pianoforte keyboard. The frequencies of a
piano having an equally tempered scale are shown in Fig. 12.54.
The relative frequencies of the natural or diatonic or just scale and tem
pered scale are shown below.
70 Olson, "Musical Engineering," McGraw-Hill Book Company, New York, N.Y.,
1952.
612 ACOUSTICAL ENGINEERING
C D I E F G A B C
-- ---- -- ---- -- --
Natural Scale 1.000 1.125 1.250 1.333 1.500 1.667 1.875 2.000
----- - ---- -- - -- -- - ---- -- - --
Tempered Scale. 1.000 1.122 1.260 1.325 1.498 1.682 1.887 2.000
FIG. 12.54. The piano keyboard and the frequencies of the equally
tempered scale as used in music according to the American Standard
Pitch.
75 Meissner, B. F., Proc. IRE, Vol. 24, No. 11, p. 1427, 1936.
78 Olson and Belar, Jour. A cous. Soc. A mer., Vol. 27, No.3, p. 595, 1955.
FREQUENCY
(PITCH)
50 100
,
1000
,
10000 leooo
FREOUENCY IN CYCLES PER SECON D
I I
120
II>
90
oJ
III
INTENSITY .. 60
(LOUDNESS) U
:!:
50
GROWTH
!-----DURATION =------l
STEADY STATE
DECAY
~~I
DURATION
TIME IN SECONDS
PORTAMENTO
Ib--==
TIME IN SECONDS
TIMBRE
1\ ~ L-~I~,~I~I~llul__~1
100
=.
1000
__
10000
'
VIBRATO
vVVV v V \T\{VV VV
AMPLITUDE FREQUENCY
MODULATION MODULATION
means. Thus it will be seen that it is possible to generate any tone produced
by a voice or a musical instrument by employing an electronic system. In
addition, it is possible to produce musical tones which cannot be produced
by the voice or conventional instruments. In other words, the process of
translating the musical notation on paper into the corresponding musical
sounds can be accomplished by an electronic system. Furthermore, the
electronic system can reproduce or create any sound or combinations of
sounds, which have or have not been produced, that may have any possihle
musical significance. The electronic system for the production of musical
sounds has been termed electronic music synthesizer.
SPEECH, MUSIC, AND HEARING 615
One of the uses for an electronic music synthesizer is for the production
of phonograph records. Of course, the phonograph records thus produced
can be played in the conventional manner. The use of a synthesizer for the
production of musical sounds opens an entirely new field for the production
of recorded music. For example, there is the possibility of entirely new
tone complexes and combinations which cannot be achieved in conventional
instruments. Furthermore, in the case of conventional instruments, the
musician is limited to the use of ten fingers on two hands and two feet to
perform the different operations. This limitation does not exist in the
synthesizer. Conventional instruments produce various noises such as the
rushing of wind in wind instruments, bow scratch in the viol family, various
clatters and rattles in plucked and struck string instruments, and mechanism
rattle in any instrument in which keys, valves, levers, and shafts are used.
These undesirable noises do not exist in the electronic music synthesizer.
With the advent of the electronic method for the production of musical tones,
new musical compositions can be written which take advantage of the
superior characteristics of the electronic synthesizer.
From the preceding discussions and Fig. 12.55, it is evident that, in order
to synthesize any musical tone whatsoever, the synthesizer must provide
the following facilities: Means for producing a tone with any fundamental
frequency within the audio frequency range: Means for producing a tone
with any overtone structure. Means for producing a tone of any growth,
duration, or decay characteristic. Means for changing the overtone struc
ture at any time. Means for introducing a vibrato. Means for changing
the intensity of the tone. Means for providing a portamento or glide from
a tone of one frequency to a tone of a different frequency.
A schematic block diagram of an electronic music synthesizer with means
for producing all the characteristics of a musical tone outlined above is
shown in Fig. 12.56. The coded paper record controls all the functions of
the electronic music synthesizer. The information is recorded and stored
in the paper record in the form of perforations. When the paper record is
run through the machine, brushes slide over the paper record. An electrical
circuit is closed when a brush passes over a perforation. The brushes which
pass over the paper record actuate electrical circuits in the relay trees. The
use of a relay tree makes it possible to record the information on the paper
record in the binary code system. There is a separate relay tree for each
of the characteristics of a tone. In this way any of the characteristics of a
tone can be obtained or changed at any instant. Thus it will be seen that
any tone whatsoever can be produced by providing the proper information
in the coded paper record. The output of the synthesizer is recorded on a
disk record. The coded paper record and the phonograph recorder are
driven in synchronism by an interconnecting cable drive. In the complete
electronic music synthesizer two complete channels, as shown in Fig. 12.56,
are used and operated from the single coded paper record. This makes it
possible for the coded paper record to set up one channel while the other
channel is in operation and producing a tone. Furthermore, one channel
CHANNEL I
CHANNEL 2
FIG. 12.56. Schematic diagram of the electronic music synt
SPEECH, MUSIC, AND HEARING 617
can start playing a tone before the other channel stops playing a tone.
Everything is duplicated in the second channel except the twelve tuning
fork oscillators which supply the tones in one octave.
A schematic perspective view of the complete electronic music synthesizer
is shown in Fig. 12.57. The seven racks contain all the electronic equipment
for performing the different functions. The information contained in the
paper record in the form of perforations is recorded by means of a puncher
actuated by keys forming a keyboard similar to that of a typewriter.
From the fundamental knowledge and the characteristics and perform
ance of the synthesizer, as outlined above, it is not difficult to prove from
RELAY
TREES
FREQUENCY
AMPLIFIERS RELAY
MIXERS TREES
OCTAVE VOLUME
GROWTH
DECAY
TIMBRE
TEST
GLIDER
VIBRATO
FREQUENCY
SOURCE
theory what can be done with the synthesizer. However, to make a practical
demonstration, it is necessary to train an operator or synthesist. In this
connection, the music synthesizer is an entirely new genus of apparatus.
Were it merely an improvement of existing species, immediate commercial
applications would be evident. One of the uses foreseen for the music
synthesizer is to make music for sale in the form of phonograph records.
To make an artistic record, or a hit, novelty alone or technical excellence is
not sufficient. There are also other ingredients many of which are intangible.
However, the synthesizer will facilitate the production of such a hit. For
example, the synthesizer can produce any kind of sound that can be imagined.
Then if a person can image a hit, then the synthesizer will facilitate the
618 ACOUSTICAL ENGINEERING
production of the hit. The hit producing attributes are not amenable to
analysis in the acoustical laboratory, so this factor must be left out. To
demonstrate the possibilities and performance of the synthesizer, an experi
ment was conducted as follows:
Two different piano selections, namely, "Polonaise in A Flat Op. S3"
(Chopin) and "Clair de Lune" (Debussy) and a violin and piano selection,
"Old Refrain" (Kreisler), all played by famous artists and reproduced from
commercial disk records were compared with synthesized versions of the
same selections. The piano selections were recorded by Iturbi, Rubinstein,
and Horowitz, and the violin in "Old Refrain" was recorded by Kreisler.
The synthesizer versions, completed on August 5, 1953, and the identical
passages from the commercial phonograph records were recorded on magnetic
tape and intermixed and played to various people. They were asked to
tell which was synthesized and which was not. Interpreting the results by
standard statistical methods, it can be said with 70 per cent certainty that
only one out of four persons can tell which is which.
It should be noted that this work was done with only the help of conven
tional equipment to aid in the analysis which is not as complete as demanded
for synthesis; thus much of the work was done by cut and try. This points
to an important feature of the synthesizer. Manual dexterity is not required.
A synthesis once learned can be added to the fund of knowledge without
further practice to be able to perform it. The ability of the synthesist is,
therefore always increasing.
If a composer has in mind what he wants to achieve, the effects can be
obtained by means of the electronic music synthesizer, regardless of whether
he can playa musical instrument or not. The composer or musician can
produce the sound of any existing musical instrument as well as other
sounds, regardless of whether they have ever existed. The results which the
composer and musician wishes to achieve can be obtained and demonstrated
as the music is being composed and played. Once a particular result has
been obtained, it can be retained forever. Thus it will be seen that the
electronic music provides a powerful tool for the composer or musician
because he can reproduce or create any sound or combination of sounds
which have or have not been produced, that may have any musical signifi
cance.
The above experiment demonstrates the potential capabilities of the
synthesizer in the ability to copy existing selections. For an acoustical
engineer to copy the performance of famous artists in the matter of a few
weeks shows the tremendous possibilities of the synthesizer. For example,
it would be impossible for even a genius with no previous experience in
playing a musical instrument to imitate all of these artists on these two
instruments with the order of fidelity that was achieved. Thus, it seems
quite obvious that a trained or professional musician could produce great
musical renditions by means of the electronic music synthesizer.
The above experiment illustrates another use for the synthesizer, in
addition to those which have been listed, namely, the rejuvenation of old
SPEECH, MUSIC, AND HEARING 619
recordings where the master is in poor condition. That is, a new record
without distortion and noise can be made by the synthesizer.
One use envisioned for the music synthesizer is the production of music
for sale in the form of phonograph records. In order to further demonstrate
the potentialities of the music synthesizer, complete musical selections were
synthesized. A partial list of these selections indicating the style of music
and the date on which the synthesis was completed was as follows:
"Blue Skies" (Berlin) In the style of a dance band.
April I, 1952.
"Nola" (Arndt) In the style of the piano.
May 28,1952
Stephen Foster Medley, "Oh Susanna," "De Camptown Races," "My
December 12, 1952 Old Kentucky Home," "Old Black Joe," "Old
Folks at Home," and "Hard Times Come Again
-No Mo." In the style of bowed, plucked, and
struck instruments; air, mechanical, and lip
reed instruments; and percussion instruments.
"Holy Night" (Adams) In the style of the organ.
December IS, 1952.
"Fugue No.2 from Well Tem In the style of ancient struck and plucked strings
pered Clavichord" (Bach) in several variations.
July IS, 1953.
"Hungarian Dance No.1" In the gypsy style without copying any particular
(Brahms) instrument, but varying colors adapted for easy
September 3,1953 synthesis.
"Sweet and Low" (Tennyson Voice and instrumental accompaniment to show
Bamby) that voice can be synthesized.
January 15,1954
"Spoken Voice" A few spoken sentences were synthesized to show
March 26, 1954 the versatility of the synthesizer.
.- _ _ _ _ _ _ _ _ _ _ _ _
~
-
J
SPECTRAL
DECODER
LETTER SYLLABLE
ENCODER
ENCODER
_ _ _ . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ .J
SPECTRAL
DISPLAY
LETTER 5916 0 0 0
MATRIX
3951 0 0 0 0 0
2637 0 0 0 0 0
>
TYPEWRITER ~1760 0 0 0 0 0
'"51174 0 0 0 0 0
'"... 7B3
II::
0 0 0 0 0
523 0 0 0 0 0
350 0 0 0 0 0
ACTUATING
MECHANISM .02.06.01 .14.18
TIME
FIG. 12.58. Schematic diagram of the phonetic typewriter.
VOLUME
CONTROL
ACTUATING
MECHANISM
REPRODUCER - RECEIVER
FIG. 13.1. A schematic diagram of the elements of a
generalized communication system consisting of a trans
mitting or recording section and a receiving or re
producing section.
AMPLIFIER
EQUALIZER
GAIN CONTROL
AMPLIFIER
LOUDSPEAKER
SELECTOR LISTEN
~~HES SWITCH
GAIN
CONTROL
AMPLIFIER
GAIN
CONTROL
LOUDSPEAKERS
MICROPHONE
PHONOGRAPH
TURNTABLE
HORN LOUDSPEAKERS
000000000
ORCHESTRA
000
AMPLIFIERS
LOUDSPEAKERS
THEATER
FIG.13.8. Schematic arrangement of the apparatus in an
auditory perspective sound reproducing system.
4 Apps, David C., Jour. Acous. Soc. Amer.. Vol. 24, No.6, p. 660. 1952.
COMPLETE SOUND REPRODUCING SYSTEMS 633
C. Auditory Perspective System. 5-In the binaural reproducing system
the ears of the auditor are effectively transferred to the original scene of
action. A system for effectively transferring the original sources of sound
from the studio to the theater stage is shown in Fig. 13.8. The sound is
picked up by three microphones and amplified in separate channels, each
channel feeding a separate loudspeaker. The three loudspeakers are
arranged on the stage in the same positions as the microphones on the pickup
stage. The system consisting of three channels is, of course, a compromise.
To give true stereophonic reproduction an infinite number of channels must
be used. However, excellent results have been obtained in sound motion
picture reproduction employing three channels in magnetic tape sound
reproduction, and in phonograph record reproduction employing two
channels. These systems will be described in greater detail in Secs. 13.SB,
13.6B, and 13.7B.
TELEPHONE LOCAL OFFICE TELEPHONE
SUB SET r----------------, SUB SET
...--11_=:....'9'"""_ _...
LINE
_II~ ~II~
~------t-......r-II~'-------'~IIl...--+--~
L ________________ J
MICROPHONE
MICROPHONE
SUBSCRIBER SUBSCRIBER
SETS SETS
LOCAL LOCAL
OFFICE OFFICE
LOCAL LOCAL
OFFICE OFFICE
FIG. 13.11. Schematic arrangement of a telephone system with four offices and inter
connecting lines.
permit the operator to send ringing current to the subscriber station. The
further consideration of circuits, switchboards, repeaters, manual and
automatic exchanges, etc. are outside the scope of this book, and the reader
is referred to books on these subjects.
13.5. Magnetic Tape Sound Reproducing System.-The use of a
moving magnetic wire as a means for reproducing audio signals was demon
strated over a half century ago. However, it is only within the last decade
that theories, techniques, and materials have been developed which make
it possible to reproduce audio signals with a satisfactory performance from
the standpoint of frequency range, signal-to-noise, and distortion. Although
satisfactory performance was obtained from wire, magnetic tape consisting
of a plastic base coated with a thin layer of iron oxide is now almost uni
versally used in magnetic sound reproducers.
A. Monaural Magnetic Tape Sound Reproducing System.-A complete
monaural magnetic tape sound recording and reproducing system is shown
in Fig. 13.12. The magnetic recorder system may be used to record in a
studio, a room in the home, an office, a factory, etc., and from a signal input,
as for example, a radio receiver. The magnetic. recorder has been described
in Sec. 9.7. Fig. 13.12 depicts all the elements of a magnetic tape sound
recording and reproducing system. The output of the microphone or signal
source is amplified and sent through an equalizer to compensate for the
characteristics of the magnetic head and magnetic tape. The magnetic
recording head, actuated by the amplifier, magnetizes the magnetic coating
on the tape in a pattern which corresponds to the undulations in the original
sound wave. That is, the magnetic flux pattern of the recorded tape consists
of a series of magnetized sections. To overcome the nonlinearity of the
COMPLETE SOUND REPRODUCING SYSTEMS 635
tape or wire, a high-frequency signal, termed a bias, is fed to the recording
head together with the audio signal. The signal level fed to the recording
head is governed by the volume control and is measured by the volume
AMPLIFIERS
MIXERS
MONITOR
ACOUSTICS OF
MONITORING ROOM
LOUDSPEAKER
AMPLIFIER
ATTENUATOR
AMPLIFIER ATTENUATOR
EQUALIZER
VOLUME INDICATOR
AMPLIFIER
EQUALIZER
VOLUME CONTROL
VOLUME INDICATOR
AMPLIFIER
LOUDSPEAKER
ACOUSTICS OF ROOM
The signal level fed to the loudspeaker is governed by the volume control
and is measured by the volume indicator.
Perspective views of complete monaural magnetic tape sound reproducing
systems are shown in Fig. 13.13. The studio or professional magnetic
recorder and reproducer consists of microphones, monitoring console,
monitoring loudspeaker, and rack or console-type magnetic recorder or
reproducer. The system is used for professional studio recording on mag
netic tape. Portable-type magnetic tape recorders are shown in Fig. 13.13.
MICROPHONES
CONSOLE TYPE
MAGNETIC TAPE
RECORDER
MICROPHONE
PORTABLE TYPE
MAGNETIC TAPE
RECORDER
PORTABLE TYPE
MAGNETIC TAPE
RECORDER
o o o
ORCHESTRA
o 0 o
o
CHANNEL A CHANNEL B
MICROPHONES
AMPLIFIERS
GAIN CONTROLS
VOLUME INDICATORS
E9-- -~~"'j
TRAC: A ' t _
§ B
VOLUME CONTROLS
AMPLIFIERS
VOLUME INDICATORS
LOUDSPEAKERS
CHANNEL A CHANNEL B
@ LISTENER
to the point of sound pickup. For this reason the binaural sound reproduc
ing is particularly useful for sUbjective investigations.
13.6. Disk Phonograph Reproducing System.-A. Monaural Disk
Phonograph Reproducing System.-A complete disk recording system is
shown in Fig. 13.15. The first element is the acoustics of the studio. The
factors which influence the collection of sound have been discussed in Sec.
11.3. The output of each microphone is amplified and fed to a mixer. If
more than one microphone is used, as for example, a soloist accompanying
an orchestra, one microphone for the singer and one for the orchestra; the
outputs of the two may be adjusted for the proper balance by means of the
mixers. A compressor may be used to reduce a large amplitude range to
that suitable for reproduction in the home. An equalizer provides the
recording characteristic described in Sec. 9.3A4. The gain control controls
STUDIO
ACOUSTICS OF STUOIO
MONITOR
ACOUSTICS OF
MONITORING ROOM
LOUDSPEAK ER
AMPLIFIER
ATTENUATOR
EQUALIZER
the over-all signal level. The cutter actuated by the amplifier cuts a wavy
path in the revolving record corresponding to the undulations in the original
sound wave striking the microphone. A minitoring system is used to
control the recording operation.
During the past decade, in the recording of sound for phonograph disk
records, the direct recording of the sound on the lacquer original has been
replaced by first recording on magnetic tape by the system shown in Fig.
13.16. The magnetic record thus produced is then reproduced and the
output used to record the lacquer original, as shown in Fig. 13.16. The
recording and rerecording procedures are similar to that of the direct
recording. There are many advantages in the use of magnetic tape as the
master as follows: The tape can be edited and monitored to obtain the best
lacquer original. Any number of similar lacquer originals can be obtained
from the tape. The tape does not deteriorate with playing.
The processes in the record plant for the mass production of records is
depicted in Fig. 13.17. The original lacquer disk termed the "original"
COMPLETE SOUND REPRODUCING SYSTEMS 639
of Fig. 13.17A is metalized and then electroplated. The plating is separated
from the lacquer and reinforced by backing with a solid metal plate. The
assembly is termed the "master" (Fig. 13.17B). The master is electro
plated. This plating is separated from the master and reinforced by backing
STUDIO MONITOR.
ACOUSTICS OF SET
MONITORING ROOM
MICROPHONES
LOUDSPEAKER
AMPLIFIERS
TELEPHONE RECEIVERS
MIXERS
ATTENUATOR
AMPLIFIER
AMPLIFIER EQUALIZER
VOLUME INDICATOR
o 0 0 MAGNETIC TAPE
RECORDS ON REELS
RECORDER
MAGNETIC TAPE RECORO
REPRODUCING HEAD
MONITOR
ACOUSTICS OF
MONITORING ROOM
LOUDSPEAKER
EQUALIZER
AMPLIFIER
GAIN CONTROL
GAIN CONTROL
AMPLIFIER
EQUALIZER
AMPLIFIER
VOLUME INDICAlOR
with a solid metal plate. The assembly is termed the "mother" (Fig.
13.l7C). Several mothers may be made from the master. The mother is
electroplated. This plating is separated from the mother and reinforced by
a solid metal plate. The assembly is termed the" stamper" (Fig. 13 .17D).
Several stampers may be made from each mother. One stamper containing
640 ACOUSTICAL ENGINEERING
A
t::
__ .
c::r-~====~.:±~
MODULATED
GROOVE
ORIGINAL
-~~=v====~~~~
B ""- .--.", MODULATED
MASTER
RIDGE
~ODULATEO
D ~MODULATED STAMPER
RIDGE
~$TAMPERX
STAMPING
RECORD
FIG. 13.17. The steps in the process for the mass pro
duction of disk phonograph records from the lacquer
original.
placed on the other side of the record is placed in the lower jaw of a hydraulic
press equipped with means for heating and cooling the stampers (Fig. 13 .17E).
HOME A preform or biscuit of thermoplastic mate
PICKUP rial such as a shellac compound or vinylite
RECORD
TURNTABLE is placed between the two stampers. The
stampers are heated, and the jaws of the
AMPLIFIER
press are closed which presses the two
EQUALIZER
AND FILTERS
stampers against the thermoplastic material.
When an impression of the stampers has
VOLUME CONTROL
been obtained in the thermoplastic material,
AMPLIFIER the stampers are cooled which cools and sets
the plastic record. The jaws of the hydraulic
LOUDSPEAKER
press are opened, and the record is removed
ACOUSTICS OF ROOM from the press. The modulated grooves in
FIG. 13.18. Schematic arrange
ment of the apparatus in a com the record correspond to those in the original
plete disk sound reproducing lacquer disk (Fig. 13.17F). The stamping
system. procedure is repeated again and again until
sufficient records are obtained. The above
process constitutes the "mass-production system" for the production of
phonograph records.
COMPLETE SOUND REPRODUCING SYSTEMS 641
The system for the reproduction of phonograph records is shown in
Fig. 13.18. The record is turned at constant speed by the turntable. The
stylus or needle of the pickup follows the wavy spiral groove in the record
and generates a voltage corresponding to the undulations in the record
MICROPHONES
1
+
I
I
t4AGNETIC TAPE
RECORDER
RECORDING
MONITORING
LACaUER DISC
RECORDER
•
I
I
CONSOLE I
r~T1J¥t1J@5TJj@rJ
t.....
@
@
©
~ABLE CONSOLE
RECORD PRESS RECORDS PHONOGRAPHS
o o o
ORCHESTRA
o o o o
CHANNEL A CHANNEL B
MICROPHONES
AMPLIFIERS
GAIN CONTROLS
VOLUME CONTROLS
AMPLIFIERS
DISK RECORDER
CUTTERS
DISK RECORD
PHONOGRAPH TURNTABLE
PICKUPS
AMPLIFIERS
VOLUME CONTROLS
AMPLIFIERS
vOLUME INDICATORS
LOUDSPEAKERS
CHANNEL A CHANNEL B
shown in Fig. 13.20. Each channel consists of the elements of the system
shown in Figs. 13.15, 13.16, 13.17, 13.18, and 13.19. In recording an
orchestra or other sound sources spread out over an area or a sound source
moving over an area, the sound is picked up by means of two microphones
in channels A and B. The distance between the microphones and the
distance from the microphones to the sources of sound are selected to provide
COMPLETE SOUND REPRODUCING SYSTEMS 643
the most realistic sound reproduction when the loudspeakers are separated
by a distance of about ten feet in an average living room. Two-channel
stereophonic disks have been quite extensively commercialized.
13.7. Sound Motion-Picture Reproducing System. 6-A. Single
Channel Sound Motion-Picture Reproducing System.-A complete sound
motion-picture recording and reproducing system is shown in Fig. 13.21.
The first element is the acoustics of the set. The factors which influence
STUDIO MONITOR
ACOUSTICS OF
ACOUSTICS OF SET MONITORING ROOM
L.OUDSPEAKER
TEL.EPHONE RECEIVERS
ATTENUATOR
AMPL.IFIER
GAIN CONTROL.
EQUAL.IZER
VOL.UME INDICATOR
uu NEGATIVE RECORD
the collection of sound have been discussed in Sec. 11.3. The output of the
microphones is amplified and fed to attenuators termed mixers. If more
than one microphone is used, as for example, a soloist accompanying an
orchestra, one microphone for the singer and one for the orchestra, the
output of the two maybe adjusted for the proper balance. A low-pass
filter is usually used to reduce ground noise above the upper limits of repro
duction. A high-pass filter is used on speech with the lower limit placed
below the speech range. This latter expediency reduces low-frequency
noises without impairing the speech quality. An equalizer is used to accen
tuate the high frequencies to compensate for the film transfer loss at high
frequencies. A compressor is used to reduce the volume range (see Sec.
9.11). The following attenuator controls the over-all volume. The output
of the amplifier feeds the light modulator and the monitoring system. By
means of the optical system and light modulator the electrical variations are
recorded on the film into the corresponding variations in density (termed
variable density recording) or in area (termed variable area recording).
(See Sec. 9.9B.) The monitoring system is also connected to the output
of the recording amplifier. An equalizer is used to adjust the frequency
characteristic to simulate that of the ultimate reproduction. If the monitor
6 Kellogg, E. W., "History of Motion Pictures," Jour. Soc. Mot. Pic. Tel. Eng., Vo!.
64, No.6, p. 291, 1955; va!. 64, No.7, p. 356, 1955; Vo!' 64, No.8, p. 422,1955.
644 ACOUSTICAL ENGINEERING
STUDIO MONITOR
ACOUSTICS OF SET ACOUSTICS OF
MONITORING ROOM
MICROPHONES
LOUDSPEAKER
AMPLIFIERS
TELEPHONE RECEIVERS
MIXERS
ATTENUATOR
AMPLIFIER
ATTENUATOR
AMPLIFIER EOUALIZER
RECORDER
MAGNETIC TAPE RECORD
'-L----v--'--' REPRODUCING HEAD
MONITOR
AMPLIFIER ACOUSTICS OF
MONITORING ROOM
EQUALIZER LOUDSPEAKER
AMPLIFIER EQUALIZER
VOLUME INDICATOR
NEGATIVE FI LM RECORD
recording on magnetic tape by the system shown in Fig. 13.22. The pro
cedures in recording on magnetic tape for sound motion pictures are similar
to those for recording on photographic film and will not be repeated. The
magnetic tape record is reproduced and the output recorded on photographic
film by the system shown in Fig. 13.22.
COMPLETE SOUND REPRODUCING SYSTEMS 645
Perspective views of all the elements in a sound motion picture recording
system are shown in Fig. 13.23. The motion picture camera is interlocked
with the magnetic tape recorder so that synchronism of the sound and
picture will be preserved. The magnetic tape recorder is in turn interlocked
with the photographic film recorder when the sound is rerecorded on photo
graphic film. If the recording is made directly on the photographic film the
magnetic tape reproducer is by-passed.
The processes7 in the film laboratory for the mass production of motion
picture positive prints is shown in Fig. 13.24. The negative record IS
7 Crabtree, J. I., Jour. Soc. Mot. Pic. Tel. Engs., Vol. 64, No.1, p. 13,1955.
646 ACOUSTICAL ENGINEERING
LABORATORY
PROCESSING- PRINTING
THEATRE
UNDEVELOPED NEGATIVE
c@ LAMP
RECORD ON FILM
g OPTICAL SYSTEM
-....L_-'O
.....
· POSITIVE RECORD
NEGATIVE DEVELOPER
PHOTO CELL
AMPLIFIER
STAGE LOUDSPEAKERS
POSITIVE FILM RECORDS
000 ON REELS ACOUSTICS OF THEATRE
FIG. 13.24. The steps in the process FIG. 13.25. Schematic arrangement of the
for the production of motion-picture apparatus in a system for the reproduction of
positive film from the negative film. sound from motion-picture film.
and fed to equalizers and filters. A low-pass filter is used to cut out the
ground noise due to film above the upper limit of reproduction. An equalizer
is used to adjust the frequency characteristic to that suitable for the best
reproduction in the theater. The attenuator is used for adjusting the level
of reproduction. The output of the power amplifier feeds the stage loud
speakers and monitoring loudspeaker. The monitoring loudspeakers and
the attenuator are located in the projection booth. As a matter of fact, the
entire system, save for the stage loudspeakers, is located in the projection
booth. A dividing network and a two-channel loudspeaker system are
shown in Fig. 13.25. Of course, any type of suitable loudspeaker described
in Chapter VII may be used. The action of a sound motion-picture repro
ducer in a theater has been discussed in Sec. 11.2G and will not be repeated
here.
Perspective views of the elements of a theater sound motion-picture
reproducing system are shown in Fig. 13.26.
COMPLETE SOUND REPRODUCING SYSTEMS 647
PROJECTOR A PROJECTOR B
VOLTAGE
'11..._..L.....I AMPLIFIER
AMPLIFIERS
STAGE LOUDSPEAKER
MICROPHONE
BOOM
PERAMBULATOR
STUDIO
ACOUSTICS OF STUDIO
MICROPHONE
EQUALIZER
GAIN CONTROL
VOLUME INDICATOR
RECORDER
ATTENUATOR
VOLUME INDICATOR
O0 0
• ••
RECORDED SOUND ON MAGNETIC STRIPS AND
PICTURE ON POSITIVE FILM ON REELS
PROJECTOR A PROJECTOR B
::j1::=+J:!§~~::E:!.IVOLTAGE AMPLIFIERS
POWER
AMPLIFIERS
AUDITORIUM
LOUDSPEAKERS
AMPLIFIERS
STAGE LOUDSPEAKERS
STUDIO
ACOUSTICS OF STUDIO MONITOR
ACOUSTICS OF
MICROPHONES MONITORING ROOM
AMPLiFIERS LOUDSPEAKER
MIXERS AMPLiFIER
GAIN CONTROL
EQUALIZER
VOLUME INDICATOR
LINE
ISOLATING AMPLIFIER
REPEATER
TRANSMITTER TO TRANSMITTER
REPEATER
REPEATER
AMPLIFIER
VOLUME INDICATOR
HOME
ACOUSTICS OF ROOM
LOUDSPEAKER
A F AMPLIFIER
VOLUME CONTROL
21'lJl DETECTOR
IF AMPLIFIER
lar DETECTOR
OSCILLATOR
RF AMPLIFIER
TRANSMITTER
LINE
EQUALIZER
LIMiTiNG HOME
AMPLIFIER ACOUSTICS OF ROOM
AMPLIFIER
AF AMPLIFIER
VOLUME INDICATOR
2~ DETECTOR
FREQ. MODULATED
OSCILLATOR
DISCRIMINATOR
FREQUENCY LIMITING
MULTIPLIERS AMPLIFIER
RF AMPLIFIER IF AMPLIFIER
RF POWER 1M DETECTOR
AMPLIFIER OSCILLATOR
RF AMPLIFIER
ANTENNA
MICROPHONES
STUDIO
MONITORING
CONSOLE
MONITORING
LOUDSPEAKER TRANSCRIPTION
TURNTABLE
MONITORING
LOUDSPEAKER
MICROPHONE
BOOM
The receiver is made very small and unobtrusive so that it may be placed in
the outer ear cavity and coupled directly to the ear canal. The bone
conduction receiver is also of the magnetic type with a mechanical trans
former that couples the vibrating system to the mastoid bone, see Sec. 9.13.
The electrical impedance of the receivers is relatively low, being about 100
ohms. Therefore, a transformer is used to couple the output of the power
vacuum tube to the receiver.
During the past two years, transistor-type hearing aids have been
developed and commercialized. The advantage of the transistor amplifying
MICROPHONE
AMPLIFIER MICROPHONE
BEHIND GRILL
TONE CONTROL
VOLUME CONTROL
AMPLIFIER
TELEPHONE RECEIVER
MAGNETIC TAPE
REPRODUCER
POWER
AMPLIFIER
RECORD PLAYER
in Fig. 13.36. The components may be placed in a single cabinet and termed
a combination sound reproducer. The components may be used in a
custom-type sound reproducing system in which one or more of the com
ponents may be housed in separate cabinets or the components may be
placed in special compartments as the part of a house or apartment.
14
OF INFORMATION
14.1. Introduction.-Speech always has been and still remains the most
important method for the transmission of information between individuals.
Manual signals involving the arms, hands, face, eyes, and mouth are second
ary to the voice. To augment the process of communication man devised
sound generators in the form of noise makers, sound signalling devices, and
musical instruments and visible signalling, as exemplified by fire, smoke, and
flags. In a further expansion of communication, orthography was evolved,
using symbolic notation which made it possible to record and store informa
tion and to transport information by symbolic recordings to a distant point.
The advent of the printing press about 400 years ago supplied a means for
the mass dissemination of information in the form of the printed page. Still
further, photography was developed which provided a process for recording
a scene in the form of a picture. A further development of photography
was the motion picture, in which a series of photographs of a changing scene
are taken and when projected depict a reproduction of the original scene.
The telegraph was the first electrical system for the coded transmission of
information over large distances by the use of wires for immediate decoding
into the message at the recording end. The invention of wireless telegraphy
at the turn of the century provided a system for the transmission of coded
messages through the ether, thereby eliminating the use of a wire trans
mission line. The transmission of pictures and the printed page by facsimile,
either by wire or through the ether, was developed and introduced about
three decades ago. The inventions and developments in the field of sound
and scene reproduction during the past half century, as exemplified by the
telephone, phonograph, sound systems, radio, sound motion picture, and
television have made it possible to pick up sound or sound and scene and
reproduce it or them, either at the same point or some other point either
at the same time or some subsequent time. The impact of the telephone,
phonograph, radio, sound motion pictures, and television upon the dissemina
tion of information, art, and culture has been tremendous. However, the
list will not end with the communication systems that have been enumerated.
There will be additional systems and adjuncts in the field of communica
tion which have not been developed that will ultimately improve, modify,
657
658 ACOUSTICAL ENGINEERING
The disk sound recording and reproducing system has also been used for
home and commercial applications as well as for the prerecorded records.
Magnetic tape phonographs or reproducers have been developed to a high
order of excellence during the past decade. In the magnetic tape system,
the information is recorded on a ferromagnetic coating on a plastic base tape
in the form of magnetized elements corresponding to the electrical input.
In reproduction, the stored magnetic information on the tape is converted
into the corresponding electrical variations. The tape system has been
applied to recording and reproducing sound for home and commercial
applications. The commercial applications include sound motion pictures,
television signals, master recordings, transcriptions, and memory systems
for computors.
14.16. Radio.-A radio broadcasting system consists of a transmitter
and a receiver. At the transmitter, the sound vibrations in the studio which
represent the program material are converted into the corresponding electrical
variations by the microphone. These electrical variations modulate a
radio-frequency system which is coupled to the antenna where the resultant
signal is radiated. A small part of the signal from the transmitter is picked
up by the radio receiver antenna. This signal is amplified and demodulated
into the original audio electrical signal. The audio electrical signal is fed
to the loudspeaker which converts the electrical variations into original
sound vibrations.
Two types of radio broadcasting are in use today, namely, amplitude
and frequency-modulated systems.
The radio, like the phonograph, is a home-type instrument. There are
3200 broadcasting stations, 125,000,000 home-type receivers and 35,000,000
automobile receivers in this country. Practically every family owns a
radio receiver, which means that they can hear famous persons in govern
ment, in industries, in the schools, in the theater, and in all manner of
musical aggregations.
In addition to conventional broadcasting, there are radio communication
systems of all types, involving stationary land and mobile land, water, and
air systems.
14.17. Sound Systems.-Sound system is a term used to designate the
combination of a microphone, amplifier, and loudspeaker and auxiliary
equipment. Sound systems include hearing aids, public address, sound
reinforcing, electric megaphone, intercommunicating, announce, paging,
and call systems.
14.18. Facsimile.-The term, facsimile, is used to designate a system
for the transmission of graphic material in the form of a photograph, draw
ing, or printed page. In existing commercial facsimile systems, the informa
tion is transmitted bit by bit in a sequential manner. This is done by
dividing the original into a large number of elementary areas. Then signals
are transmitted that correspond to the relative light shades of these areas.
Such shades are reproduced in graphic form at the receiving point. The
elementary areas are reproduced in a sequence corresponding to that at the
MEANS FOR THE COMMUNICATION OF INFORMATION 663
pickup at the transmitter. The fidelity of reproduction depends upon the
correspondence in amplitude of the number of these elementary areas in the
picture.
14.19. Sound Motion Picture.-The addition of sound to the motion
picture made this type of expression complete. It was the first commercial
system in which picture and sound were synchronised and reproduced at the
same time. In some of the early systems the sound was recorded and repro
duced by means of a disk record which was synchronized with picture.
During the past twenty-five years, the photographic method for recording
and reproducing the sound has been universally employed. The photo
graphic reproduction of sound is accomplished by recording information in
the form of a narrow band, termed the sound track, located between the
sprocket holes and the picture. The recording system produces photo
graphic undulations in the sound track which correspond to the original
sound. In reproduction, these photographic undulations are converted
into the corresponding electrical variations which, in turn, are converted
into sound by the loudspeaker.
Recently, stereophonic sound systems employing multiple sound channels
have been introduced on a wide commercial scale. In one system, four
separate magnetic sound tracks are used. Two of the tracks are outside
the sprocket holes and the other two are inside the sprocket holes. The
output from the three separate sound channels and the control channel are
recorded on the four magnetic tracks. In reproduction, four separate
magnetic head, amplifier, and loudspeaker systems are used to reproduce
the sound in perspective. A wide screen is used to project the picture.
The three loudspeakers are located behind the screen.
There are 18,000 motion picture theaters in this country. The average
weekly attendance is 90,000,000 persons.
14.20. Television.-Television is the latest commercial system in which
picture and sound are reproduced at the same time. The fundamental
objective of a television system is to extend the sense of sight and hearing
beyond their natural limits. At the pickup point the camera scans
the scene and thereby divides the scene into elementary projected areas. The
light reflected from the elementary areas is picked up by the camera in the
scanning process and converted into the corresponding electrical variations.
At the receiving end the electrical variations are converted into the cor
responding light variations and depicted upon some sort of screen in the
pattern similar to that at the pickup. The gradations of the light produced
upon the screen are proportional to the reflected light intensities of the scene
being televised, thereby producing a picture corresponding to the televised
scene.
14.21. Undeveloped Systems for the Transmission of Information.
-In the over-all plana of the systems for the transmission of information,
there are many that have not been developed or commercialized. These
3 Peterson, Gordon E., Jour. Speech and Hearing Disorders, Vol. 18, p. 116, 1953.
664 ACOUSTICAL ENGINEERING
6 Dudley R., Jour. Acous. Soc. Amer., Vol. 11, No.2, p. 169, 1927.
6 Olson and Belar, Jour. Acous. Soc. Amer., Vol. 27, No.3, p. 595.1955.
LOUDSPEAKER
~ MUSIC LS-v1
~SYNTHESIZER~
3 LOUDSPEAKER
ICODEL-J SP E ECH L-..I"V1
t..=:..:..J I SYNTHESIZERI ~
4 LOUDSPEAKER
MICROPHONE 6
UNDERWATER SOUND
k = 217
,\'
,\ = wavelength, in centimeters,
p = density, in grams per cubic centimeter, and
c= velocity of propagation, in centimeters per second.
If the ultimate acoustical resistance has been obtained, the value of the
acoustical resistance for 1 square centimeter is
rAW = pc = 144,000 15.3
where p = density of water, in grams per cubic centimeter, and
c = velocity of sound in water, in centimeters per second.
For a peak sound pressure of 106 dynes per square centimeter the power
output per square centimeter is 3.6 X 106 ergs or .36 watt. If the loud
speaker is lowered to a depth of 33 feet the output will be quadrupled.
UNDERWATER SOUND 671
The preceding presents the idealized picture of sound transmission in
water. Transmission of sound in the sea does not conform to this simple
description. There are many factors2 that influence the transmission of
sound in the sea. A few of the factors are as follows: The surface of the sea,
the bottom of the sea, the inhomogeneity of the sea.
The surface of the sea affects the transmission by the reflection produced
at the surface. Since the change in acoustical impedance from the sea to
the air is about 3000, the reflection is practically perfect. If the sea were
perfectly plane, account of the surface reflection could be accounted for by
the establishment of an image transmitter of opposite phase above the
surface. However, the sea is never quiet, with the result that there are
waves which disturb the plane surface condition.
In shallow water the bottom of the sea introduces another troublesome
boundary condition. Unlike the surface of the sea, the shape of the bottom
is not known. In addition, the absorption and reflection varies over wide
limits.
Sea water is not homogeneous. There are both vertical and horizontal
variations in the velocity of propagation of sound in the sea. Vertical
variations are perhaps more pronounced and produce refraction of the sound
waves. See Sec. 1.11. The bending of the sound waves by refraction
reduces the efficiency in echo ranging. In the case of severe refraction, the
distance of satisfactory operation may be seriously reduced.
The surface reflection, the bottom reflection, and the reflection due to
variations in the sea produce reflections. This lends to what appears to be
reverberation in the sonar receiver.
The reflection and refraction effects also produce variation in the received
sound in both passive and echo systems. The roll and pitch of the sending
and receiving vessel produce fluctuations in the received signal.
The anomalies of the sea outlined above are of great importance in both
undersea warfare as well as in commercial navigation. The application of
some of the newer principles in the fields of communication such as correla
tion techniques and recognition procedures should be of value in solving
many of the problems in the transmission of sound in the sea.
As in all transmitting media, noise produced in the medium influences the
reception of information. The different types of noise includes the ambient
noise in the sea due to natural causes and noise produced by ships and other
sound sources. The ambient noise in the sea is a function of the state of
agitation of the sea and the wind velocity. Various undersea life also
produces discrete sounds and noises. In general, underwater signals are
received from hydrophones suspended below the ship. Under these condi
tions the noise produced by the ship influences the reception of underwater
signals. The principal source of noise in water, produced by a ship, is due
to cavitation at the propellers. As in all other underwater phenomena,
noise in the sea is a very complex subject.
2 A Survey Report on Basic Problems of Underwater Acoustics Research, Committee
on Undersea Warfare, National Research Council, Washington, D.C., 1950.
672 ACOUSTICAL ENGINEERING
~rH~
ELECTRICAL CIRCUIT MECHANICAL NETWORK
., 25
c 20
w
~ 15
o
v tv """
t...... r-
g; 10
w
II:. 5
o~
• 2 4
•103 2 4 8 ,0 • 2
CROSS-SECTIONAL VIEW FREQUENCY
FIG. 15.1. Cross-sectional view, electrical circuit, mechanical network, and pressure
response frequency characteristic of a direct radiator dynamic projector. In the
electrical circuit: ZEM = the motional electrical impedance. L and rEI = the
inductance and electrical resistance of the voice coil. rEG = the electrical resistance
of the electrical generator. e = the voltage of the electrical generator. In the
mechanical network: mi = the mass of the diaphragm and coil. rMI and C MI = the
mechanical resistance and compliance of the suspension system. C M2 = the com
pliance of the air chamber behind the diaphragm. m2 and "M2 = the mass and
mechanical resistance of the water load. ma and rMa = the mass and mechanical
resistance of the aperture in the pressure equalizer. CM3 = the compliance of the
pressure equalizer.
~'~'5
MECHANICAL CIRCUIT
25
f\
m B
020
~ 15
/ n
z
o 10
0-
III
A ./
'"
IX 5
FIG. 15.2. Cross-sectional view, mechanical circuit, and voltage response fre
quency characteristic of a condenser hydrophone. In the mechanical circuit:
mo and rHo = the mass and mechanical resistance of the water load upon the
diaphragm. mI. rMI. and CMI = the mass, mechanical resistance, and compliance
of the diaphragm. m2' "H2' and CH2 = the mass, mechanical resistance, and
compliance of the air film between the diaphragm and the back plate. In
the graph: A = the open-circuit voltage response frequency characteristic
of the condenser hydrophone. B = the output voltage response characteristic of
the combination of a two-stage amplifier and condenser hydrophone for an output
impedance of 250 ohms. 0 db = 1 microvolt per dyne per square centimeter.
m r. m r.
t=s
CM' C M•
MAGNET
rH~
ELEC. CI RCUIT MECHANICAL CIRCUIT
25
~ 20
.L""'>.
r-..
OJ 15
!£o
0.. 10
'"\r
v ....,. ......
<II
w
a:
s
V r-..
"'I
<\0 15 20 30 40 50 60 80 100
CROSS-SECTIONAL VIEW fREQUENCY IN KILOCYCLES
FIG. lS.3. Cross-sectional view, the electrical circuit, the mechanical circuit,
and the pressure response frequency characteristic of a high-frequency direct
radiator dynamic projector. In the electrical circuit: ZEM = the motional
electrical impedance. L and rEI = the inductance and electrical resistance
of the voice coil. rEG = the electrical resistance of the generator. e = the
voltage output of the generator. In the mechanical circuit: mi = the mass
of the diaphragm. rMI and C MI = the mechanical resistance and compliance
of the suspension system. C M2 = the compliance of the air chamber behind
the diaphragm. m2 and rM2 = the mass and mechanical resistance of the
water load.
15.4
If the current in the coil is sinusoidal, then the expression for the current
can be written
i = i max sin wt 15.5
where i max = amplitude of the current, in abamperes,
w = 24,
f = frequency, in cycles per second, and
t = time, in seconds
Substituting equation 15.5 for the current in 15.4, the force on the arma
ture, in dynes, is
j M -_ 4?TAa
C2' 2 . 2
2 ~ max SIll wt 15.6
c:==s r H4
C C
rEG rEI L
f ~
COIL
0- 10
z
w / \
i3 -I 5
;;:
...
"'-2 o / 1\
-2
300
/ 400
\
500
SECTIONAL VIEW fRONT VIEW fREQUENCY
FIG. 15.4. Cross-sectional and front views, electrical circuit, mechanical circuit, and
efficiency frequency characteristic of a magnetic projector. In the electrical circuit:
ZEM = the motional electrical impedance. L and rEI = the inductance and electrical
resistance of the coil. rEG = the electrical resistance of the electrical generator.
e = the voltage output of the electrical generator. In the mechanical circuit: mv
rMI' and C,VI = the mass, mechanical resistance, and compliance of the diaphragm.
CM2 = the compliance of the air chamber behind the diaphragm. m2 and YM2 = the
mass and mechanical resistance of the water load.
ZEM = 15.8
15.9
From equations 15.8 and 15.9 it will be seen that the efficiency increases
with the power input. This characteristic is typical of unpolarized driving
systems.
15.7. Ma~netostriction Projector.-A magnetostriction projector is a
projector in which a diaphragm is driven by the mechanical forces generated
in a ferromagnetic rod possessing magnetostrictive properties. A cross
sectional view of a magnetostriction projector is shown in Fig. 15.5. A rod
of nickel is coupled to the diaphragm. The polarizing magnetic flux in the
rod is supplied by a permanent magnet. The coil surrounding the nickel
rod supplies the alternating magnetic flux. The resultant magnetic flux
in the rod is the sum of the polarizing and alternating flux. The variation of
magnetic flux in the nickel rod causes it to vary in length. The electrical
circuit and the mechanical network of a magnetostriction projector are
shown in Fig. 15.5.
The driving force,S in dynes, generated in the rod is given by
47rNiK .
1M = -R-- sm wt 15.10
i = current, in abamperes,
w = 27Tj,
t = time, in seconds.
HYDROPHONE
~ CM1 my'4
rl
PROJECTOR
Q
ELECTRICAL CIRCUITS
fM
l)\----r---,
MECHANICAL
o
NETWORKS
<Il
o -5 /"
W
<Il
6 -10 II \
0..
<Il
~-15
/ \
-2~2
/ 23 24 25
\
26
CROSS -SEC TIONAL VIEW FREQUENCY IN KC.
MAGNET
END PLATE
NICKEL
CYLINDER
COIL
fashion over the tube and the polarizing and alternating flux paths are
circumferential in the tube. The fundamental resonant frequency of a
radiator of this type without any external load is given by
c
f = 27TR 15.19
HYDROPHONE f ~£I S
PROJECTOR f 1£1 ~
ELECTRICAL NETWORKS MECHANICAL NETWORKS
o.---.-~~-,---,
co
Cl _5 f---H--f-~f--l
w
~ -IOf--If-----;f--l'r--l
o
n.
If) -151---->f-f--lf--l----'H
w
a:
-20"-------'__--L__...L__..JI
450 500 550
CROSS- SECTIONAL VIEW FREQUENCY IN KC.
1
ZEI = jwCEl'
15.29
m
o
DIA""I1K~\I.>M
50 55
CROSS- SECTIONAL VIEW FREQUENCY IN KC.
quency charactenstics for steel and aluminum blocks are shown in Fig.
15.9. As would be expected, blocks with a lower density yield a wider
frequency band. The quartz crystal sandwich projector conserves quartz
since only a thin layer is used as compared to the large block for the all
quartz subaqueous loudspeaker. In addition, the electrical impedance is
lower. With metal blocks of the same density as quartz, the shape of the
response frequency characteristic is approximately the same as a solid block
of quartz.
The system shown in Fig. 15.9 may also be used as a hydrophone. The
shape of the response frequency characteristic is approximately the same
for the hydrophone as that shown in Fig. 15.9 for the projector.
U Olson, "Dynamical Analogies," p. 169, D. Van Nostrand Company, Princeton,
N.J., 1943.
686 ACOUSTICAL ENGINEERING
o
:!l -5
/"\
~ ~IO
/ 1\
z
~ -15
/ \
'"~-20 / \
-25
22 23 24 25 26
\ 27
CROSS- SECTIONAL VIEW fREQUENCY IN KC.
passive sonar was an all acoustic system a barium titanate ceramic hydro
record paper. The process is repeated every few seconds. In this way, a
continuous record of the depth of the sea is obtained. Depths from 5 feet
to several thousand feet may be measured. Both magnetostriction or
crystal projectors and hydrophones have been used for sonic depth indicators.
15.16. Echo Direction and Ran~in~ Sonar. 24-The position of sub
merged submarines may be determined by means of sonic echo ranging
equipment. The general arrangement and the schematic diagram of echo
direction and ranging equipment are shown in Fig. 15.13. The oscillator
operates at a fixed frequency somewhere between 15 and 50 kilocycles, the
customary frequency being about 25 kilocycles. The sequence switch
connects the power amplifier to the projector-hydrophone used as a pro
jector for a few milliseconds. Then the sequence switch connects the
amplifier to the projector-hydrophone used as a hydrophone. If a reflected
sound impulse is received, it will be indicated as a "pip" on the cathode-ray
tube or other visual indicating means and reproduced as a "ping" on the
air loudspeaker. The horizontal sweep system drives the cathode-ray spot
to the right starting at zero when the sound impulse is sent out. The
horizontal axis of the tube is calibrated in yards. The diameter of the
diaphragm of the projector-hydrophone is usually about 5 to 10 wavelengths,
see Sec. 15.7. Since the same unit is used for both sending and receiving,
the ordinates of the directional patterns shown in Fig. 2.13 must be squared.
It will be seen that under these conditions the directional pattern of a
24 Evans, R. J., Electronics, Vol. 19, No.8, p. 88, 1946.
UNDERWATER SOUND 689
diaphragm 6 wavelengths in diameter is quite narrow and it is possible to
determine the direction of the submerged submarine quite accurately.
The combination of direction and range gives the position of the submerged
submarine.
In the detection of submarines by echo ranging sonar, use can be made of
the Doppler effect to determine the velocity of the target with respect to the
WATER SURFACE
PROJECTOR
HYDROPHONE
\ \
SOUND SUBMARINE
PULSE
CATHODE-RAY TUBE
transmitting point, see Sec. 1.10. The frequency ie, in cycles per second,
of the returning echo is given by
ie = is [U + 2 (VS cos: 8 - V t cos 8t )] 15.31
SUBMARINE
UUTGOING
_ SOUND
PULSE
ELEMENT
~ SUBMARINE
.......... REFLECTEO
SOUND
PULSE
DELAY
~E)
BEAM FORMING
AIR
LOUDSPEAKER SWITCH
ULTRASONICS
STATOR ROTOR
AIR
INPUT
~ 1~~~lltWATER COIL
COOLING
-----
DIAPHRAGM
WATER
COOLING
COIL
TERMINALS
OIL
RELEASE
FIG. 16.2. Cross-sectional view of a magnetostriction ultra
sonic generator.
OIL
QUARTZ TERMINALS
+ t +
-~
t t
d-
PLATE CYLINDER
I t
C
SPHERICAL
WATER COOLING
COIL
OIL
RMINALS
21 Williard, G. W., Jour. A cous. Soc. A mer., Vol. 25, No.4, p. 669, 1953.
22 Galloway, William J., Jour. Acous. Soc. Amer., Vol. 26, No.5, p. 849, 1954.
23 Mellen, Robert H., Jour. Acous. Soc. Amer., Vol. 26, No.3, p. 356, 1954.
24 Connolly and Fox, Jour. Acous. Soc. Amer., Vol. 26, No.5, p. 843, 1954.
26 Solner, K., Jour. Phys. Chem., Vol. 42, p. 1071, Nov. 1938.
27 Freundlick and Gillings, Faraday Soc. Trans., Vol. 35, p. 319, Feb. 1939.
28 Bergmann, "Ultrasonics," John Wiley and Sons, New York, N.Y., 1938.
33 Bondy and Sollner, Faraday Soc. Trans., Vol. 32, p. 556, March 1946.
35 Schmid and Ehret, Zeit. f. Electrochem, Vol. 43, p. 869, Nov. 1937.
36 Bergmann, "Ultrasonics," John Wiley and Sons, New York, N.Y., 1938.
38 Hiedmann, Egon A., Jour. Acous. Soc. Amer., Vol. 26, No.5, p. 831, 1954.
39 Hueter and Bolt, "Sonics," John Wiley and Sons, New York, N.Y., 1955.
40 Sollner and Bondy, Faraday Soc. Trans., Vol. 32, p. 616, April 1936.
41 Number of Authors, Faraday Soc. Trans., Vol. 32, p. 104, August 1936.
45 Hiedmann, Egon A., Jour. Acous. Soc. Amer., Vol. 26, No.5, p. 831. 1954.
ULTRASONICS 699
larger ones. The larger ones rise to the surface and are expelled. This
use of ultrasonics should lead to an improvement in castings where the
presence of bubbles is very objectionable.
16.7. Chemical Effects of Ultrasonics. 46 ,47,48,49,50,51,52,53_A large
number of experiments have been conducted on the effect of intense sound
waves upon chemical reactions. Certain types of chemical reactions have
been speeded by the application of intense sound waves. However, in some
cases it is difficult to isolate the thermal effects due to the sound and the
effects due to the sound alone. Another chemical effect is the breaking down
of molecules. For example, a chain molecule of starch has been broken
into six fragments. The application of intense sound waves to speed up
the aging of whiskey has been suggested. The explanation is that in the
aging process there is a gradual change in the structure of complex molecules
which could be accomplished in a relatively short time with the application
of sound.
16.8. Biological Effects of Ultrasonics. 54 ,55,56,57,58,59,60,61,62,63,64_
Ultrasonics have a very destructive effect upon small living organisms.
Small fish have been killed by high-power echo ranging and sounding devices.
Ultrasonics have been used in the extraction of antibodies secreted in
the cells of pathogenic bacteria. These antibodies are used in serums for
immunization against typhoid and other diseases. The bacterial cell walls
are broken down by the application of supersonic waves and the antibodies
are set free. The cell walls of the bacteria are separated from the antibodies
by centrifuging.
It appears that bacteria can be destroyed by ultrasonics. The bacteria
in milk have been reduced by the application of supersonics. This indicates
that milk can be sterilized by ultrasonics.
Another application in medicine is the use of sound to produce stimula
tion within the body. Therapeutic effects of a different nature but similar
to those produced by heat and radio·frequency diathermy may be obtained.
46 Porter and Young, Jour. Amer. Chem. Soc., Vol. 60, p. 1497, 1938.
47 Benthe H., Zeit.f. Phys. Chem., Vol. 163, p. 161, Feb. 1933.
48 Schmitt, Johnson, and Olson, Jour. Amer. Chem. Soc., Vol. 51, p. 370, 1929.
49 Barrett and Porter, Jour. Amer. Chem. Soc., Vol. 63, p. 3434, Dec. 1941.
50 Weissler, Alfred, Jour. Acous. Soc. Amer., Vol. 25, No.4, p. 651, 1953.
51 Yeager and Hovorko, Jour. Acous. Soc. Amer., Vol. 25, No.3, p. 443,1953.
52 Hueter and Bolt, "Sonics," John Wiley and Sons, New York, N.Y., 1955.
59 Bergmann, "Ultrasonics," John Wiley and Sons, New York, N.Y., 1938.
64 Danner, Ackerman, and Fring,Jour. Acous. Soc. Amer., Vol. 26, No.5, p. 731, 1954.
COIL
NICKEL
LAMINATIONS
MAGNET -t==-:l..L.---Y
PERSPECTIVE VIEW
73Kearney, Thomas, Jour. Acous. Soc. Amer., Vol. 26, No.2, p. 244, 1954.
74 Hueter and Bolt, "Sonics," John Wiley and Sons, New York, N.Y .• 1955.
76 Mattiat, Oskar, Jour. Acous. Soc. Amer., Vol. 26, No.2, p. 241, 1954.
77 Bueter and Bolt, "Sonies," John Wiley and Sons, New York, N.Y., 1955.
80 Hueter and Bolt, "Sonics," John Wiley and Sons, New York, N.Y., 1955.
82 Bergmann, "Ultrasonics," John Wiley and Sons, New York, N.Y., 1938.
84 Hayes, H. C., Jour. Acous. Soc. Amer., Vol. 8, No.4, p. 220, 1937.
85 Firestone, F. A., Jour. Acous. Soc. Amer., Vol. 17, No.4, p. 363, 1946.
87 Rueter and Bolt, "Sonics," John Wiley and Sons, New York, N.Y., 1955.
ULTRASONICS 703
equipment is quite small and portable. Furthermore, other intervening
or adjacent components need not be removed to carry out the tests.
A syst«m for detecting flaws in tires by the use of ultrasonics has been
developed. The tire is immersed in water and the transmission of an
ultrasonic wave through the tire is obtained by a projector and hydrophone
combination. Since the characteristic acoustical impedance of rubber and
water is practically the same, there will be very little attenuation or other
anomalies in the transmission of the ultrasonic wave except in the case of
a flaw or defect i:l the rubber.
16.16. Ultrasonic Delay Lines and Filters.-Delay lines 89 ,9o for the .
storage of pulses one microsecond in length and for periods up to 2000 micro
seconds have been developed. These delay lines are used for the storage of
radar pulses from one pulse to the next. Both mercury and solid lines have
been used. Quartz crystal transducers are used for the transmitter and the
receIver.
Ultrasonic band-pass filters 91 for use in the intermediate frequency
amplifiers in radio receivers consist of mass and compliance elements.
Magnetostriction transducers are used for the transmitter and the receiver.
The outstanding characteristic is the very high attenuation over a very narrow
frequency range at the upper and lower cutoff frequencies. For example, a
band-pass filter with a pass band of 6 kilocycles at 100 kilocycles shows 4S
decibels attenuation in 1 kilocycle at the cutoff frequency.
89 Arenberg, David L., Jour. Acous. Soc. Amer., Vol. 20, No. 1, 1948.
90 Rueter and Bolt, "Sanies," John Wiley and Sons, New York, N.Y., 1955.
91 Burns, Jr., Leslie L., RCA Review, Vol. 13, No.!, p. 34, 1952.
INDEX
studios, 542
piston source, 43
system, 650
nonuniform, 44
uniform, 43
configuration, 169
broadcasting studios, 542
437
sound stages, 553
435
Combination, tones, 594
423
horn and phase inverter loudspeaker,
reciprocity, 430
facsimile, 662
pistonphone, 424
orthography, 660
reciprocity, 427
phonograph,661
secondary, 433
radio, 662
electrical, 75, 81
system, 662
Cellular horn, 50
motion picture, 643
free plate, 66
radio, 650
membrane, 62
telephone, 633
~upported plate, 66
television, 653
center, 66
Compliance, mechanical, 76, 81
outside, 66
rotational, 76, 82
INDEX 707
Compound, direct radiator loud
Di rectional, characteristics (Cont.)
speaker, 157
cone surface, 53
Compressibility,S
double source, 32
Compressors, 403
exponential horn, 47
Condensation, 6
line, circular, 43
istics, 48
nonuniform, 38
finite, 106
tapered, 37
infinite, 102
uniform, 36
pickup, 359
point sources. 35
Current, 74
efficiency, 331
surface source, 50
Directivity index, loudspeaker, 211
103
Dispersion of sound, 500
Cycle, 3
absorption in passing through air,
Cylinder, pulsating, 95
535
infinite, 101
conventional sound absorber, 501
waves, 28
electronic sound absorber, 511
Decibels, 15
general announce system, 531
microphone, 262
190
frequency, 574
paging system, 531
intensity, 575
power requirements, 524
Diffraction, 17
public address system, 525
electrical, 85, 86
mobile, 534
mechanical, 85, 86
living room, 533
rotational, 85, 87
resonator, sound absorber, 509
surface, 44
theater, 523
708 INDEX
mechanism, 503
elements, 77
table, 502
impedance, 75
theaters, 521
musical instruments, 612
outdoor, 528
resistance, 75, 78
table, 536
Electronic, microphone, 273
cone, 183
Electrostatic, actuator, 426
horn, 223
microphone, 253, 307
phonograph,373,472
rotational, 77
speech,580
Energy density, sound, 15
Doppler effect, 17
motion, 5
Double source, 32
wave, 6
Duration, 576
Exponential, connectors, 112
diator loudspeaker)
finite, 108
microphone, 260
infinite, 103
pickup, 367
multiple flare, 114
Dyne, 73
Facsimile, 662
defender, 414
Flux magnetic, 189
mechanism, 558
density, 200
tor, 126
Frequency, 3
horn, 212
fundamental, 4
Electrical, abampere, 74
range, music, 579
abohm,77
musical instruments, 579, 587
analogies, 71
fundamental, 610
capacitance, 75, 81
total, 579, 587
INDEX 709
monaural, 600
loudspeaker (see loudspeaker horn)
stereophonic, 603
mani fold sections, 114
reproducers, 598
multiple flare, 114
fundamental, 610
Hyperbolic horn, 104
resonant, 122
Impedance, acoustical, 76
Gauss, 200
mechanical, 478
ers) electrical, 75
free, 99
Harmonic, 4
tube, 97
Head, diffraction, 20
pulsating cylinder, 95
magnetic, 385
pulsating sphere, 93
carbon, 406
slit, 89
testing, 495
tube, 88
transistor, 406
vibrating strip, 96
mechanism, 558
Inductor, microphone, 363
acuity, 569
localization, 576
Inertance, 76, 80
nonlinearity, 573
response, 569
Horn, cellular, 50
Intensity of sound, 15
equation, 100
conical, 106
microphone, ~64, 291
cylindrical, 105
Lapel microphone, 329
exponential, 108
Large surface microphone, 321
general, 100
Larynx, 560
cylindrical, 101
Lavalier microphone, 329
710 INDEX
Lens, acoustic, 19
Loudspeaker, di rect radiator (Cont.)
microphone, 321
locations (Cont.)
threshold, 577
television receivers, 171
Limiters, 403
measurements, 438
Loudness, 570
nonlinear distortion, 183
baffles, 149
large, 149
small space, 167
irregular, 149
transient response, 178
diator, 163
diaphragms, 242
compound, 157
direct radiator combination, 163
condenser, 205
efficiency, 212
distortion, 183
diaphragm, unloaded side, 221
cone, 183
initial, 213
subharmonic, 186
folded, 240
ionophone, 244
efficiency, 126
233
electrostatic, 205
single channel, 238
feedback, 168
nonlinear distortion, 223
labyrinth, 162
subharmonic, 228
ments, 177
throat, effect of, 223
phonographs, 173
personal radio receiver, 239
INDEX 711
Loudspeaker (Cont.)
Measurement of (Cont.)
horn (Cont.)
roughness, 497
vibration, 243
compliance, 76, 81
testing, 438
mass, 76, 79
networks, 148
materials, 198,203,204
phonograph, 358
microphone, 273
vibrating systems, 56
conversion, 390
Membrane, circular, 62
recording, 384
rectangular, 63
reproducing, 384
Microphone, barium titapate, 260
terms, 199
calibration, 423
Masking, 572
condenser, 253
Mass, 76,79
crystal, 257
487
dipole, 308
articulation, 494
loudspeakers, 438
electronic, 273
microphones, 423
high order, 311
noise, 488
noise discrimination, 312
records, 472
highly directional, gradient, 315
porosity, 484
large surface, 321
712 INDEX
simple, 323
ultradirectional, 327
ultradirectional, 327
combination pressure and velocity,
inductor, 263
291
lapel, 329
distance effect, 293
lavalier, 329
phase, effect of, 295
lens, 321
sensitivity, effect of, 293
lip, 310
electrostatic, 307
magnetic, 273
high order, 315
polydirectional, 298
307
pressure, 246
uniphase, 305
comparison of electrodynamic,
wind screening, 332
268
Minimum perceptible differences, 574
condenser, 253
Modes of vibration, 56
diaphragm, 262
torsional, 68
electronic, 273
transverse, 57
electrostatic, 253
membrane, circular, 61
gradient, 275
pipes, 69
inductor, 263
plate, circular, 63
magnetic, 270
strings, 50
conductor, 263
Moment of inertia, 76, 80
piezoelectric, 257
Monaural reproducing system, 629,
probe, 267
ribbon, 264
recording, 400
radio, 331
uniaxial, 316
transport, 403
shapes, 339
optical, film, 394
testing, 423
sound track, 394
INDEX 713
Motion picture sound system (Cont.)
One degree of freedom, acoustical, 122
optical (Cont.)
electrical, 122
recording (Cont.)
mechanical, 122
density, 397
Optical-film reproducers, 393
reproducing, 399
Orchestra, live, 606
transport, 400
reproduced, 606
organ, 612
Orifice, circular, impedance, 99
piano, 612
Orthography, 660
synthesizer, 613
Oscillating sphere, impedance, 94
voder, 565
scales, 611
Paging systems, 532
natural, 611
istics, 48
terms, 611
infini te, 102
reflector, 320
plates, 63
spherical wave, 13
Nodes, 56
stationary wave, 14
Noise, 335
Period,3
Barkhausen, 337
Periodic quantity, 3
film, 591
Phase, effect of in, loudspeakers, 138,
microphone, 335
464
record, 377
microphones, 295, 436
reduction, 536
inverter loudspeaker, 159
thermal, 335
Phonetic typewriter, 619, 664
electrons, 336
distortion, 373, 472
molecules, 335
loudspeaker locations, 173
ity, 595
ceramic, 362
ear, 573
compliance of, 373
loudspeakers, 183
crystal, 359
microphones, 333
dynamic, 367
transducers, 594
feedback, 372
Octave, 4
magnetic, 364
Ohm, acoustical, 77
variable resistance, 371
electrical, 77
record, diameter, 379
mechanical, 77
grooves, 376
rotational, 77
speed, 378
714 INDEX
Phonograph (Cont.)
Power (Cont.)
characteristics, 355
tems, 524
grooves, 376
Pressure, gradient microphone, 275,
lateral, 351
microphone, 246
vertical, 355
sound, effective, 6, 74
distortion, 373
maximum, 6, 74
noise, 377
peak,6,74
Photography, 660
sphere, impedance, 93
pickup, 359
rotational, 87
Pipes, closed, 69
99
open, 69
Radio, 650, 662
45
534
Pistonphone, 424
sound reproducing system, 650, 662
Pitch, 571
Range, frequency, ear, 569, 577
Plane waves, 10
musical instruments, 579
amplitude, 10 ",
fundamental, 610
particle velocity, 10
total, 579, 587
pressure, 10
preference, 600, 601, 603
free, 66
reproduced sound, 600, 603
supported center, 66
reproducing systems, 598
supported outside, 66
voice, 579, 587
Point source, 30
fundamental, 610
Porosity, 504
Reactance, acoustical, 76
measurement, 484
electrical, 75
speech, 587
Records, phonograph, 377, 379
INDEX 715
Reflector, 320
Sound, 2
Refraction, 17
absorber, conventional, 503
Reluctance, 199
electronic, 415
electrical, 75, 78
duration, 576
mechanical, 76, 78
energy density; 15
rotational, 76, 78
generator, 658
bars, longitudinal, 66
intensity, 15
torsional, 68
loudness, 570
transverse, 57
motion picture, system, 393, 643,
membrane, circular, 61
663
pitch, 571
pipes, 69
plate, circular, 63
pressure, 6, 74
string, 56
instantaneous, 6, 74
studio, 552
film, 393, 643
theater, 523
magnetic, 384, 390, 400, 634
Rotational, analogies, 71
radio, 650, 662
compliance, 76, 82
stereophonic, 633, 636, 642, 648
elements, 77
telephone, 633, 661
impedance, 76
television, 653, 663
ohm, 77
stages, 553
reactance, 76
timbre, 576
resistance, 76, 78
waves, cylindrical, 28
plane, 10
Semaphore, 660
Source, circular, ring, 43
Similarity principle, 26
nonuniform, 44
Sirens, 409
cone, 53
Slit, narrow, 89
curved, line, 40
definition, 687
doublet, 32
passive, 687
conical, 48
scanning, 689
exponential, 47
716 INDEX
parabolic, 48
magnetic, 201
point, 30
decibels, power and voltage ratios,
series, 35
14
tapered,37
dimensions of quantities, 86, 87
uniform, 36
musical scales, 611
super, 39
noise, in various locations, 526
mechanism, 560
Poisson's ratios, 7
power, 579
quantities; electrical, mechanical,
visible, 568
86,87
vocoder, 566
magnetic, 201
Sphere, oscillating, 94
resistivity, 128
impedance, 94
symbols; electrical, mechanical, ro
pulsating, 93
tational, and acoustical, 86, 87
impedance, 94
magnetic, 201
phase angle, 13
units, 86, 87
pressure, 12
velocities of sound, 7
Stationary wave, 14
Young's moduli, 7
Stethoscopes, 410
Tape, magnetic, 384, 401
Strings, 56
Telegraph, 660
Strip, vibrating, 96
Telephone, 633, 661
Studios, 542
receiver, bipolar, 340
broadcasting, 542
crystal, 346
vocal, 551
inductor, 348
Subharmonic, 4
distortion, 186
testing, 470
Teletype, 660
Suspensions, 192
electrical, 86
stages, 553
mechanical, 86
system, 653, 663
rotational, 87
Testing of, hearing aids, 495
460
acoustical resistances, 7
istic, 464
electrical, 85, 86
452
INDEX 717
Testing of, loudspeakers (Cont.)
Transient response, loudspeakers, 178
438
one tube to another, 117
apparatus, 439
three tubes, 119
465
transmission in, 116, 121
microphones, 423
Tuning fork, 59
calibration, 423
435
Ultrasonics, 692
423
chemical effects, 699
noise, 488
coagulation, 698
noise, 475
delay lines, 703
pickups, 473
detergent, 701
records, 472
dispersion, 697
optical, 472
drilling, 701
471
medical applications, 700
mastoid, 472
soldering, 702
outdoor, 528
quartz, 684
electrons, 336
magnetostriction, 680
molecules, 335
projectors, barium titanate, 686
Thermophone, 425
crystal, 682
Timbre, 576
dynamic, 672, 674
Torque, 73
magnetic, 675
Transducer, 4
ranging, 688
718 INDEX
signaling, 691
Viscosity, 89
sounding, 687
Visible speech, 568
291
Voice, 658
electrostatic, 307
coils, loudspeaker, 192,242
uniaxial, 303
limiter, 403
unidyne, 305
Vowel sounds, 563
electrical, 86
microphone, 319
mechanical, 86
plane, 10
rotational, 87
amplitude, 10
particle velocity, 10
particle, 10, 13
spherical, 11
potential, 6
particle velocity, 13
of sound, 7
phase angle, 13
table, 7
pressure, 12
stationary, 14
45
strip, impedance, 96
Wind, response of microphones, 437
pipes, 69
Wows, 377
pickup, 380
measurement, 475
torsional, bars, 28
Young's modulus, table, 7
transverse, bars, 57
strings, 56
Zone plate, 24
40_000006291