EC6602 Uw 2
EC6602 Uw 2
EC6602 Uw 2
com
Prepared By:
M.MAHARAJA, AP / ECE
UNIT I
FUNDAMENTALS OF
RADIATION
There are several basic properties that are common to all antennas:
Reciprocity: an antennas electrical characteristics are the same whether it is used for
transmitting or receiving. Because this is always true, throughout this lecture, we will
consider antennas as transmitting antennas.
The diagram above shows vertical and horizontal polarization. If the radio wave's
electric field vector points in some other direction, it is said to be obliquely polarized.
If the electric field rotates in space, such that its tip follows an elliptical path, it is
elliptically polarized.
Wavelength: this is the length of one RF wave. It can be computed by either of the
following formulas, depending on the units required:
Antenna gain should be expressed in dBi, wavelength and distances in m and powers
in dBm or dBW.
Here is an example:
Two dipole antennas 100 km apart are aligned and one transmits a 1 kW signal. The
frequency is 222 MHz. What is the received power?
Solution A using dB
Near field (induction field): electromagnetic field created by an antenna that is only
significant at distances of less than 2D/ from the antenna, where D is the longest
dimension of the antenna.
Near field region: A spherical region of radius 2D/ centered on the antenna.
Far field (radiation field): electromagnetic field created by the antenna that extends
throughout all space. At distances greater than 2D/ from the antenna, it is the only
field. It is the field used for communications.
Far field region: The region outside the near field region, at distances greater than
2D/.
Input Impedance: This is the impedance measured at the antenna input terminals. In
general it is complex and has two real parts and one imaginary part:
Radiation resistance: - represents conversion of power into RF waves (real)
Loss resistance represents conductor losses, ground losses, etc. (real)
reactance represents power stored in the near field (imaginary)
Efficiency: this is the ratio of radiation resistance to total antenna input resistance:
The loss resistances come from conductor losses and losses in the ground (the near
field of the antenna can interact with the ground and other objects near the antenna).
The efficiency of practical antennas varies from less than 1% for certain types of low
frequency antennas to 99% for some types of wire antennas.
Electrical length. This came up in the section on transmission lines. It is the length or
distance expressed in terms of wavelengths.
Bandwidth: generally the range of frequencies over which the antenna systems SWR
remains below a maximum value, typically 2.0
Azimuth and Elevation: These are angles used to describe a specific position in an
antenna's radiation pattern. Azimuth is a horizontal angle, generally measured from
true north. The elevation angle is a vertical angle, ranging from 0 degrees (horizon) to
90 degrees (zenith).
A dipole can be any length, but it most commonly is just under 1/2 wavelength long.
A dipole with this length, known as a resonant or half wave dipole, has an input
impedance that is purely resistive and lies between 30 and 80 ohms, which provides a
good match to commercially available 50 ohms coaxial cables as well as commercial
transmitters and receivers, most of which have 50 ohm output and input impedances.
The length of a dipole can be approximately determined from the following formula:
l = 468/f
where:
l is the length in feet and
f is the frequency in MHz.
The 3-dimensional radiation pattern in free space is a fat doughnut with the dipole
piercing its central hole. Notice that unlike an isotropic radiator that radiates equally
well in all directions, the dipole radiates more RF in some directions than others. This
means that the dipole has a gain or directivity over an isotropic radiator of
approximately 2.1 dB. That means that the radiation from the dipole is 2.1 dB stronger
in the direction of maximum radiation than the radiation from an isotropic radiator in
the same direction, when both antennas are fed with the same amount of RF power..
The input impedance of a dipole antenna also depends on its electrical length. When
the antenna is approximately an odd multiple of a half wavelength long, the input
impedance is resistive and lies between 50 and 200 ohms. For antennas that are an
even number of half wavelengths long, the input impedance is resistive and extremely
high, between 1000 and 50,000 ohms.
The chart below shows the effect of ground on the input impedance of a dipole.
As a horizontal antenna is brought closer to the surface of the earth, its input
resistance decreases at first because the electric field is being shorted by the ground.
As the antenna is brought closer, the input resistance will rise again because increases
in ground loss resistance overwhelm the decrease due to shorting of the electric field.
Over a good conductor such as sea water, the input resistance drops steadily as the
antenna is lowered, reaching a value of zero when the antenna touches the water's
surface.
As a horizontal dipole is raised above the ground, the input resistance increases until a
maximum value of approximately 90 ohms is reached at a height of 3/8 . As the
antenna is raised even higher, the input resistance slowly oscillates around the free
space value of 73 ohms. Most dipoles in actual installations show an input resistance
of 50 to 75 ohms, depending on the location.
There is a variation of the /2 dipole known as the folded dipole that is often used for
FM and TV reception. A diagram of the folded dipole is shown below.
The folded dipole is the same overall length as the /2 dipole, but has a second
conductor connected to the first only at the ends, and separated from it by
approximately /400. The input impedance of the folded dipole is approximately 300
ohms, which is a perfect match to TV twin lead and to the input of the TV set. The
folded dipole also has a larger bandwidth than the regular dipole, which is important
for proper TV reception.
The most common vertical antenna is the Marconi antenna. It is a vertical conductor
/4 high, fed at the end near ground. It is essentially a vertical dipole, in which one
side of the dipole is the RF image of the antenna in the ground. This may sound
strange, but remember that ground reflects RF as a mirror reflects light
This type of antenna, unlike the dipole, is an unbalanced antenna, and should be fed
directly with coaxial cable. The shield of the coax is connected to the ground at the
base of the antenna and the center lead of the coax is connected to the vertical radiator.
Because the ground under a vertical antenna is actually part of the antenna, it is
necessary that ground losses be minimized. To minimize the losses, the electrical
conductivity of the ground must be made as high as possible, or an artificial low loss
ground must be provided.
Ground conductivity can be improved by using ground radial wires. These are wires
buried just under the earths surface or laid on the surface that provide a low resistance
path for RF currents flowing in the ground. The ground currents are greatest in the
vicinity of the feed point of a Marconi antenna, so the radials run out from the feed
point, up to a distance of /4 from the antenna, if possible. The ground radials do not
have to be any specific length and the general rule is that a large number of short
radials is preferable to a few long radials. The diagram below shows how current
flows through the ground to the feed point of the Marconi antenna.
The radials should be laid out in a pattern that follows the ground current, that is
running radially out from the feed point of the antenna. The diagram below is a bird's
eye view of typical ground radial layouts. Note that the radials do not all have to be
the same length and that losses may be decreased by adding extra radials near the feed
point. These extra radials can be as short as /40 and still be effective.
Ground losses affect the feed point impedance and antenna efficiency. A Marconi
antenna mounted on a perfectly conducting ground would have an input impedance
that is the impedance of a dipole, or approximately 36 ohms. When mounted on a
real ground, the input impedance can range from 38 ohms for a well designed AM
broadcast antenna mounted over a specially prepared ground, to over 100 ohms for a
Marconi mounted above poor, unprepared ground that has no radials.
Ground loss reduces the antenna's efficiency, because part of the power being
delivered to the antenna is being dissipated in the ground rather than being radiated.
The efficiency can be computed from the measured value of input resistance by using
the following formula:
The radiation pattern of the Marconi antenna is a half doughnut as shown in the figure
below. There is no radiation straight up in the direction of the wire. The bulk of the
radiation occurs at a low elevation angle, which is what is needed to launch a ground
wave.
Yagi-Uda Array
In the previous examples of array design, all of the elements in the array were assumed to
be driven with some source. A Yagi-Uda array is an example of a parasitic array. Any element in
an array which is not connected to the source (in the case of a transmitting antenna) or the
receiver (in the case of a receiving antenna) is defined as a parasitic element. A parasitic array is
any array which employs parasitic elements. The general form of the N-element Yagi-Uda array is
shown below.
! Lightweight
! Low cost
! Simple construction
! Unidirectional beam (front-to-back ratio)
! Increased directivity over other simple wire antennas
! Practical for use at HF (3-30 MHz), VHF (30-300 MHz), and
UHF (300 MHz - 3 GHz)
R = sD = 0.18
sR = sD = 0.28
sR = sD = 0.38
sR = sD = 0.18
sR = sD = 0.28
sR = sD = 0.38
UNIT II
APERTURE AND SLOT ANTENNAS
Radiation from Apertures
Using image theory, the perfect electric (magnetic) conducting screen can be elimi-
nated and replaced by an image magnetic (electric) surface current, doubling its value E J cos sin E J
= (18.2.3)
over the aperture. The image eld causes the total tangential electric (magnetic) eld to H J m m sin cos H Jm m
vanish over the screen. Under the duality transformations (18.2.2), the rst two of Eqs. (18.2.1) transform
If the tangential elds E a , H a were known over the entire aperture plane (screen plus into the last two, and conversely, the last two transform into the rst two.
aperture), the three versions of the equivalence principle would generate the same radi- A useful consequence of duality is that if one has obtained expressions for the elec-
ated elds. But because we consider E a , H a only over the aperture, the three versions tric eld E, then by applying a duality transformation one can generate expressions for
give slightly different results. the magnetic eld H. We will see examples of this property shortly.
In the case of a perfectly conducting screen, the calculated radiation elds (18.4.10) The solution of Eq. (18.2.1) is obtained in terms of the usual scalar and vector po-
using the equivalent currents (18.1.2) are consistent with the boundary conditions on tentials , A, as well as two new potentials m , A m of the magnetic type:
the screen. 1
We provide a justication of the eld equivalence principle (18.1.1) in Sec. 18.10 using jA
E = Am
vector diffraqction theory and the Stratton-Chu and Kottler formulas. The modied (18.2.4)
1
forms (18.1.2) and (18.1.3) are justied in Sec. 18.17 where we derive them in two ways: m jA m +
H = A
one, using the plane-wave-spectrum representation, and two, using the Franz formulas
in conjuction with the extinction theorem discussed in Sec. 18.11, and discuss also their The expression for H can be derived from that of E by a duality transformation of
relationship to Rayleigh-Sommerfeld diffraction theory of Sec. 18.16. the form (18.2.2). The scalar and vector potentials satisfy the Lorenz conditions and
Helmholtz wave equations:
A + j = 0 A m + j m = 0
18.2 Magnetic Currents and Duality
m
2 + k 2 = and 2 m + k2 m = (18.2.5)
Next, we consider the solution of Maxwells equations driven by the ordinary electric
charge and current densities , J, and in addition, by the magnetic charge and current
2 A + k2 A = J 2 A m + k2 A m = J m
densities m , J m .
ejk|rr |
Although m , J m are ctitious, the solution of this problem will allow us to identify
the equivalent magnetic currents to be used in aperture problems, and thus, establish The solutions of the Helmholtz equations are given in terms of G(r r )= :
4|r r |
the eld equivalence principle. The generalized form of Maxwells equations is:
1 1
(r) = (r )G(r r ) dV , m (r) = m (r )G(r r ) dV
H = J + jE V V
(18.2.6)
1 A( r ) = J(r )G(r r ) dV , A m ( r) = J m (r )G(r r ) dV
E=
V V
(18.2.1)
E = J m jH where V is the volume over which the charge and current densities are nonzero. The
observation point r is taken to be outside this volume. Using the Lorenz conditions, the
1 scalar potentials may be eliminated in favor of the vector potentials, resulting in the
H= m
alternative expressions for Eq. (18.2.4):
There is now complete symmetry, or duality, between the electric and the magnetic 1 1
quantities. In fact, it can be veried easily that the following duality transformation E= A)+k2 A A m
(
j
leaves the set of four equations invariant : (18.2.7)
1 1
H= A m )+k2 A m + A
(
E H J Jm A Am j
H E m m
(duality) (18.2.2) These may also be written in the form of Eq. (15.3.9):
J m J A m A
m m 1 1
E= A) J] A m
(
j
where , A and m , A m are the corresponding scalar and vector potentials introduced (18.2.8)
below. These transformations can be recognized as a special case (for = /2) of the 1 1
H= A m ) J m ]+ A
(
following duality rotations, which also leave Maxwells equations invariant: j
Replacing A, A m in terms of Eq. (18.2.6), we may express the solutions (18.2.7) di- Noting that r (F r)=
F +
F and r F =
F
F , and similarly for Fm ,
rectly in terms of the current densities: we nd for the polar components of Eq. (18.3.4):
1 ejkr
E= k2 J G + (J )
G j J m G dV E = jk (F + Fm )+
(F Fm )
j V 4r
(18.2.9) (18.3.6)
1 jk ejkr
H= k2 J m G + (J m )
G + j J G dV H= (F Fm )+
(F + Fm )
j V 4r
Alternatively, if we also use the charge densities, we obtain from (18.2.4): The Poynting vector is given by the generalization of Eq. (16.1.1):
E= j J G + G J m G dV 1 k2
V P= Re(E H )= r |F + Fm |2 + |F Fm |2 = r Pr (18.3.7)
2 322 r 2
(18.2.10)
m
H= j J m G + G + J G dV and the radiation intensity:
V
dP k2
U(, )= = r 2 Pr = |F + Fm |2 + |F Fm |2 (18.3.8)
18.3 Radiation Fields from Magnetic Currents d 322
The radiation elds of the solutions (18.2.7) can be obtained by making the far-eld
18.4 Radiation Fields from Apertures
approximation, which consists of the replacements:
For an aperture antenna with effective surface currents given by Eq. (18.1.1), the volume
ejk|rr |
ejkr jkr
e and jk (18.3.1) integrations in Eq. (18.2.9) reduce to surface integrations over the aperture A:
4|r r | 4r
1
where k = kr. Then, the vector potentials of Eq. (18.2.6) take the simplied form: E= (J s )
G + k2 J s G j J ms G dS
j A
(18.4.1)
ejkr ejkr 1
A(r)= F(, ) , A m (r)= Fm (, ) (18.3.2) H= (J ms )
G + k2 J ms G + j J s G dS
4r 4r j A
where the radiation vectors are the Fourier transforms of the current densities: and, explicitly in terms of the aperture elds shown in Fig. 18.1.1:
F(, ) =
J(r )ejkr dV 1
V E= (
(n H a ) G)+k2 (n H a )G + j(n E a )
G dS
(radiation vectors) (18.3.3) j A
Fm (, ) = J m (r )ejkr dV 1
V H= (
(n E a ) G)k2 (n E a )G + j(n H a )
G dS
j A
Setting J = J m = 0 in Eq. (18.2.8) because we are evaluating the elds far from the (18.4.2)
current sources, and using the approximation = jk = jkr, and the relationship These are known as Kottlers formulas [12911296,1286,12971301]. We derive them
k/ = , we nd the radiated E and H elds: in Sec. 18.12. The equation for H can also be obtained from that of E by the application
of a duality transformation, that is, E a H a , H a E a and , .
ejkr In the far-eld limit, the radiation elds are still given by Eq. (18.3.6), but now the
E = j r (A r) r A m = jk r F r Fm radiation vectors are given by the two-dimensional Fourier transform-like integrals over
4r
(18.3.4) the aperture:
j jk ejkr
H= r (A m r)+r A = r F + Fm r
4r
F(, ) = J s (r )ejkr dS = n H a (r )ejkr dS
A A
These generalize Eq. (15.10.2) to magnetic currents. As in Eq. (15.10.3), we have: (18.4.3)
jkr jkr
Fm (, ) = J ms (r )e dS = n E a (r )e dS
1 A A
H= r E (18.3.5)
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
18.4. Radiation Fields from Apertures 805 806 18. Radiation from Apertures
F = z g = z (x gx + y gy )= y gx x gy
(18.4.7)
Fm = z f = z (x fx + y fy )= x fy y fx
The polar components of the radiation vectors are determined as follows:
F =
F =
(y gx x gy )= gx sin cos gy cos cos
x) and (
where we read off the dot products ( y) from Eq. (15.8.3). The remaining
polar components are found similarly, and we summarize them below:
F = gx cos + gy sin
(18.4.8)
Fm = cos (fy cos fx sin )
Fig. 18.4.1 Radiation elds from an aperture. Fm = (fx cos + fy sin )
It follows from Eq. (18.3.6) that the radiated E-eld will be:
Fig. 18.4.1 shows the polar angle conventions, where we took the origin to be some-
where in the middle of the aperture A. ejkr
E = jk (fx cos + fy sin )+ cos (gy cos gx sin )
The aperture surface A and the screen in Fig. 18.1.1 can be arbitrarily curved. How- 4r
(18.4.9)
ever, a common case is to assume that they are both at. Then, Eqs. (18.4.3) become ejkr
ordinary 2-d Fourier transform integrals. Taking the aperture plane to be the xy-plane E = jk cos (fy cos fx sin )(gx cos + gy sin )
4r
as in Fig. 18.1.1, the aperture normal becomes n = z, and thus, it can be taken out of
the integrands. Setting dS = dx dy , we rewrite Eq. (18.4.3) in the form: The radiation elds resulting from the alternative forms of the eld equivalence
principle, Eqs. (18.1.2) and (18.1.3), are obtained from Eq. (18.4.9) by removing the g- or
F(, ) = J s (r )ejkr dx dy = z H a (r )ejkr dx dy the f -terms and doubling the remaining term. We have for the PEC case:
A A
(18.4.4)
ejkr
Fm (, ) =
J ms (r )ejkr dx dy = z
E a (r )ejkr dx dy E = 2jk fx cos + fy sin
A A
4r
(18.4.10)
jkr jkx x +jky y
ejkr
where e =e and kx = k cos sin , ky = k sin sin . It proves conve- E = 2jk cos (fy cos fx sin )
4r
nient then to introduce the two-dimensional Fourier transforms of the aperture elds:
and for the PMC case:
jkr jkx x +jky y
f(, )= E a (r )e dx dy = E a (x , y )e dx dy ejkr
A A E = 2jk cos (gy cos gx sin )
(18.4.5) 4r
(18.4.11)
g(, )= H a (r )ejkr dx dy = H a (x , y )ejkx x +jky y dx dy ejkr
A A E = 2jk (gx cos + gy sin )
4r
Then, the radiation vectors become:
In all three cases, the radiated magnetic elds are obtained from:
F(, ) = z g(, )
(18.4.6) 1 1
H = E , H = E (18.4.12)
Fm (, ) = z f(, )
Because E a , H a are tangential to the aperture plane, they can be resolved into their We note that Eq. (18.4.9) is the average of Eqs. (18.4.10) and (18.4.11). Also, Eq. (18.4.11)
cartesian components, for example, E a = x Eax + y Eay . Then, the quantities f, g can be is the dual of Eq. (18.4.10). Indeed, using Eq. (18.4.12), we obtain the following H-
resolved in the same way, for example, f = x fx + y fy . Thus, we have: components for Eq. (18.4.11), which can be derived from Eq. (18.4.10) by the duality
transformation E a H a or f g , that is,
The Huygens source condition implies the same relationship for the Fourier trans-
ejkr forms of the aperture elds, that is, (with n = z)
H = 2jk gx cos + gy sin
4r
(18.4.13)
1 1 1
ejkr g= n f gx = fy , gy = fx (18.5.2)
H = 2jk cos (gy cos gx sin )
4r
At = 90o , the components E , H become tangential to the aperture screen. We Inserting these into Eq. (18.4.9) we may express the radiated electric eld in terms
note that because of the cos factors, E (resp. H ) will vanish in the PEC (resp. PMC) of f only. We nd:
case, in accordance with the boundary conditions.
ejkr 1 + cos
E = jk fx cos + fy sin
2r 2
18.5 Huygens Source (18.5.3)
ejkr 1 + cos
E = jk fy cos fx sin
The aperture elds E a , H a are referred to as Huygens source if at all points on the 2r 2
aperture they are related by the uniform plane-wave relationship:
The factor (1 + cos )/2 is known as an obliquity factor. The PEC case of Eq. (18.4.10)
1 remains unchanged for a Huygens source, but the PMC case becomes:
Ha = n E a (Huygens source) (18.5.1)
ejkr
E = jk cos fx cos + fy sin
where is the characteristic impedance of vacuum. 2r
(18.5.4)
For example, this is the case if a uniform plane wave is incident normally on the ejkr
E = jk fy cos fx sin
aperture plane from the left, as shown in Fig. 18.5.1. The aperture elds are assumed to 2r
be equal to the incident elds, E a = Einc and H a = Hinc , and the incident elds satisfy
We may summarize all three cases by the single formula:
Hinc = z Einc /.
ejkr
E = jk c fx cos + fy sin
2r
(elds from Huygens source) (18.5.5)
ejkr
E = jk c fy cos fx sin
2r
We note that the rst is the average of the last two. The obliquity factors are equal to
unity in the forward direction = 0o and vary little for near-forward angles. Therefore,
the radiation patterns predicted by the three methods are very similar in their mainlobe
behavior.
Fig. 18.5.1 Uniform plane wave incident on an aperture.
In the case of a modied Huygens source that replaces by T , Eqs. (18.5.5) retain
their form. The aperture elds and their Fourier transforms are now assumed to be
The Huygens source condition is not always satised. For example, if the uniform
related by:
plane wave is incident obliquely on the aperture, then must be replaced by the trans-
verse impedance T , which depends on the angle of incidence and the polarization of 1 1
Ha = z E a g= z f (18.5.7)
the incident wave as discussed in Sec. 7.2. T T
Similarly, if the aperture is the open end of a waveguide, then must be replaced by Inserting these into Eq. (18.4.9), we obtain the modied obliquity factors :
the waveguides transverse impedance, such as TE or TM , depending on the assumed
waveguide mode. On the other hand, if the waveguide ends are ared out into a horn 1 1
c = [1 + K cos ] , c = [K + cos ] , K= (18.5.8)
with a large aperture, then Eq. (18.5.1) is approximately valid. 2 2 T
18.6 Directivity and Effective Area of Apertures conservation and compute the total power that ows into the right half-space through
the aperture. Assuming a Huygens source, we have:
For any aperture, given the radiation elds E , E of Eqs. (18.4.9)(18.4.11), the corre-
1 1
sponding radiation intensity is: Prad = Pz dS = z Re E a Ha dS = |E a (r )|2 dS (18.6.8)
A 2 A 2 A
dP 1 1 Because = 0 corresponds to kx = ky = 0, it follows from the Fourier transform
U(, )= = r 2 Pr = r 2 |E |2 + |E |2 = r 2 |E(, )|2 (18.6.1)
d 2 2 denition (18.4.5) that:
2
2
Because the aperture radiates only into the right half-space 0 /2, the total
|f|2max =
E a ( r
)ejkr
dS
E a ( r
) dS
radiated power and the effective isotropic radiation intensity will be: A kx =ky =0 A
4 1 1
E a (r ) dS
0 0
Umax = |f |2max =
(18.6.9)
22 22 A
The directive gain is computed by D(, )= U(, )/UI , and the normalized gain
Dividing (18.6.9) by (18.6.8), we nd the directivity:
by g(, )= U(, )/Umax . For a typical aperture, the maximum intensity Umax is
towards the forward direction = 0o . In the case of a Huygens source, we have:
2
E a (r ) dS
Umax 4
A
4Aeff
k2 2 Dmax = 4 = 2 = (directivity) (18.6.10)
U(, )= c |fx cos + fy sin |2 + c2 |fy cos fx sin |2 (18.6.3) Prad 2 2
82 |E a (r )| dS
A
Assuming that the maximum is towards = 0o , then c = c = 1, and we nd for
It follows that the maximum effective area of the aperture is:
the maximum intensity:
2
k2
E a (r ) dS
82 Aeff = A
A (effective area) (18.6.11)
|E a (r )|2 dS
k2 2
k A
= |fx |2 + |fy |2 =0 = |f |2max
82 82 and the aperture efciency:
where |f|2max = |fx |2 + |fy |2 =0 . Setting k = 2/, we have:
2
E a (r ) dS
1 Aeff
eatl =
A
, epel =
A
2 (18.6.13)
|E(, )| 1 + cos |f(, )|
|E |max 2 |f |max A A
so that ea becomes the product:
The power computed by Eq. (18.6.2) is the total power that is radiated outwards from
ea = eatl epel (18.6.14)
a half-sphere of large radius r . An alternative way to compute Prad is to invoke energy
1 1 a
vx = kx a = ka sin cos = sin cos
2 2
(18.8.2)
1 1 b
vy = ky b = kb sin sin = sin sin
2 2
The pattern simplies along the two principal planes, the xz- and yz-planes, corre-
sponding to = 0o and = 90o . We have:
sin(vx ) sin (a/)sin
f (, 0o ) = =
vx (a/)sin
(18.8.3)
sin(vy ) sin (b/)sin
f (, 90o ) = =
vy (b/)sin
Fig. 18.7.1 Uniform rectangular and circular apertures. Fig. 18.8.1 shows the three-dimensional pattern of Eq. (18.7.3) as a function of the
independent variables vx , vy , for aperture dimensions a = 8 and b = 4. The x, y
The eld E a can have an arbitrary direction, with constant x- and y-components, separability of the pattern is evident. The essential MATLAB code for generating this
E a = x E0x + y E0y . Because E a is constant, its Fourier transform f(, ) becomes: gure was (note MATLABs denition of sinc(x)= sin(x)/(x)):
f(, )= E a (r )ejkr dS = E a ejkr dS A f (, ) E a (18.7.1)
A A
field strength
A A
0.5
The quantity f (, ) depends on the assumed geometry of the aperture and it, alone,
determines the radiation pattern. Noting that the quantity |E a | cancels out from the
ratio in the gain (18.6.7) and that f (0, )= (1/A) A dS = 1, we nd for the normalized
0
gain and eld strengths: 8
4 8
|E(, )| 1 + cos 0
0
4
= g(, ) = |f (, )| (18.7.3) 4
|E |max 2 vy
8 8
4
vx
18.8 Rectangular Apertures Fig. 18.8.1 Radiation pattern of rectangular aperture (a = 8, b = 4).
For a rectangular aperture of sides a, b, the area integral (18.7.2) is separable in the x-
a = 8; b = 4;
and y-directions: [theta,phi] = meshgrid(0:1:90, 0:9:360);
a/2 b/2 a/2 b/2 theta = theta*pi/180; phi = phi*pi/180;
1 1 1
f (, )= ejkx x +jky y dx dy = ejkx x dx ejky y dy
ab a/2 b/2 a a/2 b b/2 vx = a*sin(theta).*cos(phi);
vy = b*sin(theta).*sin(phi);
where we placed the origin of the r integration in the middle of the aperture. The above
integrals result in the sinc-function patterns: E = abs((1 + cos(theta))/2 .* sinc(vx) .* sinc(vy));
field strength
field strength
vy vary over the limits a/ vx a/ and b/ vy b/. In fact, the physically 13.26 dB 13.26 dB
realizable values of vx , vy are those that lie in the ellipse in the vx vy -plane: 0.5 0.5
vx2 vy2 1
2
+ 2 2 (visible region) (18.8.4)
a b
0 0
The realizable values of vx , vy are referred to as the visible region. The graph in 0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90
Fig. 18.8.1 restricts the values of vx , vy within that region. (degrees) (degrees)
The radiation pattern consists of a narrow mainlobe directed towards the forward
Fig. 18.8.2 Radiation patterns along the two principal planes (a = 8, b = 4).
direction = 0o and several sidelobes.
We note the three characteristic properties of the sinc-function patterns: (a) the 3-
dB width in v-space is vx = 0.886 (the 3-dB wavenumber is vx = 0.443); (b) the rst Therefore, for the purpose of computing the integral (18.7.2), we may set = 0. We
sidelobe is down by about 13.26 dB from the mainlobe and occurs at vx = 1.4303; and have then k r = kx x = k sin cos . Writing dS = d d , we have:
(c) the rst null occurs at vx = 1. See Sec. 20.7 for the proof of these results.
a 2
The 3-dB width in angle space can be obtained by linearizing
the relationship vx = 1 sin cos
f ()= ejk d d (18.9.1)
(a/)sin about = 0o , that is, vx = (a/) cos
=0 = a/. Thus, = a2 0 0
vx /a. This ignores also the effect of the obliquity factor. It follows that the 3-dB The - and -integrations can be done using the following integral representations
widths in the two principal planes are (in radians and in degrees): for the Bessel functions J0 (x) and J1 (x) [1449]:
2 1
x = 0.886 = 50.76o , y = 0.886 = 50.76o (18.8.5) 1 J1 (x)
a a b b J0 (x)= ejx cos d and J0 (xr)r dr = (18.9.2)
2 0 0 x
The 3-dB angles are x = x /2 = 25.4o /a and y = y /2 = 25.4o /b.
Then Eq. (18.9.1) gives:
Fig. 18.8.2 shows the two principal radiation patterns of Eq. (18.7.3) as functions of
, for the case a = 8, b = 4. The obliquity factor was included, but it makes essen- J1 (ka sin ) J1 (2u) 1 a
tially no difference near the mainlobe and rst sidelobe region, ultimately suppressing f ()= 2 =2 , u= ka sin = sin (18.9.3)
ka sin 2u 2
the response at = 90o by a factor of 0.5.
The 3-dB widths are shown on the graphs. The rst sidelobes occur at the angles This is the well-known Airy pattern [634] for a circular aperture. The function f ()
a = asin(1.4303/a)= 10.30o and b = asin(1.4303/b)= 20.95o . is normalized to unity at = 0o , because J1 (x) behaves like J1 (x) x/2 for small x.
For aperture antennas, the gain is approximately equal to the directivity because the Fig. 18.9.1 shows the three-dimensional eld pattern (18.7.3) as a function of the in-
losses tend to be very small. The gain of the uniform rectangular aperture is, therefore, dependent variables vx = (a/)sin cos and vy = (a/)sin sin , for an aperture
G D = 4(ab)/2 . Multiplying G by Eqs. (18.8.5), we obtain the gain-beamwidth radius of a = 3. The obliquity factor was not included as it makes little difference
product p = G x y = 4(0.886)2 = 9.8646 rad2 = 32 383 deg2 . Thus, we have an near the main lobe. The MATLAB code for this graph was implemented with the built-in
example of the general formula (16.3.14) (with the angles in radians and in degrees): function besselj:
9.8646 32 383 a = 3;
G= = (18.8.6)
x y ox oy [theta,phi] = meshgrid(0:1:90, 0:9:360);
theta = theta*pi/180; phi = phi*pi/180;
1
1
3 dB
field strength
field strength
17.56 dB
0.5
0.5
0
3
3 0
0 10 20 30 40 50 60 70 80 90
0
0 (degrees)
vy vx
3 3
Fig. 18.9.2 Radiation pattern of circular aperture (a = 3).
Fig. 18.9.1 Radiation pattern of circular aperture (a = 3).
The rst-null angle null = 0.6098/a is the so-called Rayleigh diffraction limit for
E(i) = abs(2*besselj(1,2*pi*u(i))./(2*pi*u(i))); the nominal angular resolution of optical instruments, such as microscopes and tele-
scopes. It is usually stated in terms of the diameter D = 2a of the optical aperture:
surfl(vx,vy,E);
shading interp; colormap(gray(16));
= 1.22 = 70o (Rayleigh limit) (18.9.7)
D D
The visible region is the circle on the vx vy -plane:
E(r), if r V
(3)
E(r ) (r r ) dV = uV (r) E(r)= (18.10.4)
V 0, if r
V
We may now solve Eq. (18.10.3) for E(r). In a similar fashion, or, performing a duality
transformation on the expression for E(r), we also obtain the corresponding magnetic
eld H(r). Using (18.10.1), we have:
Fig. 18.10.1 Fields outside a closed surface S.
1 G E
E(r) = G G J m dV +
j G J E G dS
1 V S n n
2 E + k2 E = j J + + J m
1 G H
(18.10.1) H(r) = j G J m G m + G J dV +
H
G
dS
1 V S n n
2 H + k2 H = j J m + m J
(18.10.6)
We recall that the Greens function for the Helmholtz equation is: Because of the presence of the particular surface term, we will refer to these as
the Kirchhoff diffraction formulas. Eqs. (18.10.6) can be transformed into the so-called
ejk|rr |
Stratton-Chu formulas [12911296,1286,12971301]:
2 G + k2 G = (3) (r r ) , G(r r )= (18.10.2)
4|r r |
where is the gradient with respect to r . Applying Greens second identity given by E(r)= j G J + G J m G dV
V
Eq. (C.27) of Appendix C, we obtain:
E G + j G(n H)+(n E) G + (n E)
G dS
2 2
G E E G dV = G E dS , = n S
V S+S n n n (18.10.7)
m
where G and E stand for G(r r ) and E(r ) and the integration is over r . The quantity H(r)= j G J m + G + J G dV
V
/n is the directional derivative along n. The negative sign in the right-hand side
arises from using a unit vector n that is pointing into the volume V. + j G(n E)+(n H) G + (n H)
G dS
The integral over the innite surface is taken to be zero. This may be justied more S
rigorously [1293] by assuming that E and H behave like radiation elds with asymptotic
The proof of the equivalence of (18.10.6) and (18.10.7) is rather involved. Problem
form E const. ejkr /r and H r E/. Thus, dropping the S term, and adding
18.4 breaks down the proof into its essential steps.
and subtracting k2 G E in the left-hand side, we obtain:
Term by term comparison of the volume and surface integrals in (18.10.7) yields the
E G effective surface currents of the eld equivalence principle:
G(2 E + k2 E)E (2 G + k2 G) dV = G E dS (18.10.3)
V S n n
J s = n H , J ms = n E (18.10.8)
Using Eq. (18.10.2), the second term on the left may be integrated to give E(r):
Similarly, the effective surface charge densities are:
E(r ) (2 G + k2 G) dV = E(r ) (3) (r r ) dV = E(r)
V V
s = n E , ms = n H (18.10.9)
where we assumed that r lies in V. This integral is zero if r lies in V1 because then r Technically [1299], one must set uV (r)= 1/2, if r lies on the boundary of V, that is, on S.
can never be equal to r. For arbitrary r, we may write: See [1288,1294,1300,1301] for earlier work by Larmor, Tedone, Ignatowski, and others.
Initially derived by Larmor and Love [1300,1301], and later developed fully by Schelkunoff [1287,1289].
The precise conditions are: r|E| const. and r|E H r| 0 as r .
Eqs. (18.10.7) may be transformed into the Kottler formulas [12911296,1286,1297 This happens if we choose S in Fig. 18.10.1 such that all the current sources are
1301], which eliminate the charge densities , m in favor of the currents J, J m : inside it, or, if in Fig. 18.10.2 we choose S such that all the current sources are outside
it, then, the Kirchhoff, Stratton-Chu, Kottler, and Franz formulas simplify into:
1
E(r)= k2 J G + (J )
G j J m G dV
j G E
dS
V
E(r) = E G
1
S n n
+ G + j(n E)
k2 G(n H)+ (n H) G dS
j S
= j G(n H )+(n E ) G + (n E )
G dS
S
1
H(r)= k2 J m G + (J m )
G + j J G dV 1
j V = G + j(n E )
k2 G(n H )+ (n H ) G dS
j
1
S
+ G + j(n H)
k2 G(n E) (n E) G dS
j 1
S = G(n H )dS + G(n E )dS
(18.10.10) j S S
1
1
E(r) = (A + A s ) J (A m + A ms )
j
(18.10.11)
1
1
H(r) = (A m + A ms ) J m + (A + A s )
j
where A and A m were dened in Eq. (18.2.6). The new potentials are dened by:
Fig. 18.10.3 Current sources are outside the eld region.
A s (r) = J s (r )G(r r )dS = n H(r ) G(r r )dS
S S
(18.10.12)
A ms (r) = J ms (r )G(r r )dS = n E(r ) G(r r )dS
S S
18.11 Extinction Theorem
Next, we specialize the above formulas to the case where the volume V contains
no current sources (J = J m = 0), so that the E, H elds are given only in terms of the In all of the equivalent formulas for E(r), H(r), we assumed that r lies within the volume
surface integral terms. V. The origin of the left-hand sides in these formulas can be traced to Eq. (18.10.4), and
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
18.11. Extinction Theorem 821 822 18. Radiation from Apertures
therefore, if r is not in V but is within the complementary volume V1 , then the left-hand
sides of all the formulas are zero. This does not mean that the elds inside V1 are
1
zeroit only means that the sum of the terms on the right-hand sides are zero. k2 J G + (J )
G j J m G dV
j V1
To clarify these remarks, we consider an imaginary closed surface S dividing all
1
space in two volumes V1 and V, as shown in Fig. 18.11.1. We assume that there are = G + j(n E)
k2 G(n H)+ (n H) G dS
current sources in both regions V and V1 . The surface S1 is the same as S but its unit j S
eld can be calculated by any of the four alternative formulas, Kottler, Franz, Stratton-
Chu, or Kirchhoff integralall applied to the open surface S. Fig. 18.13.1 Fresnel diffraction through rectangular aperture.
The quantity (n r2 )(n r1 ) is an obliquity factor. Next, we set r = r1 + r2 and It follows that the Fresnel approximation of the diffraction coefcient for an arbitrary
dene the free-space eld at the point P2 : aperture will be given by:
r1 + r 2 r
(18.13.7) D= = ejk(b b )/(2F) dS (18.13.11)
E0 2F S
If the origin were the point of intersection between the aperture plane and the line
A further simplication is obtained by assuming that the aperture plane is the xy-
P1 P2 , then E0 would represent the eld received at point P2 in the unobstructed case
plane and that the line P1 P2 lies on the yz plane at an angle with the z-axis, as shown
when the aperture and screen are absent.
in Fig. 18.13.2.
The ratio D = E/E0 may be called the diffraction coefcient and depends on the
aperture and the relative geometry of the points P1 , P2 :
E jk
D= = (n r2 )(n r1 ) ejk(R1 +R2 r1 r2 ) dS (18.13.8)
E0 4F S
1 1 1 r1 r2
= + F= (18.13.9)
F r1 r2 r1 + r2
where we assumed that the aperture limits are (with respect to the new origin):
To simplify this expression, we now assume that the origin is the point of intersection
of the line of sight P1 P2 and the aperture plane. Then, the vectors r1 and r2 are anti-
x1 x x2 , y1 y y2
parallel and so are their unit vectors r1 = r2 . The linear terms cancel and the quadratic
ones combine to give: The end-points y1 , y2 are shown in Fig. 18.13.2. The integrals may be expressed
1 1
in terms of the Fresnel functions C(x), S(x), and F(x)= C(x)jS(x) discussed in
R1 +R2 r1 r2 = r r (r2 r )2 =
r r2 (r r2 )
2 = 1 b b (18.13.10)
Appendix F. There, the complex function F(x) is dened by:
2F 2F 2F
x
where we dened b = r r2 (r r2 ), which is the perpendicular vector from the point F(x)= C(x)jS(x)= ej(/2)u du
2
(18.13.13)
P to the line-of-sight P1 P2 , as shown in Fig. 18.13.1. 0
Note that the quantities b1 = y1 cos and b2 = y2 cos are the perpendicular
distances from the edges to the line P1 P2 . Since du dv = (k cos /F)dx dy , we Fig. 18.14.1 Illuminated and shadow regions in straight-edge diffraction.
obtain for the diffraction coefcient:
u2 v2
j j
eju ejv
2 2
/2 /2
D= du dv = F(u2 )F(u1 ) F(v2 )F(v1 )
2
2 u1 v1 2 |E|2 1
F(v)+ 1 j
|D(v)|2 = =
(18.14.2)
|E0 |2 2 2
Noting that F(x) is an odd function and that j/2 = 1/(1 j)2 , we obtain:
or, in terms of the real and imaginary parts of F(v):
E F(u1 )+F(u2 ) F(v1 )+F(v2 )
D= = (rectangular aperture) (18.13.16) 2
E0 1j 1j 1 1 1 2
2
|D(v)| = C(v)+ + S(v)+ (18.14.3)
2 2 2
The normalization factors (1 j) correspond to the innite aperture limit u1 , u2 , v1 ,
v2 , that is, no aperture at all. Indeed, since the asymptotic value of F(x) is
The quantity |D(v)|2 is plotted versus v in Fig. 18.14.2. At v = 0, corresponding to
F()= (1 j)/2, we have:
the line P1 P2 grazing the top of the edge, we have F(0)= 0, D(0)= 1/2, and |D(0)|2 =
F(u1 )+F(u2 ) F(v1 )+F(v2 ) F()+F() F()+F() 1/4 or a 6 dB loss. The rst maximum in the illuminated region occurs at v = 1.2172
=1
1j 1j 1j 1j and has the value |D(v)|2 = 1.3704, or a gain of 1.37 dB.
In the case of a long slit along the x-direction, we only take the limit u1 , u2 : Diffraction Coefficient Diffraction Coefficient in dB
1.5
E F(v1 )+F(v2 )
D= = (diffraction by long slit) (18.13.17) 1.25
E0 1j 0
20 log10|D()|
1
6
|D()|2
, or v2 , which corresponds to keeping the lower edge of the slit. In this limit 18
0.25
F(v2 ) F()= (1 j)/2. Denoting v1 by v, we have:
0 24
3 2 1 0 1 2 3 4 5 3 2 1 0 1 2 3 4 5
1 1j k
D(v)= F(v)+ , v= b1 (18.14.1)
1j 2 F
Fig. 18.14.2 Diffraction coefcient in absolute and dB units.
Positive values of v correspond to positive values of the clearance distance b1 , plac-
ing the point P2 in the illuminated region, as shown in Fig. 18.14.1. Negative values of The asymptotic behavior of D(v) for v is obtained from Eq. (F.4). We have
v correspond to b1 < 0, placing P2 in the geometrical shadow region behind the edge. for large positive x:
The magnitude-square |D|2 represents the intensity of the diffracted eld relative 1j j jx2 /2
F(x) + e
to the intensity of the unobstructed eld. Since |1 j|2 = 2, we nd: 2 x
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
18.14. Knife-Edge Diffraction 829 830 18. Radiation from Apertures
1 j jv2 /2
1 2v e
, for v +
D(v)= (18.14.4)
1 j jv2 /2
e , for v
2v
We may combine the two expressions into one with the help of the unit-step function
u(v) by writing D(v) in the following form, which denes the asymptotic diffraction
coefcient d(v):
D(v)= u(v)+d(v)ejv
2
/2
(18.14.5)
Fig. 18.14.3 Communicating antennas over an obstacle.
where u(v)= 1 for v 0 and u(v)= 0 for v < 0.
With u(0)= 1, this denition requires d(0)= D(0)v(0)= 0.5 1 = 0.5. But if
we dene u(0)= 0.5, as is sometimes done, then, d(0)= 0. The asymptotic behavior of The distance b1 can also be expressed approximately in terms of the subtended angles 1 ,
D(v) can now be expressed in terms of the asymptotic behavior of d(v): 2 , and , shown in Fig. 18.14.3:
1j b1 l1 1 l2 2 b1 = l1 l2 1 2 (18.14.9)
d(v)= , for v (18.14.6)
2v
and in terms of , we have:
In the illuminated region D(v) tends to unity, whereas in the shadow region it de-
l2 l1 2F
creases to zero with asymptotic dB attenuation or loss: 1 = , 2 = b1 = F v= (18.14.10)
l1 + l2 l1 + l2
2
L = 10 log10
d(v)
= 10 log10 22 v2 , as v (18.14.7)
The case of multiple obstacles has been studied using appropriate modications of the
knife-edge diffraction problem and the geometrical theory of diffraction [13091324].
The MATLAB function diffr calculates the diffraction coefcient (18.14.1) at any
vector of values of v. It has usage: Example 18.14.2: Fresnel Zones. Consider two antennas separated by a distance d and an ob-
stacle at distance z from the midpoint with clearance b, as shown below. Fresnel zones and
D = diffr(v); % knife-edge diffraction coefcient D(v)
the corresponding Fresnel zone ellipsoids help answer the question of what the minimum
value of the clearance b should be for efcient communication between the antennas.
For values v 0.7, the diffraction loss can be approximated very well by the follow-
ing function [1308]: Diffraction Coefficient in dB
3
exact
2 asymptotic
L = 10 log10
D(v)
= 6.9 + 20 log10 (v + 0.1)2 +1 v 0.1 (18.14.8) 2 extrema
fresnel zone
20 log10|D()|
1
Example 18.14.1: Diffraction Loss over Obstacles. The propagation path loss over obstacles and
irregular terrain is usually determined using knife-edge diffraction. Fig. 18.14.3 illustrates 0
the case of two antennas communicating over an obstacle. For small angles , the focal
1
length F is often approximated in several forms:
2
r1 r2 d1 d2 l1 l2
F=
r1 + r 2 d1 + d2 l1 + l2 3
0 1 2 3 4 5
These approximations are valid typically when d1 , d2 are much greater than and the
height h of the obstacle, typically, at least ten times greater. The clearance distance can
be expressed in terms of the heights: The diffraction coefcient D(v) and its asymptotic form were given in Eqs. (18.14.1) and
(18.14.4), that is,
h1 d2 + h2 d1
b1 = y1 cos = h cos
d1 + d 2 1 1j k 2 d1 d2
D(v)= F(v)+ , v= b= b, F= (18.14.11)
1j 2 F F d1 + d2
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
18.14. Knife-Edge Diffraction 831 832 18. Radiation from Apertures
and for positive and large clearance b, or equivalently, for large positive v, For example, for a distance of d = 1 km, using a cell phone frequency of f = 1 GHz,
corresponding to wavelength = 30 cm, we nd b = d/2 = 8.66 meters.
1 j jv2 /2 1 2
Das (v)= 1 e =1 ej(v /2+1/4) (18.14.12) A common interpetation and derivation of Fresnel zones is to consider the path difference
2v 2v
between the rays following the straight path connecting the two antennas and the path
As can be seen in the above gure on the right, the diffraction coefcients D(v) and getting scattered from the obstacle, that is, l = l1 + l2 d. From the indicated triangles,
Das (v) agree closely even for small values of v. Therefore, the extrema can be obtained and assuming that b d1 and b d2 , we nd:
from the asymptotic form. They correspond to the values of v that cause the exponential
b2 b2
in (18.14.12) to take on its extremal values of 1, that is, the vs that satisfy v2 /2 + 1/4 = n, l1 = d21 + b2 d1 + , l2 = d22 + b2 d2 +
2d1 2d2
with integer n, or:
vn = 2n 0.5 , n = 1, 2, . . . (18.14.13) which leads to the following path length l, expressed in terms of v:
The corresponding values of D(v), shown on the gure with black dots, are given by b2 1 1 b2
l = l1 + l2 d = + = = v2
2 d1 d2 2F 4
1 1
Das (vn )= 1 ejn = 1 (1)n (18.14.14)
2vn 2vn
The corresponding phase difference between the two paths, ejkl , will be then:
An alternative set of vs, also corresponding to alternating almost extremum values, are ejkl = ejv
2 /2
(18.14.19)
those that dene the conventional Fresnel zones, that is,
which has the same form as in the diffraction coefcient Das (v). The values v = un =
un = 2n , n = 1, 2, . . . (18.14.15)
2n will make the path difference a multiple of /2, that is, l = n/2, resulting in the
alternating phase ejkl = (1)n .
These are indicated by open circles on the graph. The corresponding D(v) values are:
The discrepancy between the choices vn and un arises from using D(v) to nd the alter-
ej/4 nating maxima, versus using the plain phase (18.14.19).
Das (un )= 1 (1)n (18.14.16)
2un
The Fresnel approximation is not invariant under shifting the origin. Our choice of
For clearances b that correspond to vs that are too small, i.e., v < 0.5, the diffraction origin above is not convenient because it depends on the observation point P2 . If we
coefcient D(v) becomes too small, impeding efcient communication. The smallest ac-
choose a xed origin, such as the point O in Fig. 18.14.4, then, we must determine the
ceptable clearance b is taken to correspond to the rst maximum of D(v), that is, v = v1
corresponding Fresnel coefcient.
or more simply v = u1 = 2.
The locus of points (b, z) corresponding to a xed value of v, and hence to a xed value
of the diffraction coefcient D(v), form an ellipsoid. This can be derived from (18.14.11)
by setting d1 = d/2 + z and d2 = d/2 z, that is,
2 F (d2 /4 z2 ) 2 d1 d2 d2 /4 z2
v= b b2 = v2 = v , because F = =
F 2 2d d1 + d 2 d
The diffraction coefcient is still given by Eq. (18.13.8) but with r1 , r2 replaced by where we set ky02 /2F = v2 /2 and dened the incident eld Eedge at the edge and the
the distances l1 , l2 . The unit vectors towards P1 and P2 are: overall edge-diffraction coefcient Dedge by:
l1 = z cos 1 y sin 1 , l2 = z cos 2 y sin 2 (18.14.20) ejkl1 FF cos 1 + cos 2 2
Eedge = A1 , Dedge = ejv /2
D(v) (18.14.23)
l1 l2 2
Since r = x x + y y and n = z, we nd:
The second factor (ejkl2 / l2 ) in (18.14.22) may be interpreted as a cylindrical wave
l1 r = y sin 1 , l2 r = y sin 2 , n l1 = cos 1 , n l2 = cos 2 emanating from the edge as a result of the incident eld Eedge . The third factor Dedge is
The quadratic approximation for the lengths R1 , R2 gives, then: the angular gain of the cylindrical wave. The quantity v may be written as:
1 1 (l1 r )2
1 (l2 r )2 k kF
R1 + R2 l1 l2 = (l1 + l2 )r + + (r r ) v= y0 = (sin 1 + sin 2 ) (18.14.24)
2 l1 l2 l1 l2 F
2 Depending on the sign and relative sizes of the angles 1 and 2 , it follows that
1 1 x cos2 1 cos2 2 y2
= y (sin 1 + sin 2 )+ + + + v > 0 when P2 lies in the shadow region, and v < 0 when it lies in the illuminated
l1 l2 2 l1 l2 2
region. For large positive v, we may use Eq. (18.14.4) to obtain the asymptotic form of
1 2 1 2 the edge-diffraction coefcient Dedge :
= x + y + 2F y (sin 1 + sin 2 )
2F 2F
FF cos 1 + cos 2 jv2 /2 1 j jv2 /2 FF cos 1 + cos 2 1 j
1 2 1 1 Dedge = e e =
= x + (y + y0 )2 y02 l2 2 2v l2 2 2v
2F 2F 2F
where we dened the focal lengths F, F and the shift y0 : Writing F/l2 = l1 /(l1 + l2 ) and replacing v from Eq. (18.14.24), the F factor
cancels and we obtain:
1 1 1 1 cos2 1 cos2 2
= + , = + , y0 = F (sin 1 + sin 2 ) (18.14.21)
F l1 l2 F l1 l2 l1 (1 j)(cos 1 + cos 2 )
Dedge = (18.14.25)
Using these approximations in Eq. (18.13.6) and replacing r1 , r2 by l1 , l2 , we nd: l1 + l2 4 k(sin 1 + sin 2 )
jkA1 ejk(l1 +l2 ) This expression may be simplied further by dening the overall diffraction angle
E= (n l2 )(n l1 ) ejk(R1 +R2 l1 l2 ) dS
4l1 l2 S = 1 + 2 , as shown in Fig. 18.14.4 and using the trigonometric identity:
jkA1 ek(l1 +l2 ) 2 /2Fjk(y +y cos 1 + cos 2 1 + 2
= cot
2
ejkx dx dy
2
= (cos 1 + cos 2 )ejky0 /2F 0 ) /2F
4l1 l2 sin 1 + sin 2 2
The x -integral is over the range <x < and can be converted to a Fresnel Then, Eq. (18.14.25) may be written in the form:
integral with the change of variables u = x k/(F):
l1 (1 j)
Dedge = cot (18.14.26)
l1 + l2 4 k 2
jkx2 /2F F ju /2 2 F
e dx = e du = (1 j)
k k The asymptotic diffraction coefcient
is obtained from Eqs. (18.14.25) or (18.14.26)
by taking the limit l1 , which gives l1 /(l1 + l2 ) 1. Thus,
The y -integral is over the upper-half
of the xy-plane,that is, 0 y < . Dening
the Fresnel variables u = (y + y0 ) k/(F ) and v = y0 k/(F ), we nd: (1 j)(cos 1 + cos 2 ) (1 j)
Dedge = = cot (18.14.27)
4 k(sin 1 + sin 2 ) 4 k 2
/2F F F
ejk(y +y0 ) dy = eju
2 2
/2
du = (1 j)D(v) Eqs. (18.14.26) and (18.14.27) are equivalent to those given in [1300].
0 k v k
The two choices for the origin lead to two different expressions for the diffracted
where the function D(v) was dened in Eq. (18.14.1). Putting all the factors together, elds. However, the expressions agree near the forward direction, 0. It is easily
we may write the diffracted eld at the point P2 in the form: veried that both Eq. (18.14.1) and (18.14.26) lead to the same approximation for the
diffracted eld:
ejkl2 ejkl2 l1 1j
E = Eedge Dedge (straight-edge diffraction) (18.14.22) E = Eedge (18.14.28)
l2 l l1 + l 2 2 k
2
18.15 Geometrical Theory of Diffraction where k2 = 2 , and the two-dimensional 2 is in cylindrical coordinates:
Geometrical theory of diffraction is an extension of geometrical optics [13091324]. It 2 1 1 2
= + (18.15.2)
views diffraction as a local edge effect. In addition to the ordinary rays of geometrical 2 2
optics, it postulates the existence of diffracted rays from edges. The diffracted rays
The boundary conditions require that the tangential electric eld be zero on both
can reach into shadow regions, where geometrical optics fails.
sides of the conducting plane, that is, for = 0 and = 2. In the TE case, the
An incident ray at an edge generates an innity of diffracted rays emanating from the
tangential electric eld is Ez , and in the TM case, Ex = E cos E sin = E =
edge having different angular gains given by a diffraction coefcient Dedge . An example
of such a diffracted ray is given by Eq. (18.14.22).
(1/j)(Hz /), for = 0, 2. Thus, the boundary conditions are:
The edge-diffraction coefcient Dedge depends on (a) the type of the incident wave, (TE) Ez = 0, for = 0 and = 2
such as plane wave, or spherical, (b) the type and local geometry of the edge, such as a
Hz (18.15.3)
knife-edge or a wedge, and (c) the directions of the incident and diffracted rays. (TM) = 0, for = 0 and = 2
The diffracted eld and coefcient are usually taken to be in their asymptotic forms,
like those of Eq. (18.15.26). The asymptotic forms are derived from certain exactly In Fig. 18.15.1, we assume that 0 90o and distinguish three wedge regions
solvable canonical problems, such as a conducting edge, a wedge, and so on. dened by the half-plane and the directions along the reected and transmitted rays:
The rst and most inuential of all such problems was Sommerfelds solution of a
plane wave incident on a conducting half-plane [1286], and we discuss it below. reection region (AOB): 0
Fig. 18.15.1 shows a plane wave incident at an angle on the conducting plane transmission region (BOC): + (18.15.4)
occupying half of the xz-plane for x 0. The plane of incidence is taken to be the xy- shadow region (COA): + 2
plane. Because of the cylindrical symmetry of the problem, we may assume that there
The case when 90o 180o is shown in Fig. 18.15.2, in which has been
is no z-dependence and that the elds depend only on the cylindrical coordinates , .
redened to still be in the range 0 90o . The three wedge regions are now:
Two polarizations may be considered: TE, in which the electric eld is E = z Ez , and
TM, which has H = z Hz . Using cylindrical coordinates dened in Eq. (E.2) of Appendix
E, and setting /z = 0, Maxwells equations reduce in the two cases into:
Fig. 18.15.2 Plane wave incident on conducting half-plane.
Thus, the conditions 2 U + k2 U = 0 and 2 E + k2 E = 0 require: This immediately implies that a = 1 and A2 = k2 , so that A = k. Thus, f =
A cos = k cos . Next, we choose v = Ba cos a. Then:
E2 D + 2 E D = 0 ln E)
2 D + 2( D = 0 (18.15.6)
f = A(
cos
sin )
If we assume that E is of the form E = ejf , where f is a real-valued function, then,
equating to zero the real and imaginary parts of 2 E + k2 E = 0, we nd for f : v = Baa1
cos a
sin a
f v = ABaa1 cos cos a + sin sin a = ABaa1 cos( a)
2 E + k2 E = E k2 f f + j2 f = 0 2 f = 0 , f f = k2 (18.15.7)
v v = B2 a2 2(a1)
Next, we assume that D is of the form:
Then, the last of the conditions (18.15.12) requires that:
v
jg(u)
D = D0 e du (18.15.8) 1 2 f v 2A12a cos( a)
= =1
v v v aB2 cos a
where D0 is a constant, v is a function of , , and g(u) is a real-valued function to be
which implies that a = 1/2 and B2 = 2A/a = 4A/. But since A = k, only the
determined. Noting that D = D0 ejg v and g = g (v) v, we nd:
case A = k is compatible with a real coefcient B. Thus, we have B2 = 4k/, or,
D = D0 ejg v , 2 D = D0 ejg 2 v jg (v) v v) B = 2 k/.
In a similar fashion, we nd that if we take v = Ba sin a, then a = 1/2, but now
ln E)
Then, it follows from Eq. (18.15.6) that 2 D + 2( D = 2 D + j
f D and: B = 4A/, requiring that A = k, and B = 2 k/. In summary, we have the
2
Sommerfelds solution involves the Fresnel diffraction coefcient of Eq. (18.14.1), The corresponding solutions (18.15.11) of the Helmholtz equation are:
which can be written as follows:
k 1/2
v U(, )= e jk cos
D(v) , v = 2 cos
1 1j 1 2
eju
2
/2
D(v)= + F(v) = du (18.15.10) (18.15.14)
1j 2 1j
jk cos k 1/2
U(, )= e D(v) , v = 2 sin
2
Therefore, we are led to choose g(u)= u2 /2 and D0 = 1/(1 j). To summarize,
we may construct a solution of the Helmholtz equation in the form: The function D(v) may be replaced by the equivalent form of Eq. (18.14.5) in order
to bring out its asymptotic behavior for large v:
2 U + k2 U = 0 , U = ED = ejf D(v) (18.15.11)
k 1/2
U(, )= ejk cos u(v)+d(v)ejv /2 ,
2
v = 2 cos
where f and v must be chosen to satisfy the four conditions: 2
k 1/2
2 f = 0, f f = k2 U(, )= ejk cos u(v)+d(v)ejv /2 ,
2
v = 2 sin
(18.15.12) 2
2 f v
2 v = 0, = g (v)= v
v v Using the trigonometric identities cos = 2 cos2 (/2)1 = 1 2 sin2 (/2), we
nd for the two choices of v:
It can be veried easily that the functions u = a cos a and u = a sin a are solu-
tions of the two-dimensional Laplace equation 2 u = 0, for any value of the parameter 1
k cos v2 = k cos 2 cos2 = k
a. Taking f to be of the form f = Aa cos a, we have the condition: 2 2
1
f = Aaa1
cos a
sin a f f = A2 a2 2(a1) = k2 k cos v2 = k cos + 2 sin2 = k
2 2
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
18.15. Geometrical Theory of Diffraction 839 840 18. Radiation from Apertures
Thus, an alternative form of Eq. (18.15.14) is: The rst two terms correspond to the incident and reected elds. The third term is
the diffracted eld. The algebraic signs of vi and vr are as follows within the reection,
k 1/2 transmission, and shadow regions of Eq. (18.15.4):
jk
U(, )= e jk cos
u(v)+e d(v) , v = 2 cos
2
(18.15.15) reection region: 0 < , vi > 0, vr > 0
k 1/2 transmission region: < < + , vi > 0, vr < 0 (18.15.21)
U(, )= ejk cos u(v)+ejk d(v) , v = 2 sin shadow region: + < 2, vi < 0, vr < 0
2
The unit-step functions will be accordingly present or absent resulting in the follow-
Shifting the origin of the angle still leads to a solution. Indeed, dening =
ing elds in these three regions:
, we note the property / = /, which implies the invariance of the Laplace
operator under this change. The functions U(, ) are the elementary solutions
reection region: Ez = Ei + Er + Ed
from which the Sommerfeld solution is built.
transmission region: Ez = Ei + Ed (18.15.22)
Considering the TE case rst, the incident plane wave in Fig. 18.15.1 is E = z Ei ,
shadow region: Ez = Ed
where Ei = E0 ejkr , with r = x cos + y sin and k = k(x cos + y sin ). It
follows that: where we dened the incident, reected, and diffracted elds:
k r = k(cos cos + sin sin )= k cos( )
Ei = E0 ejk cos i
(18.15.16)
Ei = E0 ejkr = E0 ejk cos() Er = E0 ejk cos r (18.15.23)
The image of this electric eld with respect to the perfect conducting plane will Ed = E0 ejk d(vi )d(vr )
be the reected eld Er = E0 ejkr r , where kr = k(x cos + y sin ), resulting in
Er = E0 ejk cos(+) . The sum Ei + Er does vanish for = 0 and = 2, but it also The diffracted eld is present in all three regions, and in particular it is the only one
vanishes for = . Therefore, it is an appropriate solution for a full conducting plane in the shadow region. For large vi and vr (positive or negative), we may replace d(v) by
(the entire xz-plane), not for the half-plane. its asymptotic form d(v)= (1 j)/(2v) of Eq. (18.14.6), resulting in the asymptotic
Sommerfelds solution, which satises the correct boundary conditions, is obtained diffracted eld:
by forming the linear combinations of the solutions of the type of Eq. (18.15.14):
1j 1 1
Ed = E0 ejk
2 vi vr
Ez = E0 ejk cos i D(vi )ejk cos r D(vr ) (TE) (18.15.17)
1j 1 1
= E0 ejk
where 22 k/1/2 cos(i /2) cos(r /2)
k 1/2 i
i = , vi = 2 cos which can be written in the form:
2
(18.15.18)
k 1/2 r ejk
r = + , vr = 2 cos Ed = E0 Dedge (18.15.24)
2 1/2
For the TM case, we form the sum instead of the difference: with an edge-diffraction coefcient:
Hz = H0 ejk cos i D(vi )+ejk cos r D(vr ) (TM) (18.15.19)
1j
1 1
Dedge = (18.15.25)
The boundary conditions (18.15.3) are satised by both the TE and TM solutions. 4 k i r
cos cos
As we see below, the choice of the positive sign in the denitions of vi and vr was 2 2
required in order to produce the proper diffracted eld in the shadow region. Using the
Using a trigonometric identity, we may write Dedge as follows:
alternative forms (18.15.15), we separate the terms of the solution as follows:
sin sin
Ez = E0 ejk cos i
u(vi )E0 e jk cos r
u(vr )+E0 e jk
d(vi )d(vr ) (18.15.20) 1j
1 1 1
= j 2 2
Dedge = (18.15.26)
4 k + k cos + cos
cos cos
2 2
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
18.16. Rayleigh-Sommerfeld Diffraction Theory 841 842 18. Radiation from Apertures
Eqs. (18.15.22) and (18.15.24) capture the essence of the geometrical theory of diffrac-
tion: In addition to the ordinary incident and reected geometric optics rays, one also
has diffracted rays in all directions corresponding to a cylindrical wave emanating from
the edge with a directional gain of Dedge .
For the case of Fig. 18.15.2, the incident and reected plane waves have propagation
vectors k = k(z cos y sin ) and kr = k(z cos + y sin ). These correspond to
the incident and reected elds:
Ez = E0 ejk cos i D(vi )ejk cos r D(vr )
(18.15.27)
Hz = H0 ejk cos i D(vi )+ejk cos r D(vr )
where, now:
k 1/2 i
i = + , vi = 2 sin
2
(18.15.28)
Fig. 18.16.1 Fields determined from their values on the xy-plane surface.
k 1/2 r
r = , v r = 2 sin
2
The choice of signs in vi and vr are such that they are both negative within the The surface integral over S can be ignored by noting that n is the negative of the
shadow region dened by Eq. (18.15.5). The same solution can also be obtained from radial unit vector and therefore, we have after adding and subtracting the term jkEG:
Fig. 18.15.1 and Eq. (18.15.17) by replacing by .
E G E G
G E dS = G + jkE E + jkG dS
S n n S r r
18.16 Rayleigh-Sommerfeld Diffraction Theory Assuming Sommerfelds outgoing radiation condition:
In this section, we recast Kirchhoffs diffraction formula in a form that uses a Dirich- E
r + jkE 0 , as r
let Greens function (i.e., one that vanishes on the boundary surface) and obtain the r
Rayleigh-Sommerfeld diffraction formula. In the next section, we show that this refor-
and noting that G = ejkr /4r also satises the same condition, it follows that the
mulation is equivalent to the plane-wave spectrum approach to diffraction, and use it
above surface integral vanishes in the limit of large radius r . Then, in the notation of
to justify the modied forms (18.1.2) and (18.1.3) of the eld equivalence principle. In
Eq. (18.10.4), we obtain the standard Kirchhoff diffraction formula:
Sec. 18.18, we use it to obtain the usual Fresnel and Fraunhofer approximations and
discuss a few applications from Fourier optics. G E
E(r)uV (r)= E
G
dS (18.16.2)
We will work with the scalar case, but the same method can be used for the vector S n n
case. With reference to Fig. 18.16.1, we we consider a scalar eld E(r) that satises the
Thus, if r lies in the right half-space, the left-hand side will be equal to E(r), and if r
source-free Helmholtz equation, (2 + k2 )E(r)= 0, over the right half-space z 0.
is in the left half-space, it will vanish. Given a point r = (x, y, z), we dene its reection
We consider a closed surface consisting of the surface S of a sphere of very large
relative to the xy plane by r = (x, y, z). The distance between r and a source point
radius centered at the observation point r and bounded on the left by its intersection S
r = (x , y , z ) can be written in terms of the distance between the original point r and
with the xy plane, as shown in the Fig. 18.16.1. Clearly, in the limit of innite radius,
the reected source point r = (x , y , z ):
the volume V bounded by S + S is the right half-space z 0, and S becomes the entire
xy plane. Applying Eq. (18.10.3) to volume V, we have:
R = |r r | = (x x )2 +(y y )2 +(z + z )2 = |r r |
E G
G(2 E + k2 E)E (2 G + k2 G) dV = G
E
dS (18.16.1) whereas
V S+S n n
R = |r r | = (x x )2 +(y y )2 +(z z )2
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
18.16. Rayleigh-Sommerfeld Diffraction Theory 843 844 18. Radiation from Apertures
This leads us to dene the reected Greens function: This approximation will be used in Sec. 18.18 to obtain the standard Fresnel diffrac-
tion representation. The quantity cos = z/R is an obliquity factor and is usually
ejkR
G (r, r )= = G(r r )= G(r r ) (18.16.3) omitted for paraxial observation points that are near the z axis.
4R Equation (18.16.5) expresses the eld at any point in the right half-space in terms of
and the Dirichlet Greens function: its values on the xy plane. In the practical application of this result, if the plane consists
of an innite opaque screen with an aperture S cut in it, then the integration in (18.16.5)
ejkR ejkR is restricted only over the aperture S. The usual Kirchhoff approximations assume that:
Gd (r, r )= G(r, r )G (r, r )= (18.16.4)
4R 4R (a) the eld is zero over the opaque screen, and (b) the eld, E(r ), over the aperture is
equal to the incident eld from the left.
For convenience, we may choose the origin to lie on the xy plane. Then, as shown
Eq. (18.16.5) is also valid in the vectorial case for each component of the electric eld
in Fig. 18.16.1, when the source point r lies on the xy plane (i.e., z = 0), the function
E(r). However, these components are not independent of each other since they must
Gd (r, r ) will vanish because R = R . Next, we apply Eq. (18.16.2) at the observation
satisfy E = 0, and are also coupled to the magnetic eld through Maxwells equations.
point r in the right half-space and at its reection in the left half-plane, where (18.16.2)
Taking into account these constraints, one arrives at a modied form of (18.16.5). We
vanishes:
G E pursue this further in the next section.
E(r) = G E dS , at point r
Sn n
G E 18.17 Plane-Wave Spectrum Representation
0= E G dS , at point r
S n n
where G stands for G(r r ). But on the xy plane boundary, G = G so that if we The plane-wave spectrum representation builds up a (single-frequency) propagating
subtract the two expressions we may eliminate the term E/n , which is the reason wave E(x, y, z) as a linear combination of plane waves ej(kx x+ky y+kz z) . The only as-
for using the Dirichlet Greens function: sumption is that the eld must satisfy the wave equation, which for harmonic time
dependence ejt is the Helmholtz equation
Gd
E(r)= E(r ) (G G ) dS = E(r ) dS
S n S n (2 + k2 )E(x, y, z)= 0 , k= (18.17.1)
c
On the xy plane, we have n = z, and therefore
where c is the speed of light in the propagation medium (assumed here to be homoge-
G G
G G
neous, isotropic, and lossless.) In solving the Helmholtz equation, one assumes initially
= and = =
n z
z =0
n
z z =0 z
z =0
a solution of the form:
Then, the two derivative terms double resulting in the Rayleigh-Sommerfeld diffrac- E(x, y, z)= E(kx , ky , z)ejkx x ejky y
tion formula [1285,1286]:
Inserting this into Eq. (18.17.1) and replacing x jkx and y jky , we obtain:
G
E(r)= 2 E(r ) dS (Rayleigh-Sommerfeld) (18.16.5)
S z 2
k2x k2y + + k 2
E(kx , ky , z)= 0
z2
The indicated derivative of G can be expressed as follows:
or, dening k2z = k2 k2x k2y , we have:
G
z 1 ejkR 1 ejkR
= jk + = cos jk + (18.16.6)
z z =0 R R 4R R 4R 2 E(kx , ky , z)
= (k2 k2x k2y )E(kx , ky , z)= k2z E(kx , ky , z)
where is the angle between the z-axis and the direction between the source and obser- z2
vation points, as shown in Fig. 18.16.1. For distances R , or equivalently, k 1/R, Its solution describing forward-moving waves (z 0) is:
one obtains the approximation:
E(kx , ky , z)= E(kx , ky , 0)ejkz z (18.17.2)
G
ejkR
= jk cos , for R (18.16.7)
z z =0
4R
If k2x + k2y < k2 , the wavenumber kz is real-valued and the solution describes a
By
adding instead of subtracting the above integrals, we obtain the alternative Greens function Gs = propagating wave. If k2x + k2y > k2 , then kz is imaginary and the solution describes an
G + G , having vanishing derivative on the boundary.
evanescent wave decaying with distance z. The two cases can be combined into one by with the understanding that z = 0. Thus, (18.17.8) takes the form of (18.16.5).
dening kz in terms of the evansecent square-root of Eq. (7.7.9) as follows: Next, we discuss the vector case as it applies to electromagnetic elds. To simplify
the notation, we dene the two-dimensional transverse vectors r = x x + y y and k =
k2 k2x k2y , if k2x + k2y k2 x kx + y ky , as well as the transverse gradient = x x + y y , so that the full three-
kz = (18.17.3) dimensional gradient is
j k2x + k2y k2 , if k2x + k2y > k2
= x x + +y y + z z = + z z
In the latter case, we have the decaying solution:
In this notation, Eq. (18.17.6) reads E(k , z)= g(k , z)E(k , 0), with g(k , z)=
k2x +k2y k2
E(kx , ky , z)= E(kx , ky , 0)ez , z0 ejkz z . The plane-wave spectrum representations (18.17.4) and (18.17.8) now are (where
the integral sign denotes double integration):
The complete space dependence is E(kx , ky , 0)ejkx xjky y ejkz z . The most general
solution of Eq. (18.17.1) is obtained by adding up such plane-waves, that is, by the spatial
d2 k
two-dimensional inverse Fourier transform: E(r , z) = E(k , 0) ejkz z ejk r
(2)2
dkx dky
E(x, y, z)= E(kx , ky , 0)ejkx xjky y ejkz z (18.17.4) = E(r , 0)g(r r , z) d2 r
(2)2
(18.17.10)
This is the plane-wave spectrum representation. Because kz is given by Eq. (18.17.3), and
this solution is composed, in general, by both propagating and evanescent modes. Of d2 k
g(r , z)= ejkz z ejk r (18.17.11)
course, for large z, only the propagating modes survive. Setting z = 0, we recognize (2)2
E(kx , ky , 0) to be the spatial Fourier transform of the eld, E(x, y, 0), on the xy plane: In the vectorial case, E(r , z) is replaced by a three-dimensional eld, which can be
decomposed into its transverse x, y components and its longitudinal part along z:
dkx dky
E(x, y, 0) = E(kx , ky , 0)ejkx xjky y
(2)2
(18.17.5) E = x Ex + y Ey + z Ez E + z Ez
jkx x+jky y
E(kx , ky , 0) = E(x, y, 0)e dx dy
The Rayleigh-Sommerfeld and plane-wave spectrum representations apply separately
to each component and can be written vectorially as
As in Chap. 3, we may give a system-theoretic interpretation to these results. Den-
ing the propagation spatial lter g(kx , ky , z)= ejkz z , then Eq. (18.17.2) reads:
d2 k
E(r , z)= E(k , 0) ejkz z ejk r = E(r , 0)g(r r , z) d2 r
(2)2
E(kx , ky , z)= g(kx , ky , z)E(kx , ky , 0) (18.17.6) (18.17.12)
Because E must satisfy the source-free Gausss law, E = 0, this imposes certain
This multiplicative relationship in the wavenumber domain translates into a convo-
constraints among the Fourier components E that must be taken into account in writing
lutional equation in the space domain. Denoting by g(x, y, z) the spatial inverse Fourier
(18.17.12). Indeed, we have from (18.17.12)
transform of g(kx , ky , z)= ejkz z , that is,
d2 k
jkx xjky y jkz z dkx dky E = j k E(k , 0) ejkz z ejk r =0
g(x, y, z)= e e (18.17.7) (2)2
(2)2
which requires that k E(k , 0)= 0. Separating the transverse and longitudinal parts,
we may write Eq. (18.17.4) in the form:
we have:
k E
k E = k E + kz Ez = 0 Ez =
E(x, y, z)= E(x , y , 0)g(x x , y y , z)dx dy (18.17.8) kz
It follows that the Fourier vector E must have the form:
Eq. (18.17.8) is equivalent to the Rayleigh-Sommerfeld formula (18.16.5). Indeed, it
follows from Eq. (D.19) of Appendix D that k E
E = E + z Ez = E z (18.17.13)
kz
G G ejkR
g(x x , y y , z)= 2 =2 , G= , R = |r r | (18.17.9)
z z 4R
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
18.17. Plane-Wave Spectrum Representation 847 848 18. Radiation from Apertures
and, therefore, it is expressible only in terms of its transverse components E . Then, where we took S to be the xy plane with the unit vector n = z. Then, Eqs. (18.10.13) and
the correct plane-wave spectrum representation (18.17.12) becomes: (18.10.14) can be written as:
1 1
k E (k , 0) d2 k E(r)= h)+
( e, H(r)= e)+
( h
E(r , z)= E (k , 0)z ejkz z ejk r (18.17.14) j j
kz (2)2
Noting that e, h are transverse vectors and using some vector identities and the de-
But from the Weyl representations (D.18) and (D.20), we have with G = ejkr /4r :
composition = + z z , we can rewrite the above in a form that explicitly separates
G d2 k k d2 k the transverse and longitudinal parts, so that if r is in the right half-space:
2 = ejkz z ejk r , 2 G = ejkz z ejk r
z (2)2 kz (2)2 1
E(r) = h)2z h + z z (
( h) + e + z z e
Then, (18.17.14) can be written convolutionally in the form: j
(18.17.20)
G 1
E(r , z)= 2 E z G E d2 r (18.17.15) H(r) = e)2z e + z z (
( e) + h + z z h
z j
where here G = ejkR /4R with R = |r r | and z = 0, that is, R = |r r |2 + z2 , If r is chosen to be the reected point r on the left half-space, then G = G and the
and E in the integrand stands for E (r , 0). Eq. (18.17.15) follows from the observation vectors e, h remain the same, but the gradient with respect to r is now = z z ,
that in (18.17.14) the following products of Fourier transforms (in k ) appear, which arising from the replacement z z. Thus, replacing z z in (18.17.20) and
become convolutions in the r domain: setting the result to zero, we have:
1
k 0= h)2z h z z (
( h) + e z z e
E (k , 0)ejkz z and E (k , 0) ejkz z j
kz
(18.17.21)
1
Because E (r , 0) does not depend on r, it is straightforward to verify using some 0= e)2z e z z (
( e) + h z z h
vector identities that j
The same results can be derived more directly by using the Franz formulas (18.10.13) 2 2
E(r) = h)2z h + z z (
( h) = h)
(
and making use of the extinction theorem as we did in Sec. 18.16. Applying (18.10.13) j j (18.17.24)
to the closed surface S + S of Fig. 18.16.1, and dropping the S term, it follows that H(r) = 2 h + 2 z z h = 2 h
the left-hand side of (18.10.13) will be zero if the point r is not in the right half-space.
To simplify the notation, we dene the vectors: Eqs. (18.17.18) and (18.17.24) are equivalent to applying the Franz formulas with the
eld-equivalent surface currents of Eqs. (18.1.2) and (18.1.3), respectively.
e= G(z E)dS , h= G(z H)dS As in the scalar case, the vector method is applied in practice by dividing S into
S S two parts, the screen over which the tangential elds are assumed to be zero, and the
In the notation of Eq. (18.10.12), we have e = A ms / and h = A s /. aperture over which the elds are assumed to arise from incident elds from the left.
Example 18.17.1: Oblique Plane Wave. Here, we show that the plane-wave spectrum method where we used the Taylor series expansion 1 + x 1 + x/2. Assuming also that R or
correctly generates an ordinary plane wave from its transverse values at an input plane. z is much greater than the wavelength of the wave, z , so that k 1/R we obtain
Consider a TM electromagnetic wave propagating at an angle 0 with respect to the z axis,
2 /2z)
as shown in the gure below. The electric eld at an arbitrary point, and its transverse 2z 1 ejkR ejk(z+|r r | jk 2 /2z
jk + jk = ejkz ejk|r r | (18.18.2)
part evaluated on the plane z = 0, are given by R R 4R 2z 2z
This amounts to replacing the propagator impulse response g(r , z) by the approx-
imation of Eq. (18.18.2):
It follows that the spatial Fourier transform of E (r , 0) will be
0
E (k , 0)= x E0 cos 0 ejk r ejk r d2 r = x E0 cos 0 (2)2 (k k0 )
jk jk
ejkz ejk|r | ejkz ejk(x +y2 )/2z
2 2
/2z
g(r , z)= = (Fresnel) (18.18.4)
2z 2z
Then, the integrand of Eq. (18.17.14) becomes
Noting that k = 2/, the constant factor in front is often written as:
k E
E z = E0 (x cos 0 z sin 0 )(2)2 (k k0 )
kz jk j
=
2z z
and Eq. (18.17.14) gives
The above approximations can also be understood from the plane-wave spectrum
d2 k point of view. The Fourier transform of (18.18.4) is obtained from the following Fourier
E(r , z) = E0 (x cos 0 z sin 0 )(2)2 (k k0 )ejkz z ejk r
(2)2 integral, which is a special case of Eq. (3.5.18):
= E0 (x cos 0 z sin 0 )ej(kx x+kz z)
0 0
jk
ejkx
2 2
/2z
ejkx x dx = ejkx z/2k (18.18.5)
which is the correct expression for the plane wave. For a TE wave a similar result holds.
2z
Applying (18.18.5) with respect to the x and y integrations, we obtain the two-
dimensional spatial Fourier transform of g(r , z):
18.18 Fresnel Diffraction and Fourier Optics
g(k , z)= ejkz ej(kx +ky )z/2k = ejkz ej|k |
2 2 2
z/2k
The Fresnel approximation for planar apertures is obtained from the Rayleigh-Sommerfeld (18.18.6)
formula (18.16.5). Using (18.16.6), we have:
Then, Eq. (18.18.3) can be written in its plane-wave spectrum form:
jkR
1 e 2z
E(r , z)= E(r , 0)
jk + d2 r (18.18.1) d2 k
E(r , z)= ejkz E(k , 0) ej|k | ejk r
2
R R 4R z/2k
S (18.18.7)
(2)2
where R = (x x )2 +(y y )2 +z2 = |r r |2 + z2 . The Fresnel approximation
assumes that z is large enough such that |r r | z, which can realized if the Eq. (18.18.6) can be obtained from the exact form g(k , z)= ejkz z by assuming
aperture has dimension d so that |r | < d, and one assumes that the observation point that for large z the evanescent modes will be absent and assuming the approximation
r remains close to the z-axis (the paraxial approximation) such that |r | < d, and z is k2x + k2y k2 for the propagating modes. Then, we can write:
chosen such that z d. Then, we can approximate R as follows:
|k |2 1 |k |2 |k |2
kz = k2 |k |2 = k 1 k 1 =k
|r r |2 1 |r r |2 |r r |2 k2 2 k2 2k
R = |r r | + z = z 1 +
2 2 z 1+ =z+
z2 2 z2 2z
and, hence
ejkz z ejkz ej|k |
2
z/2k
Because of the assumption |k | k, the maximum transverse wavenumber will Noting that (k )= k|r |2 /2z at k = kr /z, we nally nd:
be |k | = k = 2/, and correspondingly the smallest achievable transverse spatial
jk
resolution will be r 1/|k | , that is, about one wavelength. This is the basic E(r , z)= ejkz ejk|r |
2
/2z
E(k , 0) kr (Fraunhofer) (18.18.10)
2z k =
diffraction limit of optical instruments, such as lenses and microscopes. z
Near-eld optics methods [530548], where the evanescent modes are not ignored, A simpler way of deriving (18.18.10) is by using (18.18.3) and noting that
overcome this limitation and can achieve much higher, subwavelength, resolutions.
2 /2z 2 /2z
ejk|r r | = ejk|r | ejk|r | ejkr r /z
2
/2z
Although ordinary lenses are diffraction-limited, it has been shown recently [394]
that superlenses made from metamaterials having negative refractive index can achieve 2 / 2z
The factor ejk|r | can be ignored if we assume that k|r |2 z, which leads to:
perfect resolution.
In the special case when the aperture eld E(x , y , 0) depends only on one trans- jk jkz jk|r |2 /2z
E(r , z)= e e E(r , 0) ejkr r /z d2 r
verse coordinate, say, E(x , 0), the dependence of (18.18.3) on the y direction can be 2z
integrated out using the integral and the last integral factor is recognized as E(k , 0) evaluated at k = kr /z.
jk 2 /2z Example 18.18.1: Knife-Edge Diffraction. Let us revisit the problem of knife-edge diffraction
ejk(yy ) dy = 1 (18.18.8) using the Fresnel formula (18.18.3). The innite edge is along the y direction and it occu-
2z
pies the region x < 0, as shown in the gure below. The incident plane-wave eld and the
and we obtain the following one-dimensional version of the Fresnel formula, written diffracted eld at distance z are:
convolutionally and in its plane-wave spectrum form:
Einc (x, z)= E0 ejkz
jk jkz 2
E(x, z) = e E(x , 0) ejk(xx ) /2z dx
2z jk jkz 2
(18.18.9) E(x, z)= e E(x , 0) ejk(xx ) /2z dx
dkx 2z
= ejkz
2
E(kx , 0) ejkx z/2k
2
The Fraunhofer approximation is a limiting case of the Fresnel approximation when At the input plane, E(x , 0)= E0 , for x 0, and E(x , 0)= 0, for x < 0. Then, the above
integral becomes:
the distance z is even larger than that in the Fresnel case. More precisely, it is obtained
in the far-eld limit when k|r |2 z, or, d2 z, where d is the size of the aperture. jk )2 /2z
E(x, z)= E0 ejkz ejk(xx dx
2z 0
In this approximation, the eld E(r , z) becomes proportional to the Fourier trans-
form E(k , 0) of the eld at the input plane. It is similar to the radiation-eld approxi- Making the change of variables,
!
mation in which the radiation elds are proportional to Fourier transform of the current
k k
sources, that is, to the radiation vector. (x x)= u, v= x
2z 2 z
A direct way of deriving the Fraunhofer approximation is by applying the stationary-
the above integral can be reduced to the Fresnel integral F(x) of Appendix F:
phase approximationEq. (F.22) of Appendix Fto the evaluation of the plane-wave
spectrum integral (18.18.7). Dene the phase function j 2 /2 1 1j
E(x, z)= E0 ejkz eju du = E0 ejkz F(v)+
2 2 v 1j 2
|k |2 z k2x z ky z
(k )= k r = kx x + ky y x (kx )+y (ky ) This is identical (up to the paraxial assumption) to the case discussed in Sec. 18.14. When
2k 2k 2k
x < 0, the observation point lies in the shadow region.
Then, the stationary-point with respect to the kx variable is Example 18.18.2: Diffraction by an innite slit. Consider an innite slit on an opaque screen.
The y-dimension of the slit is innite and its x-size is |x| a, as shown on the left in the
kx z xk z
x (kx )= x=0 kx = ,
x (kx )= gure below. The same gure also shows an opaque strip of the same size.
k z k
and similar expressions for y (ky ). Thus, vectorially, the stationary point is at k =
kr /z. Using Eq. (F.22), we obtain:
2
j(k ) d k 2j 2j j(k ) 1
E(k , 0) e E( k , 0) e
(2)2 x (kx ) y (ky ) (2)2 k = kr
The incident eld is a uniform plane wave, Einc (x, z)= E0 ejkz , whose values on the slit 2
diffraction by a slit
2
diffraction by a strip
are E(x , 0)= E0 . The diffracted eld at distance z is given by Eq. (18.18.9):
z=a z=a
a z = 20 a z = 20 a
jk )2 /2z jk )2 /2z 1.5 z = 100 a 1.5 z = 100 a
E(x, z)= ejkz E(x , 0) ejk(xx dx = E0 ejkz ejk(xx dx
2z 2z
|1 D(x, z)|
a
|D(x, z)|
1 1
The integral can be reduced to the Fresnel integral F(x) of Appendix F by making the
change of variables:
! 0.5 0.5
k k
(x x)= u, v = (a x)
2z 2 z
0 0
4 3 2 1 0 1 2 3 4 4 3 2 1 0 1 2 3 4
so that x/a x/a
a v+
jk )2 /2z j 2 /2 F(v+ )F(v )
ejk(xx dx = eju du = D(x, z) Fig. 18.18.1 Fresnel diffraction by a slit and a strip
2z a 2 v 1j
where we used j/2 = 1/(1 j). Thus, E(x, z) becomes:
18.19 Lenses
E(x, z)= ejkz D(x, z) (18.18.11)
In Fourier optics applications, one considers the passage of light through various optical
For the case of the strip, the limits of integration are changed to: elements that perform certain functions, such as Fourier transformation using lenses.
For example, Fig. 18.19.1 shows an input eld starting at aperture plane a, then prop-
a
jk )2 /2z F()F(v+ )+F(v )F() agating a distance z1 to a thin optical element where it is modied by a transmittance
+ ejk(xx dx = = 1 D(x, z)
2z a 1j function, and then propagating another distance z2 to an aperture plane b.
where we used F()= F()= (1 j)/2. Thus, the diffracted eld in the strip case
will be given by the complementary expression
E(x, z)= ejkz 1 D(x, z) (18.18.12)
This result is an example of the Babinet principle [634] that the sum of the elds from an
aperture and its complementary screen is equal to the eld in the absence of the aperture:
Fig. 18.18.1 shows the diffracted patterns in the two cases. The graphs plot the quantities
|D(x, z)| and |1 D(x, z)| versus x in the two cases.
The slit was chosen to be four wavelengths wide, a = 4, and the diffracted patterns Fig. 18.19.1 Field propagated from plane a to plane b through a thin optical element.
correspond to the near, medium, and far distances z = a, z = 20a, and z = 100a. The
latter case corresponds to the Fraunhofer pattern having a small ratio a2 /z = 1/25. Assuming that the input/output relationship of the optical element is multiplicative,
For example, for the slit case, the corresponding pattern approximates (but it is not quite E+ (r )= T(r )E (r ), the relationship between the output eld at plane b to the input
there yet) the typical sinc-function Fourier transform of the rectangular slit distribution eld at plane a is obtained by successively applying the propagation equation (18.17.10):
E(x , 0)= E0 , for a x a :
a Eout (r ) = g(r u , z2 )E+ (u ) d2 u = g(r u , z2 )T(u )E (u ) d2 u
sin(kx a)
E(kx , 0)= E0 ejkx x dx = 2aE0 S S
a kx a
= g(r u , z2 )T(u )g(u r , z1 )Ein (r ) d2 u d2 r
where this is to be evaluated at kx = kx/z for the diffraction pattern E(x, z). The property S
that at the center of the strip, x = 0, the diffracted pattern is not zero is an example of the
so-called Poissons spot [634].
= h(r , r )Ein (r ) d2 r
S
where the overall transfer function from plane a to plane b will be:
h(r , r )= g(r u , z2 )T(u )g(u r , z1 ) d2 u (18.19.1)
S
where we labeled the spatial x, y coordinates by r , u , and r on the planes (a), the
element, and plane (b).
In a similar fashion, one can work out the transfer function of more complicated
congurations. For example, passing through two transmittance elements as shown in
Fig. 18.19.2, we will have:
Eout (r )= h(r , r )Ein (r ) d2 r (18.19.2)
S
ejka ejkg b
where we are assuming a thin lens, which allows us to ignore the bending of the ray
arising from refraction. Because, a + b = d, we have for the net phase:
where F is the focal length. Because the Fresnel propagation factor ejk|r | /2z also has
2 (n 1)kx2
(x)= knd (n 1)ka = knd
the same type of quadratic phase, but with the opposite sign, it is possible for lenses to act 2R
as spatial dispersion compensation elements, much like the dispersion compensation If we make up a convex lens by putting together two such spherical lenses with radii
and pulse compression lters of Chap. 3. They have many uses, such as compensating R1 and R2 , as shown in Fig. 18.19.3, then the net phase change between the front and
propagation effects and focusing the waves on appropriate planes, or performing spatial back planes will be, ignoring the constant nkd terms:
Fourier transforms.
The transmittance function (18.19.4) can be derived with the help of Fig. 18.19.3, 1 1 kx2 kx2
(x)= (n 1) + (18.19.5)
which shows a wave entering from the left a (convex) spherical glass surface at a distance R1 R2 2 2F
x from the axis.
where we dened the focal length F of the lens through the lensmakers equation,
Let R and d denote the radius of the spherical element, and its maximum width
along its axis to the at back plane, and let n be its refractive index. The wave travels a
1 1 1
distance a in air and a distance b in the glass. If k is the free-space wavenumber, then = (n 1) + (18.19.6)
F R1 R2
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
18.19. Lenses 857 858 18. Radiation from Apertures
In a two-dimensional description, we replace x2 by |r |2 = x2 + y2 . Thus, the phase The integral is equal to the Dirac delta, (2)2 (kr /F)= (2)2 (r )F2 /k2 . Thus,
change and corresponding transmittance function will be:
j2F jkF
E(r , F)= e E0 (r )
k|r |2 j(r ) jk|r |2 /2F
k
(r )= T(r )= e =e
2F which is sharply focused onto the focal point r = 0 and z = F. For the second case
Some examples of the various effects that can be accomplished with lenses can be depicted on the right in Fig. 18.19.4, we rst note that the paraxial approximation for a
obtained by applying the congurations of Figs. 18.19.1 and 18.19.2 with appropriate spherical wave placed at the origin is:
choices of the aperture planes and focal lengths. We will use the Fresnel approximation
ejkr jk jkz jk|r |2 /2z |r |2
(18.18.4) for g(r , z) in all of the examples and assume that the transmittance (18.19.4) E0 E0 e e , r z
4r 4z 2z
extends over the entire xy planein effect, we are replacing the lens with the ideal case
of an innitely thin transparency with transmittance (18.19.4). If this source is placed at the left focal point of the lens, then, the diverging paraxial
The main property of a lens is to turn an incident plane wave from the left into spherical wave, after traveling distance z = F, will arrive at the left plane of the lens:
a spherical wave converging on the lens focus on the right, and similarly, if a source
jk 2 /2F 2 /2F
of a spherical wave is placed at the focus on the left, then the diverging wave will be Ein (r , 0)= E0 ejkF ejk|r | E1 ejk|r |
converted into a plane wave after it passes through the lens. These cases are shown in 4F
Fig. 18.19.4. The transmittance of the lens will compensate this propagation phase resulting into
a constant eld at the output plane of the lens, which will then propagate to the right
as a plane wave:
2 /2F 2 /2F
E(r , 0)= T(r )Ein (r , 0)= ejk|r | E1 ejk|r | = E1
Fig. 18.19.4 Spherical waves converging to, or diverging from, a lens focal point. where the integral was evaluated using twice the result (18.18.8). Thus, the transmitted
wave is a uniform plane wave propagating along the z-direction.
The case on the left corresponds to the choices z1 = 0 and z2 = F in Fig. 18.19.1, To see the Fourier transformation property of lenses, consider again the left picture
that is, the input plane coincides with the left plane of the lens. The incident wave has a in Fig. 18.19.4 with the output plane still placed at the right focal length z2 = F, but
constant amplitude on the plane Ein (r )= E0 . Noting that g(r u , 0)= 2 (r u ), take an arbitrary eld Ein (r ) incident at the left plane of the lens. The overall transfer
we obtain from Eq. (18.19.1) with z2 = F: function is still the same as in Eq. (18.19.7), thus, giving:
jk jkF jk|r |2 /2F jk|r r |2 /2F
h(r , r )= T(r )g(r r , F)= e e e E(r , F) = h(r r )Ein (r ) d2 r
2F
(18.19.8)
the quadratic phase terms combine as follows: jk jkF jk|r |2 /2F
= e e Ein (r ) ejkr r /F d2 r
2F
2 /2F 2 /2F
ejk|r | ejk|r r | = ejk|r | ejkr r /F
2
/2F
The last integral factor is recognized as the Fourier transform Ein (k ) evaluated at
and result in the following transfer function: wavenumber k = kr /F. Thus, we obtain:
jk jk
ejkF ejk|r |
2
/2F
h(r , r )= T(r )g(r r , F)= ejkF ejk|r | ejkr r /F
2
/2F E(r , F)= Ein (k ) (18.19.9)
(18.19.7) k =
k r
2F 2F F
Its integration with the constant input results in: This result is similar to the Fraunhofer case (18.18.10), but it is valid at the much
shorter Fresnel distance z = F, instead of the far-eld distances. It is analogous to
jk
ejkF E0 ejk|r | ejkr r /F d2 r
2
/2F the output of the pulse compression lter in chirp radar discussed in Chap. 3, see for
E(r , F)=
2F
example Eq. (3.10.14).
It is left as an exercise to show that the extra quadratic phase factor in (18.19.9) can The delta function forces r = (z2 /z1 )r , which is the same as (18.19.11). The
be eliminated by using the conguration of Fig. 18.19.1 with both aperture planes placed negative sign means that the image is upside down. Noting that
at the foci of the lens, that is, z1 = z2 = F, (known as a 2F system.)
kr kr z21 z1
Finally, let us look at the magnifying properties of a lens. Fig. 18.19.5 shows an + = r + r
z2 z1 k2 z2
image placed at distance z1 from the left and its image formed at distance z2 on the
right. It is well-known that the distances z1 , z2 must be related by: we obtain for the eld at the output plane:
1 1 1 z1 z1 2
Ein r ejk(z1 +z2 ) ejk|r | (z1 +z2 )/2z2
2
+ = (18.19.10) Eout (r )=
z1 z2 F z2 z2
which represents a scaled and reversed version of the input.
Some references on the Rayleigh-Sommerfeld diffraction theory, the plane-wave spec-
trum representation, and Fourier optics are [1285,1286] and [13251338].
18.20 Problems
18.1 Show that Eq. (18.4.9) can be written in the compact vectorial form:
ejkr jk ejkr
E = jk r z f r (z g) , H= r r (z f)+ z g
4r 4r
Fig. 18.19.5 Lens law of magnication. Similarly, show that Eqs. (18.4.10) and (18.4.11) can be written as:
The magnication law relates the size of the image to the size of the object: ejkr 2jk ejkr
E = 2jk r z f , H= r r (z f)
4r 4r
x2 z2
M= = (magnication ratio) (18.19.11) ejkr ejkr
x1 z1 E = 2jk r r (z g) , H = 2jk r z g
4r 4r
These properties can be derived by tracing the rays emanating from the top of the
18.2 Prove the rst pair of equations for E, H of the previous problem by working exclusively
object. The ray that is parallel to the lens axis will bend to pass through the focal point with the Kottler formulas (18.4.2) and taking their far-eld limits.
on the right. The ray from the top of the object through the left focal point will bend to
18.3 Explain in detail how the inequality (18.6.12) for the aperture efciency ea may be thought
become parallel to the axis. The intersection of these two rays denes the top point of
of as an example of the Schwarz inequality. Then, using standard properties of Schwarz
the image. From the geometry of the graph one has: inequalities, prove that the maximum of ea is unity and is achieved for uniform apertures.
x1 x2 x2 x1 As a reminder, the Schwarz inequality for single-variable complex-valued functions is:
= and =
2
z1 F F z2 F F
b
b b
f (x)g(x) dx
|f (x)|2 dx |g(x)|2 dx
The consistency of the equations requires the condition (z1 F)(z2 F)= F2 , which
a
a a
is equivalent to (18.19.10). Then, Eq. (18.19.11) follows by replacing F from (18.19.10)
18.4 To prove the equivalence of the Kirchhoff diffraction and Stratton-Chu formulas, (18.10.6)
into the ratio x2 /x1 = (z2 F)/F.
and (18.10.7), use the identities (C.29) and (C.32) of Appendix C, to obtain:
To understand (18.19.10) and (18.19.11) from the point of view of Fresnel diffrac-
1
tion, we note that the transfer function (18.19.1) involves the following quadratic phase j G J + G + G J m dV = j G J G + J m G dV
V V
factors, with the middle one being the lens transmittance:
2 /2z n G + n J m G dS
ejk|r u | ejk|u | ejk|u r |
2 2
/2z2 /2F 1
S
jk|r |2 /2z1 Then, using the identity (C.33), show that Eq. (18.10.6) can be rewritten in the form:
= ejk|r | ejk(1/z1 +1/z2 1/F)|u | ejku (r /z2 +r /z1 )
2 2
/2z2 /2
e
Because of Eq. (18.19.10), the term that depends quadratically on u cancels and E(r)= j G J G + J m G dV
V
one is left only with a linear dependence on u . This integrates into a delta function in
(18.19.1), resulting in + n G + n J m G dS
S
jkejkz1 jkejkz2 jk|r |2 /2z2 jk|r |2 /2z1 kr kr
h(r , r )= e e (2)2 + n G E (n E)
G G n (
E)(n E) G dS
2z1 2z2 z2 z1
Get useful study materials from www.rejinpaul.com
S
www.rejinpaul.com
18.20. Problems 861 862 18. Radiation from Apertures
Finally, use / = E and E + J m = jH to obtain (18.10.7). wave below the edge is given by Eqs. (18.14.22)(18.14.24), except that the eld at the edge
is Eedge = E0 , and the focal lengths are in this case F = l2 and F = l2 / cos2 2
18.5 Prove the equivalence of the Stratton-Chu and Kottler formulas, (18.10.7) and (18.10.10), by
rst proving and then using the following dual relationships: Finally, show that the asymptotic diffracted eld (when l2 ), is given near the forward
direction 0 by:
ejkl2 1 j
G (J )
j G dV = G j(n E)
(n H) G E = Eedge
V S l2 2 k
jm G (J m )
G dV = G + j(n H)
(n E) G 18.11 Assume that the edge in the previous problem is a perfectly conducting screen. Using the
V S
eld-equivalence principle with effective current densities on the aperture above the edge
To prove these, work component-wise, use Maxwells equations (18.2.1), and apply the di- J s = 0 and J ms = 2n E a , and applying the usual Fresnel approximations, show that the
vergence theorem on the volume V of Fig. 18.10.1. diffracted eld calculated by Eq. (18.4.1) is is still given by Eqs. (18.14.22)(18.14.24), except
that the factor cos 1 + cos 2 is replaced now by 2 cos 2 , and that the asymptotic eld and
18.6 Prove the equivalence of the Kottler and Franz formulas, (18.10.10) and (18.10.11), by using
edge-diffraction coefcient are:
the identity ( A)= ( A)2 A, and by replacing the quantity k2 G(r r ) by
(3) (r r )2 G. Argue that the term (3) (r r ) makes a difference only for the volume ejkl2 (1 j)2 cos 2
integrals, but not for the surface integrals. E = E0 Dedge , Dedge =
l2 4 k(sin 1 + sin 2 )
18.7 Prove the equivalence of the modied Stratton-Chu and Kirchhoff diffraction integral for-
mulas of Eq. (18.12.1) and (18.12.2) by using the identity (C.42) of Appendix C and replacing Show that this expression agrees with the exact Sommerfeld solution (18.15.26) at normal
E = 0 and E = jH in the source-less region under consideration. incidence and near the forward diffracted direction.
18.8 Prove the equivalence of the Kottler and modied Stratton-Chu formulas of Eq. (18.12.1) and 18.12 A uniform plane wave, E(x, z)= E0 ejk(x sin 0 +z cos 0 ) , is incident obliquely on a lens at an
(18.12.2) by subtracting the two expressions, replacing jE = H , and using the Stokes angle 0 with the z axis, as shown in the gure below.
identity (C.38) of Appendix C.
18.9 Consider a reector antenna fed by a horn, as shown
on the right. A closed surface S = Sr + Sa is such
that the portion Sr caps the reector and the portion
Sa is an aperture in front of the reector. The feed
lies outside the closed surface, so that the volume V
enclosed by S is free of current sources.
Applying the Kottler version of the extinction theorem of Sec. 18.10 on the volume V, show
that for points r outside V, the eld radiated by the induced surface currents on the reector
Using similar methods as for Fig. 18.19.4, show that after passing through the lens, the wave
Sr is equal to the eld radiated by the aperture elds on Sa , that is,
will converge onto the shifted focal point with coordinates z = F and x = F sin 0 .
1
Conversely, consider a point source of a spherical wave starting at the point z = F and
E rad (r) = k2 G J s + J s G dS
j Sr x = F sin 0 . Show that upon passage through the lens, the spherical wave will be converted
into the obliquely moving plane wave E(x, z)= E1 ejk(x sin 0 +z cos 0 ) . What is E1 ?
1
= G + j(n E )
k2 G(n H )+ (n H ) G dS
j Sa 18.13 Consider the three lens congurations shown below. They are special cases of Figs. 18.19.1
and 18.19.2, with appropriate choices for the input and output aperture planes a and b.
where the induced surface currents on the reector are J s = nr H and J ms = nr E, and
on the perfectly conducting reector surface, we must have J ms = 0.
This result establishes the equivalence of the so-called aperture-eld and current-distribution
methods for reector antennas [1345].
18.10 Consider an x-polarized uniform plane wave incident obliquely on the straight-edge aperture
of Fig. 18.14.4, with a wave vector direction k1 = z cos 1 + y sin 1 . First show that the
tangential elds at an aperture point r = x x + y y on the aperture above the straight-edge
are given by:
sin E0
E a = x E0 ejky 1
, H a = y cos 1 ejky sin 1
0
Then, using Kottlers formula (18.12.1), and applying the usual Fresnel approximations in
the integrand, as was done for the point source in Fig. 18.14.4, show that the diffracted
Working with Eqs. (18.19.1) and (18.19.3), show that the transfer functions h(r , r ) are
given as follows for the three cases:
jk
h(r , r )= e2jkF ejk(r r )/F , h(r , r )= ejkF ejk(r r )/F
2F
h(r , r )=
F1
r +
F1
r
Aperture Antennas
F2 F2
Show that the rst two cases perform a Fourier transformation as in Eq. (18.19.9), but without
the quadratic phase factors. Show that the third case, performs a scaling of the input with
a magnication factor M = F2 /F1
x 1 x
Ey (x )= E0 cos , Hx (x )= E0 cos (19.1.1)
a TE a
where TE = /K with K = 1 2c /2 = 1 (/2a)2 . Note that the boundary
conditions are satised at the left and right walls, x = a/2.
For larger apertures, such as a > 2, we may set K 1. For smaller apertures, such
as 0.5 a 2, we will work with the generalized Huygens source condition (18.5.7).
The radiated elds are given by Eq. (18.5.5), with fx = 0:
ejkr
E = jk c fy (, )sin
2r
(19.1.2)
ejkr
E = jk c fy (, )cos
2r
Introduction
First introduced in the late nineteenth century, horn antennas were found to be most useful in
high frequency applications such as microwave transmission. The first recorded horn antenna
to appear in an experiment was pyramidal horn used by J. Chunder Bose in 1897. In a lecture
at the Londons Royal Institution, Bose performed a demonstration using the horn, which he
referred to as acollecting funnel at an operating frequency of 60 GHz.
Interest in horn antennas dwindled at the turn of the century when Guglielmo Marconi
successfully achieved the first transatlantic wireless transmission. At that time it became
apparent that lower frequencies were better suited for long distance transmission, and the
horn was inadequate for this purpose. This disinterest continued until World War II, when
research and development focus shifted back to the microwave frequencies.[2]
Today, much of the theory behind the horn antenna is well defined. There are methods
to accurately calculate all of the major design variables. Horn antennas are now commonly
used for many applications including microwave communications, feeds for reflector antennas
and radar elements.
3
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Section 2
Literature Review
During World War II, as microwave theory became more popular amongst researchers, articles
on aperture and horn antennas started to appear in academic journals. The first of which
to focus on a true horn antenna was Theory of the Electromagnetic Horn written by W.L.
Barrow and L.J. Chu for the Proceedings of the Institute of Radio Engineers in January of
1939. This paper was a theoretical analysis of the operation of the horn antenna and provided
the reader with a set of equations that could be used to design horns for radio communication.
Since WWII, there have been many paper released on horn antennas focusing on both
theory and applications. Methods have been derived for accurate calculation of many prop-
erties of these antennas such as gain and directivity. In recent years though, as microstrip
antennas have become more popular, the number of articles written on horns has declined.
The focus has shifted from fundamental theory to improving the analysis and performance of
horn antennas, as well as presenting new uses for horns.
In [5] the authors develop a new technique for the analysis of horn antennas, taking
into account both their interior and exterior cross sections. The authors state that three-
dimensions techniques that existed are numerically too complex and limit their use. They
therefore present their technique using a modal scattering matrix based on the the Mode
Matching Method. This new method is compared to measured results from an experiment
4
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
using a conical-ridged horn antenna. The results of the comparison show their method holds
as a accurate method for this type of analysis.
The authors of [3] were not concerned about analysis techniques, but rather about im-
proving the gain of a pyramidal horn antenna. Their idea is based on placing strips of metal
in the horn near the throat of the antenna. Their research shows that these simple baffles can
increase the gain as much as 10dB over a range of 1GHz for some antennas. They present this
idea as a cheap and simple method of shortening pyramidal horns without the use of lenses,
but caution that the results are preliminary. They conclude by suggesting further avenues of
research in this technique, including looking at the relationship of the baffle and bandwidth.
In [4] the authors focus on the use of horn antennas for a very specific purpose, as an
excitation and receive element for a Ku-band active transmit array module. The authors
are concerned with why the gain and operational bandwidth for this configuration are lower
than that with the same active transmit array and microstrip patch arrays. By approaching
the problem both theoretically and experimentally, the authors show that microstrip array
provides a more uniform excitation of the transmit array module, thus achieving better per-
formance. They also suggest that near-field interactions between the horn and transmit array
module could contribute to the decreased performance.
Another paper that looks at applications of horn antennas is [6]. In this article, the authors
present a 256 element box-horn array for use in the 57.2-58.2GHz range. In the design of
this array, the authors pay particular attention to suppressing the side lobes, increasing the
antennas use in directional applications such as radar. The results of the design is a 18cm x
22cm x 3cm array with a measured gain of 36.2dBi and side lobe levels -35dB in the H-plane.
5
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Section 3
As stated earlier, horn antennas operate best at the megahertz and gigahertz frequency ranges.
As such, horns have traditionally been used in terrestrial microwave communications. They
can be found on many line-of-site microwave relay towers. These antennas are used in other
areas as well.
In 1999 NASA launched its Stardust satellite to track and collect samples from the Wild
2 comet. On board the satellite was a X-band conical horn antenna. This horn was used as
a feed element for a communications reflector dish on the satellite. It was 4 inches high and
2.5 inches in diameter and weighed less than 470 grams total. The ability to create horns of
this size makes them ideal for space applications where weight is major factor in launch cost.
Another example of small horn antennas in use is the box-horn array presented in [6]. This
antenna was designed to provide a large gain, yet be small enough to be used in applications
such as collision avoidance systems in automobiles. By using an array of horns and paying
particular attention to reducing the side lobes, the authors successfully designed an antenna
array that could easily fit on the back of most modern cars and be used as a transmit and
receive element for a radar system.
Horn antennas are not always this small though. In fact the worlds largest conical horn
is 14 feet in diameter and weighs approximately 4500 lbs. One interesting application of a
6
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
7
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Figure 3.1: Harold Ewen and his horn antenna. Taken from [7].
8
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Section 4
4.1 Introduction
The most commonly used horns today are pyramidal and conical, which can be manufactured
in many shapes and sizes. For the purposes of paper, we will look at the pyramidal horn and
the E and H-plane horn of which it is comprised. This section will derive the formulas for the
E-plane horn but leaves all other derivations up to the reader.
An E-plane horn antenna is an aperture antenna that is flared in the direction of the E-field.
This results in radiated fields that have a high directivity in the E-plane of the antenna. A
detailed geometry for a E-plane antenna is given in Figure 4.1.
From Balanis the aperture fields for the E-Plane Horn are
9
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
10
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
1 = e cos(e ) .
We can use these equations and the equivalent fields approach to find the surface current
densities for the aperture,
02
E1 j[ ky ]
J~s0 = cos( x0 )e (21 ) y (4.2)
a
and
ky 02
~ 0 = E1 cos( x0 )ej[ (21 ) ] x
M (4.3)
s
a
jkejkr
E = (L + N ) (4.4)
4r
jkejkr
E = (L + N ) (4.5)
4r
where
11
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
ZZ
0
N = [Jx cos cos + Jy cos sin Jz sin ]e+jkr cos
ds0
ZSZ
0
N = [Jx sin + Jy cos ]e+jkr cos
ds0
ZSZ
0
L = [Mx cos cos + My cos sin Mz sin ]e+jkr cos
ds0
ZSZ
0
L = [Mx sin + My cos ]e+jkr cos
ds0 .
S
we get,
Evaluating,
Za/2 "
ka
#
+jk(x0 sin cos ) a cos sin cos
cos x0 e dx0 = 2
2 2
a 2 ka
sin cos 2
a/2 2
12
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
and
b1 /2
ky 02 ky 2 1
Z r
j 21 +jk(x0 sin sin ) 0 1 j 2k
e e dx = e {[C(t2 ) C(t1 )] j [S(t2 ) S(t1 )]}
k
b1 /2
" " # #
ky 2 1
cos( kx2a )
r
a 1 j 2k cos sin
N = E1 e F (t1 , t2 ) (4.6)
2 k ( kx2a )2 ( 2 )2
where,
kx = k sin cos
ky = k sin sin
b1
t1 =
21
b1
t2 =
21
and
F (t1 , t2 ) = [C(t2 ) C(t1 )] j[S(t2 ) S(t1 )] .
Note: C(t) and S(t) are cosine and sine integrals, the solutions of which can be found in
Appendix III of the course text.
By similar analysis we can also find,
" " # #
2
cos( kx2a )
r
a 1 j ky2k1 cos
N = E1 e F (t1 , t2 )
2 k ( kx2a )2 ( 2 )2
" " # #
2
cos( kx2a )
r
a 1 j ky2k1
L = E1 e cos cos kx a 2 F (t1 , t2 ) (4.7)
2 k ( 2 ) ( 2 )2
" " # #
2
cos( kx2a )
r
a 1 j ky2k1
L = E1 e sin kx a 2 F (t1 , t2 ) .
2 k ( 2 ) ( 2 )2
13
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Now, we can substitute the results of (4.6) and (4.7) into (4.4) and (4.5) to get the radiated
E-fields,
" k2 " # #
a k1 E1 ejkr j y2k1 cos( kx2a )
E = j e sin (1 + cos ) kx a 2 F (t1 , t2 ) . (4.8)
8r ( 2 ) ( 2 )2
" 2
" # #
a k1 E1 e jkr j
ky
2k
1
cos( kx2a )
E = j e cos (1 + cos ) kx a 2 F (t1 , t2 ) (4.9)
8r ( 2 ) ( 2 )2
To find the directivity of an E-plane horn we must first find the maximum radiation,
r2
Umax = |E|2max . (4.10)
2
For most horn antennas |E|max is directed mainly along the z-axis (ie. = 0). Thus,
q
|E|max = |E |2max + |E |2max .
a k1 1
|E |max = (2) 2 |E1 sin F (t1 , t2 )| (4.11)
8r (2)
a k1 1
|E |max = (2) 2 |E1 cos F (t1 , t2 )| . (4.12)
8r (2)
14
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
and
S(t) = S(t) (4.14)
we get,
= 2[C(t) jS(t)]
so,
2a k1
|E |max = |E1 sin F (t)| (4.15)
8r
2a k1
|E |max = |E1 cos F (t)| . (4.16)
8r
2a2 k1
Umax = |E1 |2 |F (t)|2 (4.17)
3
where,
b1 b1
|F (t)|2 = [C 2 ( ) + S 2( )]
21 21
15
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Power Radiated
The total power radiated from a E-plane horn can be found by,
ZZ
1 ~0 H
~ 0 ) d~s
Prad = <e(E (4.18)
2
S
~ 0 and H
Using the E ~ 0 -fields over the aperture of the horn given in (4.1),
~0 H
<e(E ~ 0 )
(" # " #)
ky 02 ky 02 ky 02
j
E1 +j +j
= <e yE1 cos x0 e 21
x cos x0 e 21
+ zE1 sin x0 e 21
a a ka a
|E1 |2
= z cos2 x0
a
thus,
Za/2
b1
Prad = |E1 |2 cos2 x0 dx0 .
2 a
a/2
Now recall,
Z
1
cos2 udu = (u + sin u cos u) + C
2
Thus, we have,
ab1
Prad = |E1 |2 (4.19)
4
Now that we have found the maximum radiation and the power radiated from the horn, we
can find the directivity using the formula,
16
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
4Umax
DE =
Prad
(4.20)
64a1
= |F (t)|2
b1
Like the E-plane horn antenna, a H-plane horn is only flared in one direction. In this case it
is flared in the direction of the H-field, giving a better directivity in that plane. Figure 4.2 is
a diagram showing the geometry of the antenna.
17
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Using the same approach as presented for the E-plane horn, and starting from,
Ex0 = Hy0 = 0
x02
h i
jk 2
Ey0 (x0 )
= E2 cos( x0 )e 2
a1
h 02 i
0 0 E2 0 jk 2 x
Hx (x ) = cos( x )e 2
a1
2 = h cos h
b2 2
Umax = |E2 |2 [C(u) C(v)]2 + [S(u) S(v)]2
(4.21)
4
where,
r
1 2 a1
u= ( )
2 a1 2
r
1 2 a1
v= ( + )
2 a1 2
and that,
ba1
Prad = |E2 |2 .
4
4b2
[C(u) C(v)]2 + [S(u) S(v)]2
DH = (4.22)
a1
The final type of horn presented is the pyramidal horn antenna. This horn is a combination of
the E-plane and H-plane horns and as such is flared in both directions. A three dimensional
diagram of the pyramidal is given in Figure 4.3. For the E-plane and H-plane views, the
18
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
1 2
Umax = |E0 |2 [C(u) C(v)]2 + [S(u) S(v)]2
2
2 b1 2 b1
C ( ) + S ( )
21 21
a1 b 1
Prad = |E0 |2
4
(4.23)
81 2
[C(u) C(v)]2 + [S(u) S(v)]2
DP =
a1 b 1
2 b1 2 b1
C ( ) + S ( )
21 21
2
= DE DH
32ab
Also, for this type of horn antenna, the properties,
1
e 2 1 2
pe = (b1 b) ( ) (4.24)
b1 4
and
1
Note: p1 and p2 in figure 4.1 and figure 4.2 are referred to as pe and ph respectively in the pyramidal
horn equations.
19
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
1
h 2 1 2
ph = (a1 a) ( ) (4.25)
a1 4
are important. If these values are not equal, then the horn is not physically realizable.
20
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Section 5
" 2 # 12
e 1
pe = (b1 b)
b1 4
" 2 # 12
12.2 1
= (6.0 0.9)
6.0 4
10.05
and
21
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
" 2 # 12
h 1
ph = (a1 a)
a1 4
" 2 # 12
15.25 1
= (7.5 2.4)
7.5 4
10.05 .
1 2
Umax = |E0 |2 [C(u) C(v)]2 + [S(u) S(v)]2
2
2 b1 2 b1
C +S
21 21
186.05
= |E0 |2
2
[C(7.4089) C(7.5426)]2 + [S(7.4089) S(7.5426)]2
C 2 (6.687) + S 2 (6.687)
48.06 |E0 |2
=
and
a1 b 1
Prad = |E0 |2
4
11.25 |E0 |2
=
therefore
4Umax
Dp =
Prad
2
48.06 |E0 |
= 4
11.25 |E0 |2
= 53.68 .
22
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
!
b1/
e = tan1 2
e
therefore,
b1 = 2e tan e
= 2(10.5) tan(21.3)
8.2 .
Similarly,
a
1/
h = tan 1 2
h
therefore,
a1 = 2h tan h
= 2(9.2) tan(21.3)
7.2 .
23
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
64ae
DE = |F (t)|2
b1
64 (0.2) (10.5)
= |F (11.32)|2
(0.025) (8.2)
134.4
(0.463)
0.644
96.6
and
4h
[C(u) C(v)]2 + [S(u) S(v)]2
DH =
a1
4 (9.2)
[C(10.6) C(10.7)]2 + [S(10.6) S(10.7)]2
=
(7.2) (0.025)
115.6
(1.98)
0.18
1271.6
therefore
2
DP = DE DH
32ab
(0.025)2
= (96.6)(1271.6)
32(0.2)(1.2)
= 31.4
24
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Section 6
Summary
Horn antennas have been used for over a hundred years. In this time, the theory behind
them has been well defined and there are now many methods of accurately analyzing their
properties. It is for this reason that researchers are now focusing on more advanced horn topics
including gain enhancements and techniques for analysis of non-traditional horn shapes.
As shown in this paper, horn antennas have many different uses. They have long been
used as feed elements or point-to-point relay antennas, but they are now being used in new
applications such as collision avoidance radar.
This paper has presented the development of equations for calculating the directivity. In
doing this, it has also shown that there is a relationship between the directivity of the E-plane,
H-plane and pyramidal horn antennas. Also, this paper has tried to show users that care must
be taken by the designer to ensure that a pyramidal horn antenna can be constructed.
25
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
References
[1] C. A. Balanis. Antenna Theory: Analysis and Design. John Wiley & Sons, Inc., New
York, NY., 1997.
[2] A. W. Love. Electronmagnetic Horn Antennas. IEEE Press, New York, NY., 1976.
[3] R.L. Rogers M.A. Koerner. Gain enhancements of a pyramidal horn using e- and h-plane
metal baffles. IEEE Transactions on Antennas and Propgation, 48(4):529538, April 2000.
[4] H.J. Song M.E. Bialkowski. A ku-band active transmit-array module with a horn or
patch array as a signal launching/receiving device. IEEE Transactions on Antennas and
Propgation, 49(4):535541, April 2001.
[6] A.V. Raisanen T. Sehm, A. Lehto. A high-gain 58-ghz box-horn array antenna with
suppressed grating lobes. IEEE Transactions on Antennas and Propgation, 47(7):1125
1130, July 1999.
[9] L.J. Chu W.L. Borrow. Thoery of the electromagnetic horn. Proceedings of the Institute
of Radio Engineers, 27(1):5164, January 1939.
26
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
The horns can be also flared exponentially. This provides better matching in
a broad frequency band, but is technologically more difficult and expensive.
The rectangular horns are ideally suited for rectangular waveguide feeders.
The horn acts as a gradual transition from a waveguide mode to a free-space
mode of the EM wave. When the feeder is a cylindrical waveguide, the antenna
lH
R x
R0
a A
aH z
RH
2
A
l=
2
H R02 + , (18.1)
2
A
a H = arctan , (18.2)
0
2 R
( A a ) H .
l 1
RH = (18.3)
A 4
The two fundamental dimensions for the construction of the horn are A and
RH .
The tangential field arriving at the input of the horn is composed of the
transverse field components of the waveguide dominant mode TE10:
E y ( x ) = E0 cos x e j g z
a (18.4)
H x ( x) = E y ( x) / Z g
where
Zg = is the wave impedance of the TE10 mode;
2
1
2a
2
g 0
= 1 is the propagation constant of the TE10 mode.
2a
=
Here, 0 = 2 / , and is the free-space wavelength. The field that
is illuminating the aperture of the horn is essentially a spatially expanded
version of the waveguide field. Note that the wave impedance of the flared
waveguide (the horn) gradually approaches the intrinsic impedance of open
space , as A (the H-plane width) increases.
The complication in the analysis arises from the fact that the waves arriving
at the horn aperture are not in phase due to the different path lengths from the
horn apex. The aperture phase variation is given by
e j ( R R0 ) . (18.5)
Since the aperture is not flared in the y-direction, the phase is uniform in this
direction. We first approximate the path of the wave in the horn:
x
2
1 x 2
R= R02 + x 2 = R0 1 + R0 1 + . (18.6)
R0 2 R0
The last approximation holds if x 0 R0 , or A / 2 0 R0 . Then, we can assume
that
1 x2
R R0 . (18.7)
2 R0
Using (18.7), the field at the aperture is approximated as
j x2
Ea y ( x ) E0 cos x e 2 R0 . (18.8)
A
The field at the aperture plane outside the aperture is assumed equal to zero.
The field expression (18.8) is substituted in the integral I yE (see Lecture 17):
I yE = Ea y ( x, y)e j ( x sin cos j + y sin sin j ) dxdy , (18.9)
SA
+ A /2 b + b /2
j x 2
I yE E0 cos x e 2 R0 e j b x sin cos j dx e j b y sin sin j dy . (18.10)
A b /2
((((((((((((
A /2
I ( ,j )
The second integral has been already encountered. The first integral is
cumbersome and the final result only is given below:
bb
sin sin sin
1 R0 2
=
I y E0
E I ( , ) b , (18.11)
b b
sin sin
2 b
2
where
2
R0
sin cos j +
[C ( s2 ) jS ( s2 ) C ( s1 ) + jS ( s1 ) ]
j
I ( ,j ) e
= 2 A
2
(18.12)
R
j 0 sin cos j
+ e 2 A [C (t2 ) jS (t2 ) C (t1 ) + jS (t1 ) ]
and
1 A R0
s1
= R0 u ;
R0 2 A
1 A R
s2
= + R0 u 0 ;
R0 2 A
1 A R0
t1
= R0 u + ;
R0 2 A
1 A R0
t2
= + R0 u + ;
R0 2 A
u = sin cos .
(
I ( ,j ) sin j + cos j . )
H-plane ( = 0 ):
1 + cos
)
FH (= f H (=)
2
(18.19)
1 + cos I ( , =
0)
=
2 I ( =0, =0)
The H-plane pattern in terms of the I ( , ) integral is an approximation, which
is a consequence of the phase approximation made in (18.7). Accurate value for
f H ( ) is found by integrating numerically the field as given in (18.14), i.e.,
+ A /2
x j
A e
R02 + x2
f H ( ) cos e j x sin dx . (18.20)
A /2
4 S Ea ds
=
D0 A
. (18.21)
2 | Ea |2 ds
S A
where
8
t = ;
2
=
H
ph
p2
64t
{[C ( p ) C ( p )] + [ S ( p ) S ( p )] } ;
1 2
2
1 2
2
1 1
=
p1 2 t 1 + , =
p2 2 t 1 + ;
8t 8t
2
1 A 1
t= .
8 R0 /
The factor t explicitly shows the aperture efficiency associated with the
aperture cosine taper. The factor phH is the aperture efficiency associated with
R0 = 100
[Stutzman]
It can be shown that the optimal directivity is obtained if the relation between A
and R0 is
A = 3 R0 , (18.24)
or
A R0
= 3 . (18.25)
lE
R y
R0
b B
aE z
RE
E-plane (y-z) cut of an E-plane
sectoral horn
The geometry of the E-plane sectoral horn in the E-plane (y-z plane) is
analogous to that of the H-plane sectoral horn in the H-plane. The analysis is
following the same steps as in the previous section. The field at the aperture is
approximated by [compare with (18.8)]
j y2
Ea y = E0 cos x e 2 R0 . (18.26)
a
Here, the approximations
y
2
1 y 2
R= R02 + y 2 = R0 1 + R0 1 + (18.27)
R0 2 R0
and
1 y2
R R0 (18.28)
2 R0
are made, which are analogous to (18.6) and (18.7).
Principal-plane patterns
The normalized H-plane pattern is found by substituting = 0 in (18.29):
a
cos sin
1 + cos 2 .
H ( )
= (18.31)
a
2
2
1 sin
2
The second factor in this expression is the pattern of a uniform-phase cosine-
amplitude tapered line source.
The normalized E-plane pattern is found by substituting = 90 in
(18.29):
(1+ cos )
E ( ) = f E ( )
(1+ cos ) [C ( r2 ) C ( r1 )]2 +[ S ( r2 ) S ( r1 )]2 . (18.32)
2 2 4= ) + S 2 ( r 0 )
C 2 ( r 0 =
Here, the arguments of the Fresnel integrals are calculated for = 90 :
B B
=r1 R0 sin ,
R0 2 2
(18.33)
B B
=
r2 + R0 sin ,
R0 2 2
and
B
r ==
0 r2 (= 0)
= . (18.34)
2 R0
Similar to the H-plane sectoral horn, the principal E-plane pattern can be
accurately calculated if no approximation of the phase distribution is made.
Then, the function f E ( ) has to be calculated by numerical integration of
(compare with (18.20))
B /2
R02 + y 2
f E ( ) e j e j sin y dy . (18.35)
B /2
Directivity
The directivity of the E-plane sectoral horn is found in a manner analogous
to the H-plane sectoral horn:
a 32 B E 4p
=DE = ph E aB ,
2 t ph
(18.36)
p
where
8 C 2 (q) + S 2 (q) B
t = 2 , ph =
= E , q .
q2 2 R0
A family of universal directivity curves DE / a vs. B / with R0 being a
parameter is given below.
R0 = 100
The optimal relation between the flared height B and the horn apex length R0
that produces the maximum possible directivity is
B = 2 R0 . (18.37)
where
8
t = ;
2
=
H
ph
p2
64t
{[ C ( p1 ) C ( p2 ) ]2
+ [ S ( p1 ) S ( p2 ) ]2
};
2
1 1 1 A 1
=
p1 2 t 1 + , =
p2 2 t 1 + , t = H ;
8t 8t 8 R0 /
C 2 (q) + S 2 (q) B
=E
ph = , q .
q2 2 R0E
The gain of a horn is usually very close to its directivity because the radiation
efficiency is very good (low losses). The directivity as calculated with (18.39)
is very close to measurements. The above expression is a physical optics
approximation, and it does not take into account only multiple diffractions, and
the diffraction at the edges of the horn arising from reflections from the horn
interior. These phenomena, which are unaccounted for, lead to only very minor
fluctuations of the measured results about the prediction of (18.39). That is why
horns are often used as gain standards in antenna measurements.
The optimal directivity of an E-plane horn is achieved at q = 1 [see also
(18.37)], ph
E = 0.8 . The optimal directivity of an H-plane horn is achieved at
aperture efficiency of
ph
= P =
ph ph
H E 0.632 . (18.40)
The total aperture efficiency includes the taper factor, too:
ph
P = t ph
HE =
ph 0.81 0.632 =0.51. (18.41)
Therefore, the best achievable directivity for a rectangular waveguide horn is
about half that of a uniform rectangular aperture.
We reiterate that best accuracy is achieved if phH and E are calculated
ph
numerically without using the second-order phase approximations in (18.7) and
(18.28).
y x
R0E R0H
b B
E A z
H
RE
RH
B=
1
2
(
b + b 2 + 8 RE . ) (18.46)
Similarly, the maximum-gain condition for the H-plane of (18.24) together with
(18.43) yields
A a A2 ( A a)
=RH = A . (18.47)
A 3 3
Since RE = RH must be fulfilled, (18.47) is substituted in (18.46), which gives
1 8 A( A a )
B= b + b 2 +
. (18.48)
2 3
Substituting in the expression for the horns gain,
4p
G= ap AB , (18.49)
2
gives the relation between A, the gain G, and the aperture efficiency ap :
4p 1 8 A(a a )
=G ap A b + b 2 +
, (18.50)
2 2 3
3bG 2 3G 2 4
A4 aA3 + A =
0. (18.51)
8p ap 32p 2 ap
2
The gain increases with frequency, which is typical for aperture antennas.
However, the curve shows saturation at higher frequencies. This is due to the
decrease of the aperture efficiency, which is a result of an increased phase
difference in the field distribution at the aperture.
Note the multiple side lobes and the significant back lobe. They are due to
diffraction at the horn edges, which are perpendicular to the E field. To reduce
edge diffraction, enhancements are proposed for horn antennas such as
corrugated horns
aperture-matched horns
The corrugated horns achieve tapering of the E field in the vertical direction,
thus, reducing the side-lobes and the diffraction from the top and bottom edges.
The overall main beam becomes smooth and nearly rotationally symmetrical
(esp. for A B ). This is important when the horn is used as a feed to a reflector
antenna.
2 Circular apertures
2.1 A uniform circular aperture
The uniform circular aperture is approximated by a circular opening in a
ground plane illuminated by a uniform plane wave normally incident from
behind.
z
a
E y
0 e
E= dj d 2 E0 J 0 ( sin )d . (18.56)
sin cos(j j )
I xE j
0
0 0
Here, J 0 is the Bessel function of the first kind of order zero. Applying the
identity
Principal-plane patterns
2 J1 ( a sin )
E-plane ( = 0 ): E ( ) = (18.60)
a sin
2 J1 ( a sin )
H-plane ( = 90 ): E=
( ) cos (18.61)
a sin
The larger the aperture, the less significant the cos factor is in (18.61)
because the main beam in the = 0 direction is very narrow and in this small
solid angle cos 1. Thus, the 3-D pattern of a large circular aperture features
a fairly symmetrical beam.
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
-15 -10 -5 0 5 10 15
6*pi*sin(theta)
The half-power angle for the f ( ) factor is obtained at a sin = 1.6 . So,
the HPBW for large apertures ( a ) is given by
1.6 1.6
HPBW 21/2 . 2arcsin=
= a 2 58.4 , deg. (18.63)
a 2a
For example, if the diameter of the aperture is 2a = 10 , then HPBW
= 5.84 .
The side-lobe level of any uniform circular aperture is 0.1332 (-17.5 dB).
Any uniform aperture has unity taper aperture efficiency, and its directivity
can be found directly in terms of its physical area,
4pp 4
=
Du = Ap p a2 . (18.64)
2 2
In (18.65), we still assume that the field has axial symmetry, i.e., it does not
depend on . Often used approximation is the parabolic taper of order n:
n
2
) E0 1
Ea (= (18.66)
a
where E0 is a constant. This is substituted in (18.65) to calculate the respective
component of the radiation integral:
n
2
a
I xE ( ) 2 E0 1 J 0 ( sin )d .
= (18.67)
0 a
The following relation is used to solve (18.67):
1
2n n!
(1 x 2 ) n xJ 0 (bx ) dx =
b n +1
J n +1 (b) . (18.68)
0
is the normalized pattern (neglecting the angular factors such as cos and
cos sin ).
The aperture taper efficiency is calculated to be
Nikolova 2010 1
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
AIRCRAFT RADAR
Nikolova 2010 2
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Nikolova 2010 3
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
0.4
0.3
0.2
0.1
0
rho/D
0 Focal point
-0.1
-0.2
-0.3
-0.4
-0.5
-0.5 -0.4 -0.3 -0.2 -0.1
zf
Nikolova 2010 5
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
The angle from the feed (focal) point to the reflectors rim is related to F / D as
1
0 2arctan . (19.4)
4( F / D )
The focal distance F of a given reflector can be calculated after measuring
its diameter D and its height H 0 :
D2
F . (19.5)
16 H 0
Eq. (19.5) is found by solving (19.1) with D / 2 and z f F H 0 . For
example, if F / D 1 / 4 , then H 0 D / 4 H 0 F , i.e., the focal point is on
the reflectors rim plane.
The reflector design problem involves mainly the matching of the feed
antenna pattern to the reflector. The usual goal is to have the feed pattern at
about a 10 dB level in the direction of the rim, i.e. F f ( 0 ) 10 dB
(0.316 of the normalized amplitude pattern).
The geometry of the paraboloidal reflector has two valuable features:
All rays leaving the focal point O are collimated along the reflectors axis
after reflection.
All overall ray path lengths (from the focal point to the reflector and on to
the aperture plane) are the same and equal to 2F .
The above properties are proven by the GO methods, therefore, they are true
only if the following conditions hold:
The radius of the curvature of the reflector is large compared to the
wavelength and the local region around each reflection point can be
treated as planar.
The radius of the curvature of the incoming wave from the feed is large
and can be treated locally at the reflection point as a plane wave.
The reflector is a perfect conductor, i.e., 1 .
The collimating property of the parabolic reflector is easily established after
finding the unit normal of the parabola,
C p
n . (19.6)
C p
Nikolova 2010 6
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Here,
C p F rf cos 2 f / 2 0 (19.7)
is the parabolic curve equation [see equation (19.2)]. After applying the
operator in spherical coordinates, C p is obtained as
f f f
C p r f cos 2 f cos sin , (19.8)
2 2 2
and, therefore,
f f
n r f cos f sin . (19.9)
2 2
The angles between n and the incident and reflected rays are found below:
f
cos i r f n cos . (19.10)
2
According to Snells law, i r . It is easy to show that this is fulfilled only if
the ray is reflected in the z-direction:
f f
cos r z n (r f cos f f sin f ) r f cos f sin
2 2
(19.11)
f f f
cos f cos sin f sin cos .
2 2 2
Thus, we proved that for any angle of incidence f the reflected wave is z-
directed.
The equal-path-length property follows from (19.2). The total path-length L
for a ray reflected at the point R is
L OR RA rf rf cos f rf (1 cos f ) 2 F . (19.12)
Notice that L is a constant equal to 2F regardless of the angle of incidence.
Nikolova 2010 7
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Nikolova 2010 8
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
dA
d
rf
d f
f
z
GO assumes that power density in free space follows straight paths. Applied
to the power transmitted by the feed, the power in a conical wedge stays
confined within as it progresses along the cones axis. Consider a conical
wedge of solid angle d whose cross-section angle is d f . It confines power,
which after being reflected from the paraboloid, arrives at the aperture plane
confined within a cylindrical ring of thickness d and area dA 2 d .
Let us assume that the feed is isotropic and it has radiation intensity
U t / 4 , where t is the transmitted power. The power confined in the
conical wedge is d Ud ( t / 4 )d . This power reaches the aperture
plane with a density of
d t d
Pa ( )
. (19.13)
dA 4 dA
The generic relation between the solid angle increment and the directional-
angle increments is
d sin d d , (19.14)
(see Lecture 4). In this case, the structure is rotationally symmetric, so we
define the solid angle of the conical wedge as
Nikolova 2010 9
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
2
d (sin f d f )d f 2 sin f d f . (19.15)
0
Equation (19.19) shows the spherical nature of the feed radiation, and it is
referred to as spherical spreading loss. Since Ea Pa ,
1
Ea . (19.20)
rf
If the primary feed is not isotropic, the effect of its normalized field pattern
F f ( f , f ) is easily incorporated in (19.20) as
F f ( f , f )
Ea . (19.21)
rf
Thus, we can conclude that the field at the aperture is described as
F f ( f , f )
Ea ( f , f ) Em e j 2 F . (19.22)
rf
The coordinates ( , ) are more suitable for the description of the aperture
field. Obviously, f . As for r f and f , they are transformed as
4 F 2 2
rf , (19.23)
4F
Nikolova 2010 10
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
f 2arctan
. (19.24)
2F
The last thing to be determined is the polarization of the aperture field
provided the polarization of the primary-feed field is known (denoted with u i ).
The law of reflection at a perfectly conducting wall states that n bisects the
incident and the reflected rays, and that the total electric field has zero
tangential component at the surface, i.e.,
Ei Er 0 , (19.25)
and
Er Ei 2(n Ei )n Er 2(n Ei )n Ei . (19.26)
Ei Er
n
However, more sophisticated designs often use an offset feed. In such cases, the
PO method (i.e., the current distribution method) is preferred.
Equation (19.29) follows directly from the relation between the far-zone
electric field and the magnetic vector potential A,
E far j A , (19.30)
which can written more formally as
E far j A ( j A r )r j ( A A ) . (19.31)
This approach is also known as Ruschs method after the name of the person
who first introduced it. The integral in (19.29) is usually evaluated numerically
by computer codes in order to render the approach versatile with respect to any
aperture and any aperture field distribution.
In conclusion, we note that both the GO and the PO methods produce very
accurate results for the main beam and first side lobe. The pattern far out the
main beam can be accurately predicted by including diffraction effects
(scattering) from the reflectors rim. This is done by augmenting GO with the
use of geometrical theory of diffraction (GTD) (J.B. Keller, 1962), or by
augmenting the PO method with the physical theory of diffraction (PTD) (P.I.
Ufimtsev, 1957).
Nikolova 2010 12
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Nikolova 2010 14
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
y H-plane
E-plane
Nikolova 2010 15
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
CO-POLARIZATION
Nikolova 2010 16
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Nikolova 2010 17
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
CROSS-POLARIZATION
45 ,135
The results above are obtained using commercial software (GRASP) using PO
methods (surface current integration).
Cross-polarization of reflectors is measured as the ratio of the peak cross-
Nikolova 2010 18
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
The analysis techniques given in the previous sections are general and can
be applied to these reflectors, too. Generally, the PO method (surface currents
integration) is believed to yield better accuracy. Both, the PO and the GO
methods, are accurate only at the main beam and the first couple of side-lobes.
Offset reflectors are popular for antenna systems producing contour beams.
To obtain such beams, multiple primary feeds (usually horns) are needed to
illuminate the reflector at different angles. Such multiple-antenna feeds may
constitute a significant obstacle at the antenna aperture and offset reflectors are
indeed necessary.
Nikolova 2010 19
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
7. Dual-reflector antennas
The dual-reflector antenna consists of two reflectors and a feed antenna. The
feed is conveniently located at the apex of the main reflector. This makes the
system mechanically robust, the transmission lines are shorter and easier to
construct (especially in the case of waveguides).
The virtual focal point F is the point from which transmitted rays appear to
emanate with a spherical wave front after reflection from the subreflector.
The most popular dual reflector is the axisymmetric Cassegrain antenna.
The main reflector is parabolic and the subreflector is hyperbolic (convex).
A second form of the dual reflector is the Gregorian reflector. It has a
concave elliptic subreflector. The Gregorian subreflector is more distant from
the main reflector and, thus, it requires more support.
Dual-reflector antennas for earth terminals have another important
advantage beside the location of the main feed. They have almost no spillover
toward the noisy ground, as do the single-feed reflector antennas. Their
spillover (if any) is directed toward the much less noisy sky region. Both, the
Cassegrain and the Gregorian reflector systems have their origins in optical
telescopes and are named after their inventors.
Nikolova 2010 20
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
The subreflector is defined by its diameter Ds and its eccentricity e . The shape
(or curvature) is controlled by the eccentricity:
c 1, hyperbola
e (19.41)
a < 1, ellipse
Special cases are
e , straight line (plane)
e 0 , circle (sphere)
e 1, parabola
Both, the ellipse and the hyperbola, are described by the equation
zs2 xs2
1. (19.42)
a2 c2 a2
The function of a hyperbolic subreflector is to convert the incoming wave
from a feed antenna located at the focal point F to a spherical wave front w
that appears to originate from the virtual focal point F. This means that the
optical path from F to w must be constant with respect to the angle of
Nikolova 2010 21
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
incidence:
F R RA F V VB c a VB . (19.43)
Since
RA FA FR FB FR , (19.44)
( FA FB because the reflected wave must be spherical)
F R FR c a ( FB VB) c a (c a) 2a . (19.45)
Note: Another definition of a hyperbola is: a hyperbola is the locus of a point
that moves so that the difference of the distances from its two focal points,
F R FR , is equal to a constant, 2a .
The equivalent parabola has the same diameter, De D , but its focal length is
longer than that of the main reflector:
Nikolova 2010 22
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
e 1
Fe F M F . (19.46)
e 1
Here, M (e 1) / (e 1) is called magnification.
The increased equivalent focal length has several advantages:
less cross-polarization;
less spherical-spread loss at the reflectors rim, and therefore, improved
aperture efficiency.
The synthesis of dual-reflector systems is an advanced topic. Many factors
are taken into account when shaped reflectors are designed for improved
aperture efficiency. These are: minimized spillover, less phase error, improved
amplitude distribution in the reflectors aperture.
Nikolova 2010 23
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
2
S Ea ds
4
D0 2 . A
(19.50)
| Ea |2 ds
S A
S Ea ds
Aeff A
(19.51)
S | Ea |2 ds
A
Aeff 1 S Ea ds
t A
. (19.52)
Ap Ap S | Ea | ds
A
2
Expression (19.52) can be written directly in terms of the known feed antenna
pattern. If the aperture is circular, then
2 a 2
1
Ea ( , ) d d
t 0 0
2 a
. (19.53)
a2
| Ea ( , ) |2 d d
0 0
All that is needed to calculate the taper efficiency is the feed pattern F f ( f , ) .
If the feed pattern extends beyond the reflectors rim, certain amount of
power is not redirected by the reflector, i.e., it is lost. This power-loss is
referred to as spillover. The spillover efficiency measures that portion of the
feed pattern, which is intercepted by the reflector relative to the total feed
power:
Nikolova 2010 24
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
2 0
| F f ( f , ) |2 sin f d f d
s 0 0
2
. (19.55)
| F f ( f , ) |2 sin f d f d
0 0
The reflector design problem includes a trade-off between aperture taper and
spillover when the feed antenna is chosen. Taper and spillover efficiencies are
combined to form the so-called illumination efficiency i t s . Multiplying
(19.54) and (19.55), and using a 2 F tan( 0 / 2) yields
2
2 o
Df f
i 2 cot 2 0
4 2 F f ( f , ) tan 2
d f d . (19.56)
0 0
Here,
4
Df 2
, (19.57)
| F f ( f , ) |2 sin f d f d
0 0
is the directivity of the feed antenna. An ideal feed antenna pattern would
compensate for the spherical spreading loss by increasing the field strength as
f increases, and then would abruptly fall to zero in the direction of the
reflectors rim in order to avoid spillover:
cos 2 ( o / 2)
, f o
F f ( f , ) cos 2 ( f / 2) (19.58)
0, f o
This ideal feed is not realizable. For practical purposes, (19.56) has to be
optimized with respect to the edge-illumination level. The function specified by
(19.56) is well-behaved with a single maximum with respect to the edge-
illumination.
The achievement efficiency a is an integral factor including losses due to:
random surface error, cross-polarization loss, aperture blockage, reflector phase
error (profile accuracy), feed phase error.
A well-designed and well-made aperture antenna should have an overall
Nikolova 2010 25
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
1
D.K. Cheng, Effects of arbitrary phase errors on the gain and beamwidth characteristics of radiation pattern, IRE Trans. AP,
vol. AP-3, No. 3, pp. 145-147, July 1955.
Nikolova 2010 26
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Achieving optimum performance from a microwave dish antenna requires that the feed
antenna be matched to the parabolic reflector. Traditionally, we have relied on rules of
thumb to choose a feed for the dish. Alternatively, a computer program can be used to
analyze performance of a feed antenna based on measured or calculated radiation patterns
and, more importantly, present the data in a graphical format for easy comprehension.
This tool is used to explore a number of published feed designs in an effort to enhance
understanding of the performance of dish antennas and feeds.
Parabolic dish antenna fundamentals were covered in detail in Chapter 4, but a short
review of key points is in order here. Figure 11-1 illustrates the operation of the dish
antenna: a feed antenna at the focus of the parabola illuminates, or radiates energy
toward, the reflector, which reflects it into a narrow beam of energy. Part of the feed
antenna radiation misses the reflector; this loss is called spillover. Another part of the
feed energy is reflected back into the feed antenna and doesnt become part of the main
beam; this loss is referred to as feed blockage.
Ideally, all areas of the reflector should be illuminated with equal energy from the feed.
Figure 11-2 shows this desired feed pattern as a broken line; since the edges of the
parabolic curve are farther away from the focus than the center of the curve, more energy
is required at the edges than at the center, but with no energy missing the reflector. An
additional requirement is that all the feed energy be in phase, so that it appears to be
radiated from a single point at the focus. The desired radiation pattern cannot be realized
with real feed antennas, so perfectly uniform illumination cannot be achieved. Figure 11-2
also shows an idealized typical feed antenna pattern; the difference between the desired
feed pattern and the actual feed radiation pattern results in illumination loss because
some areas of the reflector are unable to work as effectively as others, as well as the
spillover loss of the energy that misses the reflector and continues in an undesired
direction.
For each reflector, we try to choose a feed that provides a compromise of illumination loss
and spillover loss which yields maximum performance, which we measure by aperture
efficiency, a comparison of the actual gain to the maximum theoretical gain achievable for
the same aperture area. The traditional rule of thumb for this compromise is that best
efficiency occurs when the illumination energy is 10 dB down at the edge of the dish, so
the feed should be designed for a radiation pattern which is 10 dB down at the edge of the
dish. It isnt necessary to do this for each individual dish; all dishes with the same f/D, the
Spillover
Feed
Focus
Blockage
Figure 11-1
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Illumination loss
Spillover loss
0 dB -10 dB -20 dB
f/D = 0.4
Illumination taper = 10 dB
ratio of focal length to diameter, have the same geometry regardless of reflector diameter.
Thus, all dishes with the same f/D can use the same feed design, and good feed designs are
available for several common values of f/D.
Efficiency calculation
The aperture efficiency of a dish antenna is the amount energy concentrated into the beam
divided by the total energy radiated by the feed. The efficiency can calculated by
integrating (remember calculus?) the feed pattern radiated over the area of the reflector
and dividing the result by the total integrated feed pattern. When this calculation was
done by hand, it was usually done by approximating the feed pattern with an idealized cosn
feed pattern (n = 3 in this example) as shown in Figure 11-3, making the integration much
easier. With a computer, we can do numerical integration of actual feed patterns,
performing the tedious calculations for many data points.
The numerical integration routine I used is borrowed from a BASIC program by W7PUA1
which is based on a 1947 paper by Cutler 2. I translated the routine to C++, then added
some enhancements:
the data interpolation is more flexible to use whatever feed pattern data is
available,
feed blockage loss is calculated,
and the output is graphical for visual comprehension.
I find that a simple curve is easier to understand than tables of numbers or long
descriptions. The output format is PostScript, which can be displayed or printed using
the free Ghostscript software 3.
The bottom half of Figure 11-3 is an example of the graphical output, a plot of efficiency
vs. f/D for the cosn feed pattern shown as a polar plot in the top half of Figure 11-3. It is
obvious at a glance that this feed pattern is best suited for a reflector with an f/D of 0.4 to
0.5. The calculated efficiency of 80% for this idealized feed pattern provides a benchmark
against which real feed antennas may be compared.
The efficiency curve in Figure 11-3 shows decreasing efficiency for f/D less than 0.4.
These are deep dishes, requiring the feed to provide illumination over a very wide angle.
For a dish with an f/D = 0.25, the focus is level with the rim of the dish, so that the feed
must provide illumination over 180 degrees, as shown in Figure 11-4a. The polar plot of
feed radiation in Figure 11-3 shows almost no energy radiated straight up and down,
toward the edges of this dish. So it is not surprising that the illumination loss increases for
small values of f/D.
The other end of the efficiency curve, for f/D greater than 0.5, also shows a decreasing
efficiency. These shallower dishes, like the one illustrated in Figure 11-4b, require a
narrower angle of illumination, so more of the energy from the feed in Figure 11-3 misses
the dish, and spillover loss increases as the f/D increases.
H-plane
N1BWT 1997 Dish diameter = 10
Feed diameter = 1
70
2 dB
Efficiency %
60
50 3 dB
40 4 dB
5 dB
30
6 dB
20 7 dB
8 dB
10
Illumination loss
Spillover loss
0 dB -10 -20
Illumination loss
Spillover loss
0 dB -10 -20
Illumination taper = 10 dB
Figure 11-4b: Dish Illumination f/D = 0.75
Now lets look at an example of a real antenna. There are many big TVRO dishes around
which are being replaced by the small DSS systems and becoming available. A 12-foot
TVRO dish might be usable for EME operation on 1296 MHz. The dishes typically have
an f/D around 0.35 to 0.45, so we would like to find a suitable feedhorn. Popular
feedhorn designs for 1296 MHz have been described by W2IMU 4 and VE4MA5, so lets
look at graphs of the published patterns for these two feeds. The W2IMU dual-mode
feedhorn in Figure 11-5 provides good efficiency at an f/D around 0.6, but is not very
good around our target of 0.35 to 0.45. On the other hand, the VE4MA feedhorn in
Figure 11-6 provides its best efficiency at an f/D around 0.4, so it is a much better choice
for a TVRO dish. The maximum calculated efficiency is just a bit lower than the 80% for
the idealized feed in Figure 11-3. Later we will compare it with other real feeds.
The feed blockage loss shown in Figure 11-6 is under 10% for a twelve foot dish, where
the dish diameter is 8.6 times larger than the feedhorn diameter. Since the feed diameter
does not change when it is used on larger or smaller dishes, lets look at a few more
examples and see what happens. Figure 11-7 shows efficiency curves for the VE4MA
feed on a range of dish sizes. At the top is the curve for a 28 foot dish; since the reflector
diameter is 20 times as large as the feed diameter, feed blockage is small and efficiency is
high. The next curve, for an 8 foot dish, 5.7 times larger than the reflector, shows
efficiency a bit lower than the 12 and 28 foot dishes. Going to smaller dishes makes the
efficiency much lower: the 4 foot dish, 2.9 times larger than the feed, has significantly
reduced efficiency, while the 2 foot dish, only 1.4 times larger than the feed, hardly works
at all. Of course, a 2 foot dish at 1296 MHz is a pretty small antenna, but with efficiency
this low the gain would be perhaps 11 dB, not much higher than the feedhorn alone. A
more important point is that any blockage, whether by the feed or by the structure
supporting it, reduces the efficiency of the dish.
Feed blockage is more significant on small dishes, but small is a relative term; any dish
with a diameter less than 10 can be considered small. Thus, a 2 foot dish at 10 GHz,
about 20 in diameter, is a moderately large dish, while a 20 foot dish at 432 MHz, less
than 10 , is a small dish.
Occasionally a surplus dish is found with the original feed attached. One example I have
seen is a dish fed with WR-90 waveguide, which covers X-band (8-12 GHz). This dish is
fed by from the open end of the waveguide pointing at the dish; an open waveguide is
known to act as a moderate gain antenna. I located a published radiation pattern for open
WR-90 waveguide6 and graphed it in Figure 11-8. Clearly the efficiency this simple feed
provides is far lower than the previous ones. The moral of this story is that just because a
E-plane
N1BWT 1997 Dish diameter = 15.8
Feed diameter = 1.31
70
2 dB
Efficiency %
60
50 3 dB
40 4 dB
5 dB
30
6 dB
20 7 dB
8 dB
10
H-plane
N1BWT 1997 Dish diameter = 15.8
Feed diameter = 1.84
70
2 dB
Efficiency %
60
50 3 dB
40 4 dB
5 dB
30
6 dB
20 7 dB
8 dB
10
H-plane
N1BWT 1997 Dish diameter as shown
Feed diameter = 1.84
70
2 dB
Efficiency %
60
4 feet
50 3 dB
40 4 dB
5 dB
30
2 feet 6 dB
20 7 dB
8 dB
10
dish already has a feed does not mean it is a good feed the original design goal may not
have been maximum gain.
The purpose of these graphs is to help in visualizing the performance of various dish feeds
and comparing them so that the best feed available may be chosen for each application.
The underlying assumption is that we wish to obtain the maximum efficiency from a given
dish, and thus the maximum gain. After all, a dish doesnt get any lighter or have any less
wind resistance if we get less than maximum gain from it. On the right side of each graph
is a dB scale, relating the efficiency to a loss in dB from the theoretical gain for that
aperture. The ripple in some of the curves is an artifact of the discrete points used in the
numerical integration process, and could be removed by integrating at smaller intervals.
The program only accounts for the losses that are unavoidable: illumination loss, spillover
loss, and feed blockage loss. There are several other losses found in a real dish:
phase error
feed not at focus
diffraction from the edge of the dish
polarization shift due to reflector geometry
blockage by feed supports
surface error in the parabolic reflector
feedline loss
feed VSWR
These losses occur in greater or lesser amounts in a given antenna, so that the real
efficiency is lower than the maximum possible efficiency shown in the curves. The best
antennas I have measured have efficiencies perhaps 15% lower than the curves, while
others are significantly worse. A typical efficiency for a moderate-sized dish is about
50%, for a gain 3 dB below the theoretical gain for a given aperture size. A really good
dish has an efficiency of 60% or so, about 1 dB better than a typical dish, while a poorly
chosen feed or a poor installation can make the gain several dB worse. One dB difference
may not seem like much, but it is a huge difference for an EME station that cant squeeze
another dB from the preamp or power amplifier.
Another limitation of these curves is the accuracy of the data available for each feed
pattern. Many of the published articles only give data for the major part of the pattern,
but not the backlobes. This is sufficient to calculate the shape of the efficiency curve, but
the whole pattern is required to calculate the maximum efficiency, so I have estimated the
rest of the pattern based on similar feeds where data for the whole pattern is available.
Since some of the feeds are physically larger than others and would have more feed
blockage loss on a given dish, comparisons on any fixed dish size would make smaller
feeds look better. Therefore, most of the graphs for the various feeds described in
Chapter 6 use a reflector diameter about ten times larger than the feed diameter. If you
H-plane
N1BWT 1997 Dish diameter = 10
Feed diameter = 1
70
2 dB
Efficiency %
60
50 3 dB
40 4 dB
5 dB
30
6 dB
20 7 dB
8 dB
10
are comparing these feeds for use on your dish, you can run the program for the actual
reflector diameter of your dish.
Phase errors
All of these graphs are based on amplitude patterns only for the feeds, because phase data
is much more difficult to measure and is rarely available. If the phase of the radiated
energy is not uniform over all areas of the reflector, then different parts may reflect energy
into the main beam which is not in phase and reduces the total energy in the main beam,
lowering the gain. Another common problem is feeds that do not have the same phase
center in the E-plane and the H-plane, which has the same effect as not having the phase
center at the focal point: reduced gain and pattern distortion.
Phase errors are probably the largest cause for low efficiency, so you should not expect to
get efficiencies near the calculated values unless the feed has good phase performance. A
feed with small phase error still suffers from all the other losses listed above, so the
expected performance of a real dish might be only 15% lower than the calculated
efficiency curve.
The only feeds which have published phase data are the Kumar (VE4MA), Chaparral,
Chaparral with slots, and the Koch Multi-ring feed discussed, all described in Chapter 6.
All of these have excellent phase performance over a wide illumination angle, so the
efficiency curves for these feeds are good for any f/D greater than 0.3. None of these
feeds can adequately illuminate an f/D of 0.25, but the bent-dipole Handlebar feeds of
W7PUA 1 show promise at the lower frequencies.
Perhaps we should take a lesson from the radio astronomers. The radiotelescope feeds
described in Chapter 6 all operate at a point on the efficiency curve to the left of the peak,
or lower f/D, for reduced spillover. W7PUA suggests that since spillover increases
sidelobes and sidelobes are always bad, we should make any compromise to the left side of
the peak.
Computer program
The FEEDPATT program does all the calculations and plots graphs like the ones above.
For those with access to the Internet, the FEEDPATT program and all the data files for
feed patterns are available on my 10 GHz Web page:
www.tiac.net/users/wade/feedpatt.zip or www.qsl.net/~n1bwt
The output graphs from the program are files in PostScript format, ready for printing on
a laser printer or for viewing and printing on a PC using the free Ghostscript 3 software.
On my PC running Windows95 or WindowsNT, I run the FEEDPATT program in
one window and Ghostscript gsview in another to view the output as I work.
With the data files for the feeds described in Chapter 6, it should be possible to graph the
potential performance of any of them on your dish for any frequency of interest. For other
feeds, if you can find, calculate, or measure a radiation pattern, you can calculate a graph
of estimated efficiency. Please send me a copy of any new feed data.
Summary
Using the FEEDPATT computer program, we can accurately analyze the pattern data for
various feed antennas. The output is in graphical format for easy visual comparison. We
used this tool in Chapter 6 to analyze a wide range of feed designs and followed the
evolution from early WWII-vintage designs to modern high-performance feeds. The
program and feed pattern data files are available to help further understanding of dish
performance.
Typical Applications
Microstrip Antenna Integrated into a System: HIC Antenna Base-Station for 28-43 GHz
L x
h
x
L
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Hence E z Ez ( x, y )
E z ( x, y )
1
H = E
j
( zE z ( x, y ) )
1
=
j
=
1
j
( z Ez ( x, y ) )
Hence
H ( x, y ) =
j
(
1
z E ( x, y ) )
z
H ( x, y ) =
1
j
( z E ( x, y ) )
z
E z ( x, y )
h
H ( x, y )
J s n = 0 Js
W n
Also, on lower surface of t
patch conductor we have
x
J s = ( z H ) L
Js
n
Hence, h
Ht = 0
x
H t = 0 (PMC) L
n
h
PMC
H ( x, y ) =
1
j
( z E ( x, y ) )
z
Js
W t n
Hence,
x
n ( z Ez ( x, y ) ) = 0 L
z ( n Ez ( x, y ) ) = 0
n
h
Ez
=0
n PMC
Resonance Frequencies
y
Ez + k Ez = 0
2 2
m 2 n 2
+ k Ez = 0
2
L W
m 2 n 2
+k =0
2
Hence
L W
k =
2
+
L W (x0, y0)
W
Recall that
k = 0 0 r x
L
= 2 f
Hence
m n
2 2
c
f = + c = 1/ 0 0
2 r L W
Hence f = f mn
(x0, y0)
W
(resonance frequency of
(m, n) mode)
x
L
m n
2 2
c
f mn = +
2 r L W
(1,0) Mode
y
current
This mode is usually used because the
radiation pattern has a broadside beam.
W
x
=
Ez cos
L x
L
c1
f10 =
2 r L This mode acts as a wide
microstrip line (width W)
that has a resonant length
1 x of 0.5 guided wavelengths
J s = x sin
j 0 L L in the x direction.
Approximately,
Note: this is equivalent to saying that
c 1 the length L is one-half of a
f10 = wavelength in the dielectric:
r
2 L
0 / 2
kL = L = d / 2 =
r
Note: a higher substrate permittivity allows for a smaller
antenna (miniaturization) but lower bandwidth.
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
y
Le = L + 2L
L L
c 1
f10 = L
2 r Le x
Le
eff W
( r + 0.3) h + 0.264
L / h = 0.412
eff 0.258 W + 0.8
( r ) h
1/ 2
r +1 r 1 h
eff
r = + 1 + 12
2 2 W
Note: L 0.5 h
1
NORMALIZED FREQUENCY Hammerstad
Measured
0.95
0.9
0.85
0.8
0.75
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
h / 0
W = 1.5 L is typical.
30
25
r = 10.8
BANDWIDTH (%)
20
15
10
5 2.2
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
h / 0
The discrete data points are measured values. The solid curves are
from a CAD formula.
r = 2.2 or 10.8 W/ L = 1.5
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
(x0, y0)
W
L L
(x0, y0)
W
x
L
Rin Ez2 ( x0 , y0 )
y
(x0, y0)
For (1,0) mode:
W
x0
Rin cos 2
L x
L
150 formula.
r = 10.8
100
2.2
50
y
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
h / 0 (x0, y0)
W
r = 2.2 or 10.8
W/L = 1.5 x0 = L/4, y0 = W/2 x
L
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Pr
er =
Ptot
TM0
surface wave
x
cos () pattern
Hence,
Pr Pr
er = =
Ptot Pr + ( Pc + Pd + Psw )
100
2.2
80
EFFICIENCY (%)
60
10.8
40
exact
CAD
20
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
h / 0
r = 2.2 or 10.8 W/L = 1.5 Note: CAD plot uses Pozar formulas
100
2.2
80
EFFICIENCY (%)
exact
60
CAD
r = 10.8
40
20
0
0 0.02 0.04 0.06 0.08 0.1
h / 0
r = 2.2 or 10.8 W/L = 1.5 Note: CAD plot uses Pozar formulas
30 -30
-10
60 -20 -60
-30
120 240
150 210
180
45 -10 -45
-20
-30
135 225
180
Results: Directivity
10
r = 2.2
8
DIRECTIVITY (dB)
10.8
6
4
exact
CAD
2
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
h / 0
L
Lp
Zin R C
L
Lp
R C
L
Lp
R C
80
70
60
CAD
50 exact
Rin ( )
40
frequency where the
30 input resistance is
20
maximum
10
0
4 4.5 5 5.5 6
FREQUENCY (GHz)
60
CAD
exact
40
Xin ( )
20
-40
4 4.5 5 5.5 6
FREQUENCY (GHz)
frequency where the
input impedance is real
0 2
Xf = ( k0 h ) + ln
2 r ( k 0 a )
X f = Lp
0 = 0 / 0 = 376.73
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
0 2
Xf = ( k0 h) + ln
2 ( k
r 0 a )
40
35 r = 2.2
CAD
exact
30
W/L = 1.5
25
Xf ( )
20 h = 0.0254 0
15
a = 0.5 mm
10
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Xr
CAD Formulas
where
1
3
1 3
= 2 ( k0 h ) 60 c1 1
3
Pswhed
0 r
Hence we have
1
ehed
r = 3
3 1 1
1 + ( k0 h ) 1
4 c1 r
1 2/5
c1 = 1 +
r r2
a2 3 1
p = 1 + ( k 0 W ) + ( a2 + 2 a 4 ) ( k0 W ) + c2 ( k0 L )
2 2 4 2
10 560 5
1
+ a2 c2 ( k0 W ) ( k0 L )
2 2
70
c2 = 0.0914153
a2 = 0.16605
a4 = 0.00761
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
1
e hed
=
2 ( 0 ) (
c1 )
r 1
Pswhed
2
P hed
= k h 80 2
1 + hed sp
0
Psp
r ( x 1)
3/ 2
k 2 2
=
hed 0
P 0 0
4 r (1 + x1 ) + (k0 h) x02 1 (1 + r2 x1 )
sw
x02 1 r2 + 01 + r r2 2 01 + 02
x1 = x0 = 1 +
r x02 r2 12
0 = s tan ( k0 h ) s
1
1 = tan ( k0 h ) s +
( k0 h ) s
s cos ( k0 h ) s
2
s = r 1
1 Rs 1 16 p c1 h W 1
BW = + +
d
0 h / 0 3 r 0 L hed
er
2
f 2 f1
BW = (multiply by 100 if you want to get %)
f0
x0
R = Redge cos
2
L
4 L h
(0 )
W
0
Redge =
Rs 1 16 p c1 W h 1
d + + hed
0 h / 0 3
r L
0 er
3 r
D= ( ( k1h ) )
2
tanc
pc1 r + tan ( k1h )
2
where
tanc ( x ) tan ( x ) / x
3
D
p c1
x W E-plane
J s = x cos x
L
The probe is on the x axis.
L
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
ky W kx L
sin cos
WL 2 2
Ei (r , , ) = Ei ( r , , )
hex
2 ky W 2 k L 2
x
i = or 2 2 2
k x = k0 sin cos
k y = k0 sin sin
j 0 jk0 r
where E0 = e
4 r
2 tan ( k0 h N ( ) )
F ( ) = 1 + TE
( ) =
tan ( k0 h N ( ) ) j N ( ) sec
2 tan ( k0 h N ( ) ) cos
G ( ) = cos (1 + TM
( ) ) = r
tan ( k0 h N ( ) ) j cos
N ( )
N ( ) = r sin 2 ( )
Circular Polarization
y y
L L
L x L x
y
Phase shift realized with delay line L
P
L
P+g/4 LHCP
Z0 / 2 Z0
L
Z0
feed
g/4 Z0
50 Ohm load L
g/4
x LHCP
-180o
-90o
-270o
0o
Circular Patch
a
x
Ez
=0 J m ( ka ) = 0
=a
ka = xmn
a PMC
(nth root of Jm Bessel function)
c
f mn =
xmn
2 r
c
f11 = x11 x11 1.842
2 a r
Long/Shen Formula :
2h a h a
ae = a 1 + ln 2h + 1.7726 a = +
r 2h
or ln 1.7726
a r
In patch cavity:
J1 (k ) 1
E z ( ,) = cos
J1 (ka ) h
(The edge voltage has a maximum of one volt.)
where
tanc ( x ) = tan ( x ) / x
2 jN ( )
P ( ) = cos (1 ( )) = cos
TE
tan ( k 0 hN ( )) jN ( ) sec
2 j r cos
N ( )
Q ( ) = 1 ( ) =
TM
tan (k0 h N ( )) j r cos
N ( )
N ( ) = r sin 2 ( )
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
J12 ( k 0 )
Rin Redge 2
J1 ( ka )
Q ( ) J1 ( k0 a sin ) + P ( ) J inc
2 2 2 2
( k0 a sin ) sin d
J inc ( x ) = J1 ( x ) / x
CAD Formula:
Psp = ( k0 a ) 2 I c
80
6 e0 = 1
pc = ( k0 a ) e2 k
4 2k
I c = pc e2 = 0.400000
3 k =0
e4 = 0.0785710
e6 = 7.27509 103
e8 = 3.81786 104
e10 = 1.09839 105
e12 = 1.47731 107
Feeding Methods
Advantages:
simple
easy to obtain input match
x0
R = Redge cos
2
L
Disadvantages:
difficult to obtain input match for thicker substrates,
due to probe inductance.
significant probe radiation for thicker substrates
Advantages:
simple
allows for planar feeding
easy to obtain input match
Disadvantages:
significant line radiation for thicker substrates
for deep notches, pattern may shown distortion.
Advantages:
allows for planar feeding
less line radiation compared
to microstrip feed patch
microstrip line
Disadvantages:
requires multilayer fabrication
alignment is important for input match
Disadvantages: slot
requires multilayer fabrication
microstrip line
alignment is important for input match
Improving Bandwidth
L-shaped probe:
foam
microstrip
substrate
microstrip line slot
microstrip line
slot
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
-5
-10
-30
-35
-40
3 4 5 6 7 8 9 10 11 12
Frequency (GHz)
stacked patch with ACP feed
Single Layer Single Patch Wideband Microstrip Antenna, T. Huynh and K. F. Lee,
Electronics Letters, Vol. 31, No. 16, pp. 1310-1312, 1986.
Multi-Band Antennas
General Principle:
low-band
low-band
high-band
Miniaturization
High Permittivity
Quarter-Wave Patch
PIFA
Capacitive Loading
Slots
Meandering
r = 1
r = 4
W E-plane W=W/2
L=L/2
L
short-circuit
Ez = 0 vias
W E-plane W E-plane
L L=L/2
side view
side view
The capacitive loading allows for the length of the PIFA to be reduced.
top view
0o 90o
linear CP
The slot forces the current to flow through a longer path, increasing
the effective dimensions of the patch.
Miniaturization: Meandering
via
via
feed feed
Improving Performance:
Reducing Surface-Wave Excitation and
Lateral Radiation
z feed
b
shorted annular ring 0
o
a
x a b
h
30 -30 30 -30
-10 -10
-30 -30
180 180
conventional
conventional RSW
RSW
space-wave radiation
lateral radiation
surface waves
0
RSW - Measured
-10
RSW - Theory
-20 Conv - Measured
-30 Conv - Theory
-40
S12 [dB]
-50
-60
-70
-80
-90
-100
0 1 2 3 4 5 6 7 8 9 10
Separation [Wavelengths]
UNIT III
1 Introduction Antenna Arrays
Antenna arrays are becoming increasingly important in
wireless communications. Advantages of using antenna
arrays:
1. They can provide the capability of a steerable beam
(radiation direction change) as in smart antennas.
2. They can provide a high gain (array gain) by using
simple antenna elements.
3. They provide a diversity gain in multipath signal
reception.
4. They enable array signal processing.
1
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
z
Far field
observation
r1 point
Dipole 2
I 2 Ie j
d r1 r d cos , 0
r
x
Dipole 1 I1 I
E2 j sin j cos I 2
4 r1 2 4 r1
Use the following far-field approximations: 0
1 1
r1 r
e jkr1 e jk r d cos
4
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
E E1 E2 a
kd e jkr jkd cos
a j cos
I I e
4 r
1 2
where
5
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
1
1
cos kd cos
j jkd cos
j kd cos
1 e e 2e 2
2
The magnitude of the total E field is:
1
I1
I1 I 2e jkd cos
kId e
jkr
E a j cos AF
4 r
radiation pattern of a single Hertzian dipole AF
6
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
7
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
1 1 1 1
AFn 2cos kd cos 2cos cos
2 2 2
8
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
1 1 1 1
AFn 2cos kd cos 2cos cos
2 2 2 2
9
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
11
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
|AFn( )|
= kd cos +
kd
kdcos
16
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
2nd grating lobe 1st grating lobe Main lobe 1st grating lobe 2nd grating lobe
|AFn( )|
Visible region
(1) When d 0.5, no General conditions to
grating lobes can be kd avoid grating lobes
formed for whatever with [0,2] and d
value of . (2) When d [0.5,]:
, grating lobe(s) is 1.For 0 < , the
(are) formed for =kdcos requirement is:
whatever value of . kd + 2
(3) When 0.5 <d< , 2. For < 2, the
formation of grating requirement is:
lobes depends on . kd - 0
17
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
19
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Example 1
A uniform linear array consists of 10 half-wave dipoles with
an inter-element separation d = /4 and equal current
amplitude. Find the excitation current phase difference
such that the main beam direction is at 60 (max = 60).
Solutions d = /4, max = 60, N = 10
main beam dirction max 60 cos 1
m 2
2 d
2
m2 cos 60 0.5
m2 45 360 315, when m 1
4
20
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
sin 5 cos
1 2
AFn
1
sin cos
2 2
10
21
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
x
Dipole 1
d Dipole 2
24
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Antenna 1 Antenna 2
I1 I2
Vs1 Vs2 a2
a1
Zg1 b1 Zg2 b2
Terminal
current
d
Excitation Coupled
voltage voltage
source I1 a I2 a2
1
25
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Voc12 a c
b I2 d
26
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
I1
a Voc21 c Note that for
b d passive antennas,
Z12 = Z21
27
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
28
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Is1 and Is2 are the terminal currents at the antennas when
there is no mutual coupling effect.
Vs1 Vs 2
I s1 I s2
Z g1 Z11 Z g 2 Z11
29
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
That is:
1 1
I1 I s1 Z12 I s 2 I 2 I s 2 Z 21
I s1
D D
30
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
where Z12 Z 21
D 1
Z11 Z g1 Z 22 Z g 2
Z12
Z12
Z11 Z g1
Z 21
Z 21
Z 22 Z g 2
Now if we want to find the array pattern E on the
horizontal plane (=/2) with mutual coupling effect, then
E is just equal to the array factor (see pages 10 and 6).
1 jkd cos
Vector
magnitude, not
E =AF I1 I 2e
absolute value
I1
31
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
1
E I1 I 2e jkd cos
I1
1
I s1 Z12 I s 2 I s 2 Z 21
I s1 e jkd cos
I1D
1
I s1 I s 2e jkd cos Z12 I s 2 I s1e jkd cos with Z12 Z 21
I1D
I s1 Is2 j
1 e e
j jkd cos
Z12 e e
j jkd cos
with e
I1D I s1
I s1
1 e
I1D
j kd cos
j
Z12e 1 e j kd cos
original pattern additional pattern
32
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Z12 e j 1 e j kd cos
33
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
34
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Solution
As the required array pattern |En| is on
j
I s1 1, Is2 e the horizontal plane, it is equal to the
normalized array factor |AFn|.
I s1 I s 2 1
I s1 0, I s 2 150 2.62 rad
2
kd
4 2
Z12 Z 21
Z12 0.16 j 0.26
Z11 Z g 1 Z 22 Z g 2
Z12 Z 21
D 1 1.042 j 0.09
Z11 Z g1 Z 22 Z g 2
35
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
E AF
I s1
I1D
1 e j kd cos Z12 e j 1 e j kd cos
0.95 j 0.08
1 e j 2.62e j 2 cos
I1
0.16 j 0.26 e j 2.62 1 e j 2.62e j 2 cos
0.94 j 0.37
1 1.14 j 0.40 e j 2 cos
I1
36
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Normalization
The pattern of f attains the maximum value when = 180.
When = 180,
E 180
0.94 j 0.37
1 1.14 j 0.40 e j 2 cos
I1 180
1.83
I1
38
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
0.94 j 0.37
1 1.14 j 0.40 e j 2 cos
I1
En
1.83
I1
0.52 1 1.14 j 0.40 e j 2 cos
39
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
40
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
41
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
ANTENNA ARRAYS
An antenna array is an antenna that is composed of more than one conductor. There
are two types of antenna arrays:
Driven arrays all elements in the antenna are fed RF from the transmitter
Parasitic arrays only one element is connected to the transmitter. The other elements
are coupled to the driven element through the electric fields and magnetic fields that
exist in the near field region of the driven element
There are many types of driven arrays. The four most common types are:
Collinear array
Broadside array
Log Periodic Array
Yagi-Uda Array
COLLINEAR ARRAY
The collinear array consists of /2 dipoles oriented end-to-end. The center dipole is
fed by the transmitter and sections of shorted transmission line known as phasing lines
connect the ends of the dipoles as shown below.
Feed Line
Phasing Lines
Phasing Lines
The length of the phasing lines are adjusted so that the currents in all the dipole
sections are in phase, as shown below.
The input impedance of a collinear array is approximately 300 ohms. The directivity
of a collinear array slowly increases as the number of collinear sections is increased.
BROADSIDE ARRAY
A broadside array consists of an array of dipoles mounted one above another as shown
below. Each dipole has its own feed line and the lengths of all feed lines are equal so
that the currents in all the dipoles are in phase.
Rows of broadside arrays can be combined to form a two dimensional array as shown
below:
The two-dimensional array is used in high performance radar systems. The amplitude
and phase of each input current is adjusted so that the antenna radiates its RF in a
narrow beam. By making changes to the input phase and amplitude, the beam can be
made to scan over a wide range of angles. Electronic scanning is much faster than
mechanical scanning (which uses a rotating antenna) and permits rapid tracking of
large numbers of targets.
UNIT - IV
SPECIAL ANTENNAS
Helical Antennas
2.1 Introduction
The helical antenna is a hybrid of two simple radiating elements, the dipole and
loop antennas. A helix becomes a linear antenna when its diameter approaches zero or
pitch angle goes to 90 o . On the other hand, a helix of fixed diameter can be seen as a
loop antenna when the spacing between the turns vanishes ( = 0 o ) .
Helical antennas have been widely used as simple and practical radiators over the
last five decades due to their remarkable and unique properties. The rigorous analysis of
a helix is extremely complicated. Therefore, radiation properties of the helix, such as
gain, far-field pattern, axial ratio, and input impedance have been investigated using
experimental methods, approximate analytical techniques, and numerical analyses. Basic
radiation properties of helical antennas are reviewed in this chapter.
If one turn of the helix is unrolled, as shown in Figure 2.1(b), the relationships between
S ,C , and the length of wire per turn, L , are obtained as:
S = L sin = C tan
L = ( S 2 + C 2 )1 / 2 = ( S 2 + 2 D 2 )1 / 2
When the helix is limited in length, it radiates and can be used as an antenna.
There are two radiation modes of important practical applications, the normal mode and
the axial mode. Important properties of normal-mode and axial-mode helixes are
summarized below.
Figure 2.1 (a) Geometry of helical antenna; (b) Unrolled turn of helical antenna
Figure 2.2 Instantaneous charge distribution for transmission modes: (a) The lowest-order
mode (T0); (b) The second-order mode (T1)
For a helical antenna with dimensions much smaller than wavelength ( NL << ) ,
the current may be assumed to be of uniform magnitude and with a constant phase along
the helix [5]. The maximum radiation occurs in the plane perpendicular to the helix axis,
as shown in Figure 2.3a. This mode of operation is referred to as the normal mode. In
general, the radiation field of this mode is elliptically polarized in all directions. But,
under particular conditions, the radiation field can be circularly polarized. Because of its
small size compared to the wavelength, the normal-mode helix has low efficiency and
narrow bandwidth.
C << C
(a) (b)
Figure 2.3 Radiation patterns of helix: (a) Normal mode; (b) Axial mode
Unlike the dipole and loop antennas, the helix has a complicated geometry. There
are no exact solutions that describe the behavior of a helix. However, using experimental
methods and approximate analytical or numerical techniques, it is possible to study the
radiation properties of this antenna with sufficient accuracy. This section briefly
discusses the analysis of normal-mode and axial-mode helices.
Since the helix dimensions are much smaller than wavelength, the far-field pattern
is independent of the number of turns. It is possible to calculate the total far-field of the
normal-mode helix by combining the fields of a small loop and a short dipole connected
in series. Doing so, the result for the electric field is expressed as [6]
r kI e jkr 2D2
= j 0 sin ( S j ) , (2.1)
4r 2
2
where k = is the propagation constant, = is the intrinsic impedance of the
medium, and I 0 is a current amplitude. As noted in (2.1), the and components of
the field are in phase quadrature. Generally, the polarization of this mode is elliptical
with an axial ratio given by
E 2 S
AR = = . (2.2)
E 2D2
Unlike the case of a normal-mode helix, simple analytical solutions for the axial-
mode helix do not exist. Thus, radiation properties and current distributions are obtained
using experimental and approximate analytical or numerical methods.
The current distribution of a typical axial-mode helix is shown in Figure 2.5 [5].
As noted, the current distribution can be divided into two regions. Near the feed region,
the current attenuates smoothly to a minimum, while the current amplitude over the
remaining length of the helix is relatively uniform. Since the near-feed region is small
compared to the length of the helix, the current can be approximated as a travelling wave
of constant amplitude. Using this approximation, the far-field pattern of the axial-mode
helix can be analytically determined. There are two methods for the analysis of far-field
pattern. In the first method, an N-turn helix is considered as an array of N elements with
an element spacing equal to S . The total field pattern is then obtained by multiplying the
pattern of one turn of the helix by the array factor. The result is
sin( N )
F ( ) = c 0 cos 2 , (2.4)
sin( )
2
where c 0 is a constant coefficient and = kS cos + . Here, is the phase shift
sin( N )
In (2.4), cos is the element pattern and 2 is the array factor for a uniform array
sin( )
2
of N equally-spaced elements. As noted from (2.5), the Hansen-Woodyard condition is
satisfied. This condition is necessary in order to achieve agreement between the
measured and calculated patterns.
where
l = the length of wire from the beginning of the helix to an
arbitrary point
LT
g=
pc m
1
= (according
sin + (2 N + 1) ( cos )
N C
to Hansen-Woodyard condition)
= 2N
C and S are, respectively, the circumference and spacing between turns of the
helix normalized to the free space wavelength ( ) . Since the axial-mode helix is nearly
lossless, the directivity and the gain expressions are approximately the same.
In 1980, King and Wong [8] reported that Krauss gain formula (2.10)
overestimates the actual gain and proposed a new gain expression using a much larger
experimental data base. The new expression is given as
N + 2 1 0.8 N
D NS tan 12.5o 2
G P = 8.3 , (2.11)
P P tan
where P is the free-space wavelength at peak gain.
In 1995, Emerson [9] proposed a simple empirical expression for the maximum
gain based on numerical modeling of the helix. This expression gives the maximum gain
Half-Power Beamwidth
A more accurate formula was later presented by King and Wong using a larger
experimental data base [10]. This result is
0.6
2N
61.5 N
N +5 tan 4
HPBW = o
(degrees). (2.14)
D
N
4
NS
0.7 tan 12.5
Input Impedance
The input impedance, however, is sensitive to feed geometry. Our numerical modeling of
the helix indicated that (2.15) is at best a crude approximation of the input impedance.
Bandwidth
Based on the work of King and Wong [8], an empirical expression for gain
bandwidth, as a frequency ratio, has been developed:
4
(3 N )
fU 0.91
1.07 , (2.16)
fL GG
P
where f U and f L are the upper and lower frequencies, respectively, G P is the peak gain
from equation (2.11), and G is the gain drop with respect to the peak gain.
summarized in Table 2.1. There are some helices with parameters outside the ranges in
Table 2.1 that exhibit unique properties. However, such designs are not regarded as
optimum, because not all radiation characteristics meet desired specifications. A
summary of the effects of various parameters on the performance of helix is presented
below [2].
Circumference
As shown in Figure 2.6, it is noted that the optimum circumference for achieving
the peak gain is around 1.1 and is relatively independent of the length of the helix.
Other results show that the peak gain smoothly drops as the diameter of the helix
decreases (Figure 2.7). Since other parameters of the helix also affect its properties, a
circumference of 1.1 is viewed as a good estimate for an optimum performance.
Pitch Angle
Keeping the circumference and the length of a helix fixed, the gain increases
smoothly when the pitch angle is reduced, as seen in Figure 2.8. However, the reduction
Figure 2.6 Gain of helix for different lengths as function of normalized circumference (C ) [9]
Figure 2.7 Peak gain of various diameter as D and varied (circles), D fixed and varied
(triangle) [8].
of pitch angle is limited by the bandwidth performance. That is, a narrower bandwidth is
obtained for a helix with a smaller pitch angle. For this reason, it has been generally
agreed that the optimal pitch angle for the axial-mode helix is about 12.5o .
Number of Turns
Many properties, such as gain, axial ratio, and beamwidth, are affected by the
number of turns. Figure 2.9 shows the variation of gain versus the number of turns. It is
noted that as the number of turns increases, the gain increases too. The increase in gain is
simply explained using the uniformly excited equally-spaced array theory. However, the
gain does not increase linearly with the number of turns, and, for very large number of
turns, adding more turns has little effect. Also, as shown in Figure 2.10, the beamwidth
becomes narrower for larger number of turns. Although adding more turns improves the
gain, it makes the helix larger, heavier, and more costly. Practical helices have between 6
and 16 turns. If high gain is required, array of helices may be used.
Conductor Diameter
This parameter does not significantly affect the radiation properties of the helix.
For larger conductor diameters, slightly wider bandwidths are obtained. Also, thicker
conductors can be used for supporting a longer antenna.
Ground Plain
Figure 2.8 Gain versus frequency of 30.8-inch length and 4.3-inch diameter helix for different
pitch angles [8].
Figure 2.9 Gain versus frequency for 5 to 35-turn helical antennas with 4.23-inch diameter [8]
Various modifications of the conventional helical antenna have been proposed for
the purpose of improving its radiation characteristics. A summary of these modifications
is presented below.
Nakano and Yamauchi [11] have proposed a modified helix in which the open
end section is tapered as illustrated in Figure 2.11. This structure provides significant
improvement in the axial ratio over a wide bandwidth. According to them, the axial ratio
improves as the cone angle t is increased. For a helix with pitch angle of 12.5 o and 6
turns followed by few tapered turns, they obtained an axial ratio of 1:1.3 over a frequency
range of 2.6 to 3.5 GHz.
To reduce the size of a helix operating in the axial mode, a novel geometry
referred to as stub-loaded helix has been recently proposed [15]. Each turn contains four
stubs as illustrated in Figure 2.13. The stub-loaded helix provides comparable radiation
properties to the conventional helix with the same number of turns, while offering an
approximately 4:1 reduction in the physical size.
This antenna consists of a helix and a monopole, as shown in Figure 2.14, [16].
The purpose of this modified antenna is to maintain operation at two different
frequencies, applicable to dual-band cellular phone systems operating in two different
frequency bands (900 MHz for GSM and 1800 MHz for DCS1800).
r0
t
c
X
A
log periodic antenna is a system of driven elements, designed to be operated over a wide range of
frequencies. Its advantage is that it exhibits essentially constant characteristics over the frequency
rangethe same radiation resistance (and therefore the same SWR), and the same pattern character-
istics (approximately the same gain and the same front-to-back ratio). Not all elements in the system are
active on a single frequency of operation; the design of the array is such that the active region shifts among
the elements with changes in operating frequency. R. H. DuHamel and D. E. Isbell published the first
information on log periodic arrays in professional literature in the late 1950s. The first log-periodic antenna
article to be published in amateur literature appeared in November 1959 QST, and was written by Carl T.
Milner, W1FVY. (See the Bibliography at the end of this chapter.)
Several varieties of log periodic antenna systems exist, such as the zig-zag, planar, trapezoidal,
slot, V, and the dipole. The type favored by amateurs is the log-periodic dipole array, often abbreviated
LPDA. The LPDA, shown in Fig 1, was invented by D. E. Isbell at the University of Illinois in 1958.
Similar to a Yagi antenna in construction and appearance, a log-periodic dipole array may be built as a
rotatable system for all the upper HF bands, such as 18 to 30 MHz. The longest element, at the rear of
the array, is a half wavelength at the lower design frequency.
Depending on its design parameters, the LPDA can be operated over a range of frequencies having a
ratio of 2:1 or higher. Over this range its electrical characteristicsgain, feed-point impedance, front-to-
back ratio, and so forthremain more or less constant. This is not true of any other type of antenna dis-
cussed in this book. With a Yagi or quad antenna,
for example, either the gain factor or the front-to-
back ratio, or both, deteriorate rapidly as the fre-
quency of operation departs from the optimum de-
sign frequency of the array. And because those an-
tennas are based on resonant elements, off-reso-
nance operation introduces reactance which causes
the SWR in the feeder system to increase. Even
terminated antennas such as a rhombic exhibit sig-
nificant changes in gain over a 2:1 frequency ratio.
As may be seen in Fig 1, the log periodic ar-
ray consists of several dipole elements which are
each of different lengths and different relative
spacings. A distributive type of feeder system is
Fig 1A log periodic dipole array. All elements
used to excite the individual elements. The ele- are driven, as shown. The forward direction of
ment lengths and relative spacings, beginning the array as drawn here is to the right.
from the feed point for the array, are seen to in- Sometimes the elements are sloped forward,
crease smoothly in dimension, being greater for and sometimes parasitic elements are used to
each element than for the previous element in the enhance the gain and front-to-back ratio.
Log Periodic Arrays 10-1
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
array. It is this feature upon which the design of the LPDA is based, and which permits changes in
frequency to be made without greatly affecting the electrical operation. With changes in operating
frequency, there is a smooth transition along the array of the elements which comprise the active re-
gion. The following information is based on a November 1973 QST article by Peter Rhodes, K4EWG.
A good LPDA may be designed for any single amateur band or for adjacent bands, HF to UHF, and can
be built to meet the amateurs requirements at nominal cost: high forward gain, good front-to-back ratio,
low SWR, and a boom length equivalent to a full-sized 3-element Yagi. The LPDA exhibits a relatively low
SWR (usually not greater than 2:1) over a wide band of frequencies. A well-designed LPDA can yield a
1.3:1 SWR over a 1.8-to-1 frequency range with a typical gain of 7.0 dB over an isotropic radiator (dBi)
assuming a lossless system. This equates to approximately 4.9 dB gain over a half-wave dipole (dBd).
BASIC THEORY
The LPDA is frequency independent in that the electrical properties vary periodically with the
logarithm of the frequency. As the frequency, f1, is shifted to another frequency, f2, within the pass-
band of the antenna, the relationship is
f2 = f1/ (Eq 1)
where
= a design parameter, a constant; < 1.0. Also,
f3 = f1/2
f4 = f1/3
.
.
.
fn = f1/n 1
n = 1, 2, 3, . . . n
f1 = lowest frequency
fn = highest frequency
where ln dn,n 1
= =
ln 1 dn 2,n 1
l n = shortest element length, and
d23 = d 12 dn,n 1
d34 = d 23 =
2ln 1
.
where
. l = element length
. d = element spacing
= design constant
dn 1,n = dn 2,n 1 (Eq 3) = relative spacing constant
S = feeder spacing
where d23 = spacing between elements 2 and 3. Z0 = characteristic impedance of antenna feeder
10-2 Chapter 10
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Each element is driven with a phase shift of 180 by switching or alternating element connections,
as shown in Fig 2. At a median frequency the dipoles near the input, being nearly out of phase and close
together, nearly cancel each others radiation. As the element spacing, d, increases along
the array, there comes a point where the phase delay in the transmission line combined with the 180
switch gives a total of 360. This puts the radiated fields from the two dipoles in phase in a direction
toward the apex. Hence, a lobe coming off the apex results.
This phase relationship exists in a set of dipoles known as the active region. If we assume that an
LPDA is designed for a given frequency range, then that design must include an active region of di-
poles for the highest and lowest design frequency. It has a bandwidth which we shall call B ar, band-
width of the active region.
Assume for the moment that we have a 12-element LPDA. Currents flowing in the elements are both
real and imaginary, the real current flowing in the resistive component of the impedance of a particular
dipole, and the imaginary flowing in the reactive component. Assume that the operating frequency is such
that element number 6 is near to being half-wave resonant. The imaginary parts of the currents in shorter
elements 7 to 12 are capacitive, while those in longer elements 1 to 5 are inductive. The capacitive current
components in shorter elements 9 and 10 exceed the conductive components; hence, these elements receive
little power from the feeder and act as parasitic directors. The inductive current components in longer ele-
ments 4 and 5 are dominant and they act as parasitic reflectors. Elements 6, 7 and 8 receive most of their
power from the feeder and act as driven elements. The amplitudes of the currents in the remaining elements
are small and they may be ignored as primary contributors to the radiation field. Hence, we have a general-
ized Yagi array with seven elements comprising the active region. It should be noted that this active region
is for a specific set of design parameters ( = 0.93, = 0.175). The number of elements making up the active
region varies with and . Adding more elements on either side of the active region cannot significantly
modify the circuit or field properties of the array.
This active region determines the basic design parameters for the array, and sets the bandwidth for
the structure, Bs . That is, for a design-frequency coverage of bandwidth B, there exists an associated
bandwidth of the active region such that
B s = B Bar (Eq 4)
where
f
B = operating bandwidth = n (Eq 5)
f1
f1 = lowest frequency, MHz
f n = highest frequency, MHz
B ar varies with and as shown in Fig 3. Element lengths which fall outside Bar play an insignifi-
10-4 Chapter 10
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Zav = average characteristic impedance of a dipole
l
= 120 In n 2.25 (Eq 10)
d n
l n /diamn = length to diameter ratio of nth element
= mean spacing factor = (Eq 11)
The mean spacing factor, , is a function of and (Eqs 6 and 11). For a fixed value of Z 0, R0
decreases with increasing and increasing .
If all element diameters are identical, then the element l/diam ratios will increase along the array.
Ideally the ratios should remain constant, but from a practical standpoint the SWR performance of a single-
band LPDA will not be noticeably degraded if all elements are of the same diameter. But to minimize SWR
variations for multiband designs, the LPDA may be constructed with progressively increasing element di-
ameters from the front to the back of the array. This approach also offers structural advantages for self-
supporting elements, as larger conductors will be in place for the longer elements.
The standing-wave ratio varies periodically with frequency. The mean value of SWR, with respect to
R0, has a minimum of about 1.1:1 at opt (Eq 8), and rises to a value of 1.8:1 at = 0.05. In other words, the
periodic SWR variation (with frequency changes) swings over a wider range of SWR values with lower
values of . These SWR ranges are acceptable when using standard 52 and 72- coax for the feed line.
However, a 1:1 SWR match can be obtained at the transmitter end by using a coax-to-coax Transmatch. A
Transmatch enables the transmitter low-pass filter to see a 52- load on each frequency within the array
passband. The Transmatch also eliminates possible harmonic radiation caused by the frequency-indepen-
dent nature of the array.
R 0 should be chosen for the intended balun and feed-line characteristics. For HF arrays, a value of
208 for R0 usually works well with a 4:1 balun and 52- coax. Direct 52- feed is usually not
possible. (Attempts may result in smaller conductor spacing for the antenna feeder than the conductor
diameter, a physical impossibility.)
For VHF and UHF designs, the antenna feeder may also serve as the boom. With this technique,
element halves are supported by feeder conductors of tubing that are closely spaced. If R0 is selected
as 72 , direct feed with 72- cable is possible. An effective balun exists if the coax is
passed through one of the feeder conductors from the rear of the array to the feed point. Fig 5 shows
such a feed-point arrangement.
If the design bandwidth of the array is fairly small (single band), another possible approach is to
design the array for a 100- R0 and use a 1 /4-wave matching section of 72- coax between the feed
point and 52- feed line. In any case, select the element feeder diameters based on mechanical
considerations. The required feeder spacing may then be calculated.
The antenna feeder termination, Zt, is a short circuit at a distance of max/8 or less behind element no. 1,
the longest element. In his 1961 paper on LPDAs, Dr Robert L. Carrel reported satisfactory results in some
cases by using a short circuit at the terminals of el-
ement no. 1. If this is done, the shorted element acts
as a passive reflector at the lowest frequencies.
Some constructors indicate that Zt may be elimi-
nated altogether without significant effect on the
results. The terminating stub impedance tends to
increase the front-to-back ratio for the lowest fre-
quencies. If used, its length may be adjusted for the
best results, but in any case it should be no longer
than max/8. For HF-band operation a 6-inch short-
ing jumper wire may be used for Zt.
It might also be noted that one could increase Fig 5A method of feeding the LPDA for VHF
the front-to-back ratio on the lowest frequency by and UHF designs.
S= ( diam
2 )
10 Z 0 /276
(Eq 20)
where
S = center-to-center distance between conductors
diam = outer diameter of conductor (in same units as S)
Z0 = intended characteristic impedance for antenna feeder
Note: This equation assumes round feeder conductors.
If an impractical spacing results for the antenna feeder, select a different conductor diameter and
repeat step 11. In severe cases it may be necessary to select a different R0 and repeat steps 10 and 11.
Once a satisfactory feeder arrangement is found, the LPDA design is completed.
Design ExampleShort Four-Band Array
Suppose we wish to design a log periodic dipole array to cover the frequency range 18.06 to
29.7 MHz. Such an array will offer operation on any frequency in the 17, 15, 12 and 10-meter amateur
bands. In addition, we desire for this to be a short array, constructed on a boom of no more than
10 feet in length.
To follow through this example, it is suggested that you write the parameter names and their values
as they are calculated, in columns, on your worksheet. This will provide a ready reference for the
values needed in subsequent calculations.
We begin the design procedure with step 1 and determine the operating bandwidth from Eq 5: f1 =
18.06, f n = 29.7, and B = 29.7/18.06 = 1.6445. (Note: Because log periodics have reduced gain at the
low-frequency end, some designers lower f1 by several percent to assure satisfactory gain at the lower
operating frequencies. Increasing f n, the design frequency at the high end, however, appears to offer no
advantage other than extended frequency coverage.) Because we wish to have a compact design, we
choose not to extend the lower frequency range.
Next, step 2, we examine Fig 4 and choose values for , and gain. Knowing from the basic
theory section that larger values of call for a longer boom, we choose the not-too-large value of 0.06.
Also knowing that a compact array will not exhibit high gain, we choose a modest gain, 8.0 dBi. For
these values of and gain, Fig 4 shows the required to be 0.885.
From step 3 and Eq 12, we determine the value for cot to be 4 0.06/(1 0.885) = 2.0870. We need
not determine , the apex half angle, but if we wish to go to the trouble we can use the relationship
= arc cot 2.0870 = arc tan (1/2.0870) = 25.6
This means the angle at the apex of the array will be 2 25.6 = 51.2.
From step 4 and Eq 13, we calculate the value for Bar as 1.1 + 7.7(1 0.885)2 2.097 = 1.3125.
Next, from step 5 and Eq 4, we determine the structure bandwidth Bs to be 1.6445 1.3125 = 2.1584.
From step 6 and the associated equations we determine the boom length, number of elements, and
Log Periodic Arrays 10-7
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
longest element length.
( 1
) 984
L = 1/4 1 2.1584 2.0870 18.06 = 15.26 ft
log 2.1584
N =1 + = 1 + 0.3341 = 7.30
log (1/0.885) 0.05306
(Because a ratio of logarithmic values is determined here, either common or natural logarithms
may be used in the equation, as long as both the numerator and the denominator are the same type; the
results are identical.)
l 1 = 492/18.06 = 27.243 ft
The 15.26-foot boom length is greater than the 10-foot limit we desired, so some change in the
design is necessary. The 7.30 elements should be increased to 8 elements if we chose to proceed with
this design, adding still more to the boom length. The longest element length is a function solely of the
lowest operating frequency, so we do not wish to change that.
Decreasing either or will yield a shorter boom. Because is already close to the minimum
value of 0.05, we decide to retain the value of 0.06 and decrease the value of . Lets try = 0.8.
Repeating
1
/4 steps 2 through 6 with these values, we calculate the following.
Gain = 5.3 dBi? (outside curves of graph)
cot = 1.2000
Bar = 1.4696
Bs = 2.4168
L = 9.58 ft
N = 4.95
l 1 = 27.243 ft
These results nicely meet our requirement for a boom length not to exceed 10 feet. The 4.95 elements
obviously must be increased to 5. The 5.3 dBi gain (3.2 dBd) is nothing spectacular, but the array should
have a reasonable front-to-back ratio. For four-band coverage with a short boom, we decide this gain and
array dimensions are acceptable, and we choose to go ahead with the design. The variables summarized on
our worksheet now should be those shown in the first portion of Table 1.
Continuing at step 7, we make plans to use a
6-inch shorted jumper for the terminating stub, Table 1
Zt . Design Parameters for the 4-Band LPDA
From step 8 and Eq 2 we determine the ele- f1 = 18.06 MHz Element lengths:
f n = 29.7 MHz l 1 = 27.243 ft
ment lengths: B = 1.6445 l 2 = 21.794 ft
l2 = l1 = 0.8 27.243 = 21.794 ft
= 0.8 l 3 = 17.436 ft
= 0.06 l 4 = 13.948 ft
l 3 = 0.8 21.794 = 17.436 ft Gain = 5.3 dBi = 3.2 dBd l 5 = 11.159 ft
l 4 = 0.8 17.436 = 13.948 ft cot = 1.2000 Element spacings:
Bar = 1.4696 d12 = 3.269 ft
l 5 = 0.8 13.948 = 11.159 ft Bs = 2.4168 d23 = 2.616 ft
L = 9.58 ft d34 = 2.092 ft
From step 9 and Eq 18 we calculate the ele- N = 4.95 elements (in- d45 = 1.674 ft
ment spacing d12 as 1/2 (27.243 21.794) 1.2 = crease to 5) Element diameters:
3.269 ft. Then from Eq 3 we determine the re- Zt = 6-in. shorted jumper diam5 = 1 / 2 in.;
Ro = 208 l 5/diam5 = 267.8
maining element spacings: Zav = 400.8 diam4 = 5 / 8 in.;
= 0.06708 l 4/diam4 = 267.8
d23 = 0.8 3.269 = 2.616 ft Z0 = 490.5 diam3 = 3 / 4 in.;
d34 = 0.8 2.616 = 2.092 ft Antenna feeder: l 3/diam3 = 279.0
d45 = 0.8 2.092 = 1.674 ft #12 wire spaced 2.4 in. diam2 = 1 in.;
l 2/diam2 = 261.5
Balun: 4 to 1
Feed line: 52- coax diam 1 = 1 1/ 4 in.;
This completes the calculations of the array l 1/diam1 = 261.5
dimensions. The work remaining is to design the
10-8 Chapter 10
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
antenna feeder. From step 10, we wish to feed the LPDA with 52- line and a 4:1 balun, so we select
R 0 as 4 52 = 208 .
Before we calculate Z0 from Eq 19 we must first determine Zav from Eq 10. At this point we must
assign a diameter to element no. 5. We wish to make the array rotatable with self-supporting elements,
so we shall use aluminum tubing for all elements. For element no. 5, the shortest element, we plan to
use tubing of 1 /2-inch OD. We calculate the length to diameter ratio by first converting the length to
inches:
At this point in the design process we may also assign diameters to the other elements. To maintain
an essentially constant l /diam ratio along the array, we shall use larger tubing for the longer elements.
(From a practical standpoint for large values of , 2 or 3 adjacent elements could have the same diam-
eter. For a single-band design, they could all have the same diameter.) From data in Chapter 21 we see
that, above 1/2 inch, aluminum tubing is available in diameter steps of 1 /8 inch. We assign additional
element diameters and determine l /diam ratios as follows:
Tapered elements with telescoping tubing at the ends may certainly be used. From a matching
standpoint, the difference from cylindrical elements is of minor consequence. (Performance at the
low-frequency end may suffer slightly, as tapered elements are electrically shorter than their cylindri-
cal counterparts having a diameter equal to the average of the tapered sections. See Chapter 2.)
In Eq 10 the required length to diameter ratio is that for element no. 5, or 267.8. Now we may
determine Zav as
Zav = 120 [ln 267.8 2.25] = 120 [5.590 2.25] = 400.8
Additionally, before solving for Z0 from Eq 19, we must determine from Eq 11.
0.06 = 0.06708
=
0.8
And now we use Eq 19 to calculate Z 0.
2
From step 11, we are to determine the conductor size and spacing for a Z0 of 490.5 for the
antenna feeder. We elect to use #12 wire, and from data in Chapter 20 learn that its diameter is 80.8
mils or 0.0808 inch. We determine the spacing from Eq 20 as
( )
S = 0.0808 10 490.5/276 = 0.0808 101.777
2 2
= 0.0808 59.865 = 2.42 in.
2
An open-wire line of #12 wire with 2.4-inch spacers may be used for the feeder. This completes the
design of the four-band LPDA. The design data are summarized in Table 1.
Fig 7Pieces to be fabricated for the LPDA. At A, the forward connector, made from 1/2-in. Lexan. At B,
the rear connector, also made from 1/2-in. Lexan. At C is the pattern for the feed-line spacers, made from
1
/4-in. Plexiglas. Two of these spacers are required.
From the design procedure, the feeder spacings for the two arrays are slightly different, 0.58 inch
for the 3.5-MHz array and 0.66 inch for the 7-MHz version. As a compromise toward the use of
common spacers for both bands, a spacing of 5 /8 inch is quite satisfactory. Surprisingly, the feeder
spacing is not at all critical here from a matching standpoint, as may be verified from Z0 = 276 log (2S/
diam) and from Eq 9. Increasing the spacing to as much as 3/4 inch results in an R0 SWR of less than 1.1
to 1 on both bands.
Constructing the Arrays
The construction techniques are the same for both the 3.5 and the 7-MHz versions of the array. Once
the designs are completed, the next step is to fabricate the fittings; see Fig 7 for details. Cut the wire
elements and feed lines to the proper sizes and mark them for identification. After the wires are cut and
placed aside, it will be difficult to remember which is which unless they are marked. When you have
finished fabricating the connectors and cutting all of the wires, the antenna can be assembled. Use your
ingenuity when building one of these antennas; it isnt necessary to duplicate these LPDAs exactly.
The elements are made of standard #14 stranded copper wire. The two parallel feed lines are made
of #12 solid copper-coated steel wire, such as Copperweld. This will not stretch when placed under
tension. The front and rear connectors are cut from 1/2 -inch thick Lexan sheeting, and the feed-line
spacers from 1/4-inch Plexiglas sheeting.
Study the drawings carefully and be familiar with the way the wire elements are connected to the
two feed lines, through the front, rear and spacer connectors. Details are sketched in Figs 8 and 9.
Connections made this way prevent the wire from breaking. All of the rope, string and connectors must
be made of materials that can withstand the effects of tension and weathering. Use nylon rope and
strings, the type that yachtsmen use. Fig 6 shows the front stay rope coming down to ground level at a
point 120 feet from the base of a 75-foot tower. It may not be possible to do this in all cases. An
alternative installation technique is to put a pulley 40 feet up in a tree and run the front stay rope
through the pulley and down to ground level at the base of the tree. The front stay rope will have to be
tightened with a block and tackle at ground level.
Log Periodic Arrays 10-11
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Putting an LPDA together is not difficult if it is assembled in an orderly manner. It is easier to
connect the elements to the feeder lines when the feed-line assembly is stretched between two points.
Use the tower and a block and tackle. Attaching the rear connector to the tower and assembling the
LPDA at the base of the tower makes raising the antenna into place a much simpler task. Tie the rear
connector securely to the base of the tower and attach the two feeder lines to it. Then thread the two
feed-line spacers onto the feed line. The spacers will be loose at this time, but will be positioned
properly when the elements are connected. Now connect the front connector to the feed lines. A word
of caution: Measure accurately and carefully! Double-check all measurements before you make per-
manent connections.
Connect the elements to the feeder lines through their respective plastic connectors, beginning
with element 1, then element 2, and so on. Keep all of the element wires securely coiled. If they
unravel, you will have a tangled mess of kinked wire. Check that the element-to-feeder connections
have been made properly. (See Figs 8 and 9.) Once you have completed all of the element connec-
tions, attach the 4:1 balun to the underside of the front connector. Connect the feeder lines and the
coaxial cable to the balun.
You will need a separate piece of rope and a pulley to raise the completed LPDA into position.
First secure the eight element ends with nylon string, referring to Figs 6 and 8. The string must be long
enough to reach the tie-down points. Connect the front stay rope to the front connector, and the com-
Fig 8Typical layout for the LPDA. Use a 4:1 Fig 9Details of electrical and mechanical
balun at the point indicated. See Tables 2 and 3 connections of the elements to the feed line.
for dimensions. Knots in the nylon stay lines are not shown.
10-12 Chapter 10
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
pleted LPDA is now ready to be raised into position. While raising the antenna, uncoil the element
wires to prevent their getting away and tangling up into a mess. Use care! Raise the rear connector to
the proper height and attach it securely to the tower, then pull the front stay rope tight and secure
it. Move the elements so they form a 60-degree angle with the feed lines, in the direction of the front,
and space them properly relative to one another. By adjusting the end positions of the elements as you
walk back and forth, you will be able to align all the elements properly. Now it is time to hook your rig
to the system and make some contacts.
Performance
The reports received from these LPDAs were compared with an inverted-V dipole. All of the
antennas are fixed; the LPDAs radiate to the northeast, and the dipole to the northeast and southwest.
The apex of the dipole is at 70 feet, and the 40 and 80-meter LPDAs are at 60 and 50 feet, respectively.
The gain of the LPDAs is several dB over the dipole. This was apparent from many of the reports
received. During pileups, it was possible to break in with a few tries on the LPDAs, yet it was impos-
sible to break in the same pileups using the dipole.
During the CQ WW DX Contest some big pileups were broken after a few calls with the LPDAs.
Switching to the dipole, it was found impossible to break in after many, many calls. Then, after switch-
ing back to the LPDA, it was easy to break into the same pileup and make the contact.
Think of the possibilities that these wire LPDA systems offer hams worldwide. They are easy to
design and to construct, real advantages in countries where commercially built antennas and parts are
not available at reasonable cost. The wire needed can be obtained in all parts of the world, and cost of
construction is low! If damaged, the LPDAs can be repaired easily with pliers and solder. For those
who travel on DXpeditions where space and weight are large considerations, LPDAs are lightweight
but sturdy, and they perform well.
Fig 11Measured
radiation pattern
of the 13-30 MHz
LPDA at 14 MHz.
The front-to-back
ratio is 14.4 dB at
14 MHz, and
increases to
21 dB at 28 MHz.
Fig 12Construction diagram of the 13-30 MHz log periodic array. At B and C are shown the method
of making electrical connection to each half element, and at D is shown how the boom sections are
joined.
Log Periodic Arrays 10-15
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
bowstring it is important to use a wire that will not
stretch, such as #14 Copperweld. This bowstring is ac-
tually the inter-element transmission line. See Fig 17.
Secure the rear ends of the feeder to the two
rear insulators, soldering the wrap. Before secur-
ing the fronts, slip the 12 insulators onto the two
feed lines. A rope can be used temporarily to form
the bow and to aid in mounting the feeder line. The
end-to-end length of the feeder should be 30.24 feet.
Now lift both bows to their upright position
and tie the feeder line and the cross-arm bowstring
together where they cross, directly over and ap- Fig 15The spider web antenna, as shown in
proximately 3 feet above the hub. this somewhat deceptive photo, might bring to
The next step is to install the no. 1 rear element mind a rotatable clothesline. Of course it is
from the rear egg insulators to the right and left cross- much larger than a clothesline, as indicated by
arms using other egg insulators to provide the proper Figs 13 and 14. It can be lifted by hand.
element length. Be sure to solder the element halves
to the transmission line. Complete this portion of the
construction by installing the nylon cord catenaries
from the front arm to the cross-arm tips. Use egg insu-
lators where needed to prevent cutting the nylon cords.
In preparing the fiberglass front spreader, keep
in mind that it should be 14.75 feet long before
bowing and is approximately 13.75 feet when
bowed. Secure the center of the bowstring to the
end of the front arm. Lay the spreader on top of the
feed line, then tie the feeder to the spreader with
nylon fish line. String the catenary from the
spreader tips to the cross-arm tips.
At this point of assembly, antenna elements 2
through 13 should be prepared. There will be two
segments for each element. At the outer tip make a
small loop and solder the wrap. This will be for the
nylon leader. Measure the length plus 0.4 inch for
wrapping and soldering the element segment to the Fig 16The simple arrangement of the hub of
feeder. Seven-strand #22 antenna wire is suggested the spider web. See Fig 13 and the text for
for use here. Slide the feed-line insulators to their details.
proper position and secure them temporarily.
The drawings show the necessary transposition
scheme. Each element half of elements 1, 3, 5, 7, 9,
11 and 13 is connected to its own side of the feeder,
while elements 2, 4, 6, 8, 10 and 12 cross over to the
opposite side of the transmission line.
There are four holes in each of the transmis-
sion-line insulators (see Fig 13). The inner holes
are for the transmission line, and the outer ones
are for the elements. Since the array elements are
slanted forward, they should pass through the in-
sulator from front to back, then back over the insu-
lator to the front side and be soldered to the trans-
mission line. The small drawings of Fig 13 show Fig 17The elements, balun, transmission line
the details of the element transpositions. and main bow of the spider web antenna.
Log Periodic Arrays 10-19
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Construction
The general construction approach for the Pounder may be seen in the photographs. Drilled and tapped
pieces of Plexiglas sheet, 1/4-inch thick, serve as insulating spacers for the angle aluminum feeder. Two
spacers are used, one near the front and one near the rear of the array. Four no. 6-32 1/4-inch pan head screws
secure each aluminum angle section to the Plexiglas spacers, Figs 20 and 21. Use flat washers with each screw
to prevent it from touching the angle stock on the opposite side of the spacer. Be sure the screws are not so
long as to short out the feeder! A clearance of about 1/16 inch has been found sufficient. If you have doubts
about the screw lengths, check the assembled boom for a short with your ohmmeter on a megohms range.
Either of two mounting techniques may be used for the Pounder. As shown in Figs 18 and 19, the
rear spacer measures 10 21 /2 inches, with 45 corners to avoid sharp points. This spacer also accom-
modates a boom extension of PVC tubing, which is attached with two no. 10-32 1-inch screws. This
tubing provides for side mounting the Pounder away from a mast or tower.
An alternative support arrangement is shown in Fig 20. Two 1/2 3-inch Plexiglas spacers are used at the
front and rear of the array. Each spacer has four holes drilled 5/8 inch apart and tapped with no. 6-32 threads.
Two screws enter each spacer from either side to make a tight aluminum-Plexiglas-aluminum sandwich. At
the center of the boom, secured with only two screws, is a 2 18-inch strip of 1/4-inch Plexiglas. This strip is
slotted about 2 inches from each end to accept hose clamps for mounting the Pounder atop a mast. As shown,
the strip is attached for vertical polarization. Alternate mounting holes, visible on the now-horizontal lip of
the angle stock, provide for horizontal polarization. Al-
though sufficient, this mounting arrangement is not as
sturdy as that shown in Fig 18.
The elements are lengths of thick-wall aluminum
tubing, 1/4-inch OD. The inside wall conveniently
accepts a no. 10-32 tap. The threads should penetrate
the tubing to a depth of at least 1 inch. Eight no. 10-
32 1-inch screws are attached to the boom at the
proper element spacings and held in place with no.
10-32 nuts, Fig 19. For assembly, the elements are
then simply screwed into place.
Note that with this construction arrangement, the
two halves of any individual element are not precisely
collinear; their axes are offset by about 3/4 inch. This
offset does not seem to affect performance.
R
SWR = Rmax (Eq 7)
min
The weighted mean resistance level between the /2 and 3/2 modes is defined by
SWR 3/2
Rwm = R 01/2 R 01/2 SWR (Eq 8)
1/2
where
SWR1/2 = SWR in /2 mode
SWR3/2 = SWR in 3/2 mode
10-24 Chapter 10
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Once Z 0 and have been chosen, Fig 26 can be used to estimate the Rwm value for a given LPVA.
Notice the dominant role that Z0 (feeder impedance) plays in the array impedance.
It is apparent from the preceding data that the LPVA is useful for covering a number of different bands
spread over a wide range of the spectrum. It is fortunate that most of the amateur bands are harmonically
related. By choosing a large design parameter, = 0.9, a small relative spacing constant, = 0.02, and a tilt
angle of = 40, an LPVA could easily cover the amateur bands from 7 through 54 MHz!
DESIGN PROCEDURE
A step-by-step design procedure for the log periodic V array follows.
1) Determine the operational bandwidth, B,
in the /2 (fundamental) mode:
fn
B= (Eq 9)
f1
where
f1 = lowest frequency, MHz
f n = highest frequency, MHz
2) Determine for a desired number of elements,
n, using Fig 27.
3) Determine element lengths l1 to ln using Eqs
1 and 2 of this section.
4) Choose the highest operating mode desired and
determine and from Fig 28.
5) Determine cell boom length, L, from
2(l1 l n )
L= (Eq 10)
1
Note: If more than one LPVA cell is to be driven
by a common feeder, the spacing between cells
Fig 26Weighted mean resistance level, Rwm ,
can be determined from versus characteristic impedance of the feeder, Z0,
D12 = 21l n1 (Eq 11) for various angles.
Fig 27Design parameter, , versus number of elements, n, for various operational bandwidths, B.
Table 1
Design Dimensions for the LPVA
Element Element Design
Lengths, ft Spacings, ft Parameters Fig 29The element-to-boom detail is depicted here.
Aluminum angle brackets, U bolts, and sections of
l1 = 56.22* d 12 = 9.15 = 0.8
PVC tubing are shown securing each element to the
l2 = 56.22 d 23 = 7.32 = 0.05
boom at two points. The 300- twin-lead, threaded
l3 = 45.00 d 34 = 5.86 = 38.2 through a piece of polystyrene and attached to the
l4 = 36.00 d 45 = 4.67 L = 27 ft** foremost element, may be seen entering the picture
l5 = 28.79 = 45 at the top left. The end of the linear loading line for l1
*l1 is a shortened element; the full-size dimension is 70.28 ft. is visible near the bottom.
** The total physical boom length is L plus the distance to the l5 cross
bracing. The cross braces are 3 ft. long, and = 45; hence, the
total boom length is 27 ft + 1.5 ft = 28.5 ft.
Table 2
Basic Materials for the LPVA
10-26 Chapter 10
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
hat size if construction constraints prohibit a full-sized array. However, performance in higher mode
operations was less than optimum when shortened elements were used.
Linear Loading Stub Design
The following linear loading stub design equation may be used for approximating the stub length
(one half of element, two stubs required).
[ ] ( )
fh
2
24h
33.9 In d 1 1
234
Ls = 2.734 arctan
()
(Eq 12)
f b
fh log
a
where
Ls = linear loading stub length in feet required for each half element
h = element half length in feet
f = element resonant frequency in MHz
b = loading stub spacing in inches
a = radius (not diameter) of loading stub conductors in inches
d = average element diam in inches
Note: The resonant frequency, f, of an individual element of length, l , can be found from:
f = 468 (Eq 13)
l
The capacitance hat dimensions for each half element can be found from data in Chapter 2.
10-28 Chapter 10
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Table 2
Log-Yag Array Dimensions
Length Spacing
Element Feet Feet
Reflector 36.40 6.00 (Ref. to l 1)
l1 35.14 3.51 (d 12)
l2 33.27 3.32 (d 23)
l3 31.49 3.14 (d 34)
l4 29.81 10.57 ( l 4 to dir)
Director 32.20
Fig 33The attachment of the elements to the
boom.
Table 3
Element Material Requirements,
Log-Yag Array
1-in. 7
/ 8 -in. 3
/ 4 -in. 1/ 4 -in. 11/ 4-in.
Tubing Tubing Tubing Angle Bar
Lth Lth Lth Lth Lth
Ft Qty Ft Qty Ft Qty Ft Ft
Reflector 12 1 6 2 8 2 None None
l1 6 2 6 2 8 2 3 1
l2 6 2 6 2 8 2 3 1
l3 6 2 6 2 6 2 3 1
Fig 34Looking from the front to the back of the l4 6 2 6 2 6 2 3 1
Log-Yag array. A truss provides lateral and Director 12 1 6 2 6 2 None None
vertical support.
Table 4
Materials List, Log-Yag Array
10-30 Chapter 10
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Reconfigurable Antennas
Recongurable antennas change polarization, operating frequency, or far-eld pattern in order to cope with changing
system parameters. This paper reviews some of the past and current technology applicable to recongurable antennas,
with several examples of implementations. Mechanically movable parts and arrays are discussed, as well as more-
recent semiconductor-component and tunable-material technologies applicable to recongurable antennas.
Keywords: Recongurable antennas; recongurable arrays; antenna arrays; wideband antennas; reector antennas;
microstrip patch antennas
1. Introduction and phase of each element, the number of active elements, the
T
elements positions, or the polarization.
o configure means to arrange or organize the parts of
something to achieve a purpose. For instance, configuring The idea of reconfiguring an antenna is relatively old.
a microstrip antenna consists of determining the patch shape, In the early 1930s, the nulls of a two-element array were
substrate parameters, type and location of the feed, etc., in steered by using a calibrated variable phase changer in order
order for the antenna to radiate at a desired frequency and to determine the direction of arrival of a signal [1]. Bruce and
polarization. If the desired operating characteristics of the Beck changed the size of the rhombic antenna in Figure 1 by
antenna change, then the antenna must be reconfigured or stretching the wires with a motor and weights [2]. The element
rebuilt to meet the new specifications. Reconfigurable anten- lengths of the rhombus were 184 m, and the interior angles
nas change their performance characteristics by altering the were between 132 and 164. Reconfiguring the antenna con-
current flow on an antenna, using mechanically movable parts, sisted of adjusting the distance of the major axis by more
phase shifters, attenuators, diodes, tunable materials, or active than 100 m. Changing the shape of the rhombic wire antenna
materials. A reconfigurable antenna can be a single antenna or steered the beam in elevation. The Multiple-Unit Steerable
an array. Antenna (MUSA) was a six-element array of rhombic anten-
nas with phase shifters at five of the elements [3]. A linear
A reconfigurable antenna modifies the antennas pattern, phase shift steered the main beam. This concept was perfected
polarization, or frequency/bandwidth in some desirable fash- and introduced for azimuth scanning with a 14-row by three-
ion. For instance, a reconfigurable antenna might steer or place column array of polyrod antennas. This array had 13 rotary
a null in an antennas pattern, switch from right-hand circularly phase changers for beam steering [4]. The Wullenweber array
polarized to left-hand circularly polarized, or move the resonant was invented during WW II. It is a large direction-finding cir-
frequency from 2 GHz to 2.4 GHz. In this way, a single antenna cular array with a narrow beam that scans 360 in azimuth by
replaces two or more antennas, in order to achieve multiple activating a small group of adjacent elements along the circle
goals. Arrays are the ultimate reconfigurable antenna, because G
they have many avenues for controlling the current across
the aperture. A reconfigurable array may have reconfigurable In 1979, reconfigurability was defined as the ability
elements, or a reconfigurable array may change the amplitude to adjust beam shapes upon command [6]. The authors used
1
fc = . (1)
2 Coff Ron Figure 8. A recongurable slotted-patch antenna.
6. Tunable Materials
Tunability
Method Q Stimulus Frequency
(%)
Electric Field
Dielectric Thin Film [42] 30 45 1-20 GHz
E = 70 kV/cm
Electric Field
Dielectric Bulk Ceramic [43] 16 > 100 1-10 GHz
E = 15 kV/cm
Magnetic Field
Magnetic Film [44] 15 5 < Q < 11 0.3-5 GHz
80 kA/m (1000 Oe)
3.4* Magnetic Field
Magnetic Bulk Ceramic [45] < 1700 7 GHz
(12)
25* Dielectric Translation
Displacement Microstrip [46] 40 < Q < 100 3-7 GHz
(625) (100 m)
Mechanical Displacement
MEMS variable capacitor [47] 80 100 < Q < 300 0.5-4 GHz
1 m
Bias Voltage
Varactor [48] 60 30 < Q < 60 10 GHz
(20 V)
*Tunable frequency is proportional to the square root of the tunable permeability or permittivity, and the squares
of the values are shown for comparison with the other methods (shown in parenthesis).
Silicon is a viable candidate for optically controlling where Q is the inverse of the devices loss; Q ( 0 ) and Q (Vmax )
conductivity. In [52], the authors reported on creating planar are at zero bias and at the maximum voltage, respectively; and
antennas on a silicon wafer by precisely forming plasma chan- nr is the tunability from Equation (2). K factors as high as 500
nels in the shape of the desired antenna. Varying the conduc-
have been reported for paraelectric SrTiO3 thin films on
tivity of silicon using an IR LED was reported for a recon-
figurable patch antenna [53], and for a partially adaptive array SrRuO3 conductors.
with broadband monopoles [54].
Dielectric materials with the highest tunability have
paraelectric/ferroelectric transition temperatures that are below
the operating temperature. The magnitude of the tunability
6.2 Tunable Permittivity is directly proportional to the dielectric permittivity, which
approaches 1,000 for materials with the highest tunability [58].
A number of approaches have been explored to achieve The intrinsic dielectric loss increases with frequency, which
economical solutions for a high ability to tune with low losses may be a concern as the operating frequency increases into the
and fast response, for the purpose of controlling microwave mm-wave range [59].
filters and phased arrays. The relative tunability of the per-
mittivity of a system is defined as The dielectric permittivities of thin films and bulk materi-
als are altered through an externally applied electric field. Large
r (0) r ( E ) changes in permittivity under a dc voltage bias have been
nr = , (2) demonstrated in epitaxial paraelectric Ba 0.5Sr0.5TiO3 (BST)
r (0)
thin films that are deposited on single-crystal LaAlO3 and
where r ( 0 ) and r ( E ) are the small-signal relative permit- MgO substrates [42]. The loss of the dielectric film is represented
as an average Q value between the zero-field and high-electric-
tivities without bias and with a bias of strength E [V/cm], field bias. Bulk BST ceramics offer high tunability, and have
respectively. Tunabilities as high as 75% have been reported for higher Q values that BST films. Recent results have shown that
(Ba,Sr)TiO3 thin films at 1500 kV/cm [55]. A high electric high electric fields can be applied to bulk ceramics with
field is required to generate a large tunability; however, the nanometer-sized grains, corresponding to higher high overall
tuning voltage is low ( < 25 V) because the films are in the tunability [43].
nanometer-thickness range. A figure-of-merit for tunable
dielectric devices has been established, which includes the
device loss [56]:
6.3 Tunable Permeability
2
Q ( 0 ) Q (Vmax ) ( nr 1)
K= , (3) Similar to the tunable response in dielectric materials,
nr the magnetic permeability decreases with the application of a
Antenna Measurement
1 Antenna Ranges
An antenna range is a facility where antenna radiation
characteristics are measured. An antenna range includes
the following typical components:
1. A substantial space for hosting the test antenna and the
source antenna
2. A source antenna
3. An antenna positioner
4. A transmitter and receiver system (e.g. a Network
Analyser)
1
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Realized by
a network
analyser
2 Pattern Measurement
2.1 Reciprocity for Antenna Radiation Patterns
7
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
9
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
3 Gain Measurement
3.1 Comparison Method
The gain of an antenna can be measured by the comparison
method using a standard gain antenna whose gain and
reflection coefficient are known accurately. The power
received by the standard gain antenna and the test antenna
are measured, respectively, under the same conditions.
10
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
GT
PT 1 s
G
2
P 1
2 S
S T
PT 1 T
2
GT dB GS dB 10log10 10log10 2
PS 1 S
GT = gain of the test antenna
GS = gain of the standard gain antenna
PT = power received by the test antenna
PS = power received by the standard gain antenna
T = reflection coefficient of the test antenna
S = reflection coefficient of the standard gain antenna
11
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Example 1
A standard gain antenna has a gain of 63 (18 dB). It is used
to measure the gain of a test antenna. The received power
with the standard gain antenna Ps = 3.16 mW (5 dBm) and
with the test antenna PT = 31.6 mW (15 dBm). The standard
antenna has a VSWRS = 1.1 and the test antennas VSWRT =
1.3. Find the gain of the test antenna assuming both
antennas are linearly polarized (LP).
13
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Solution
VSWR S 1 1.1 1
S 0.05
VSWR S 1 1.1 1
VSWR T 1
T 0.13
VSWR T 1
PT 1 T
2
GT dB GS dB 10log10 10log10 2
PS 1 S
31.6 1 0.132
18 dB 10log10 10log10 2
3.16 1 0.05
28.06 dB
14
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
4 Polarization Measurement
4.1 Polarization Pattern Method
This method can be used to measure the AR and the tilt
angle of the polarization ellipse but not the sense of
polarization.
The test antenna is connected as the source antenna while
a linearly polarized antenna such as a dipole antenna is
used to receive the power at different rotation angles.
The square root of the received power plotted against the
rotation angle indicate the AR and title .
15
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
Test antenna
(transmitting) Receiving dipole antenna
(receiving)
16
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
AR = OA/OB
r A2 cos2 B 2 sin 2
A
B
O r
17
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
3 dB AR beamwidth
Radiation pattern
obtained with a
rotating linear
source
AR at
Test antenna
(receiving) Fast-rotating dipole
Rotating Source Method antenna (transmitting)
19
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
20
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
1
VSWR (dimensionless)
1
22
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
23
Get useful study materials from www.rejinpaul.com
www.rejinpaul.com
UNIT V
PROPAGATION OF RADIO WAVES
The three basic types of propagation: Ground wave, space wave and sky wave
propagation.
Sky Wave Propagation: Structure of the ionosphere Effective dielectric constant of
ionized region Mechanism of refraction Refractive index Critical frequency Skip
distance Effect of earths magnetic field Energy loss in the ionosphere due to
collisions Maximum usable frequency Fading and diversity reception.
Space Wave Propagation: Reflection from ground for vertically and horizontally
polarized waves Reflection characteristics of earth Resultant of direct and reflected
ray at the receiver Duct propagation.
Ground Wave Propagation: Attenuation characteristics for ground wave propagation
Calculation of field strength at a distance.
Propagation of Waves
The process of communication involves the transmission of information from one
location to another. As we have seen, modulation is used to encode the information onto
a carrier wave, and may involve analog or digital methods. It is only the characteristics
of the carrier wave which determine how the signal will propagate over any significant
distance. This chapter describes the different ways that electromagnetic waves propagate.
RADIO WAVES
x Electric
Field, E
y Direction of
z Propagation
Magnetic
Field, H
Longitudinal waves:
POLARIZATION
The polarization of an antenna is the orientation of the electric field with respect to
the Earth's surface and is determined by the physical structure of the antenna and
by its orientation
Radio waves from a vertical antenna will usually be vertically polarized.
Radio waves from a horizontal antenna are usually horizontally polarized.
RADIO
WAVES
SPACE GROUND
Ground Wave is a Surface Wave that propagates or travels close to the surface of
the Earth.
Line of Sight (Ground Wave or Direct Wave) is propagation of waves travelling in a
straight line. These waves are deviated (reflected) by obstructions and cannot travel
over the horizon or behind obstacles. Most common direct wave occurs with VHF
modes and higher frequencies. At higher frequencies and in lower levels of the
atmosphere, any obstruction between the transmitting antenna and the receiving
antenna will block the signal, just like the light that the eye senses.
Space Waves: travel directly from an antenna to another without reflection on the
ground. Occurs when both antennas are within line of sight of each another,
distance is longer that line of sight because most space waves bend near the ground
and follow practically a curved path. Antennas must display a very low angle of
emission in order that all the power is radiated in direction of the horizon instead of
escaping in the sky. A high gain and horizontally polarized antenna is thus highly
recommended.
Sky Wave (Skip/ Hop/ Ionospheric Wave) is the propagation of radio waves bent
(refracted) back to the Earth's surface by the ionosphere. HF radio communication
(3 and 30 MHz) is a result of sky wave propagation.
Ground-Wave Propagation
RF Propagation
Ground Wave
Ionospheric
Line of Sight (LOS)
Ground wave propagation follows the curvature of the Earth. Ground waves have carrier
frequencies up to 2 MHz. AM radio is an example of ground wave propagation.
Ionospheric propagation bounces off of the Earth's ionospheric layer in the upper atmosphere. It
is sometimes called double hop propagation. It operates in the frequency range of 30 - 85 MHz.
Because it depends on the Earth's ionosphere, it changes with the weather and time of day. The
signal bounces off of the ionosphere and back to earth. Ham radios operate in this range.
Line of sight propagation transmits exactly in the line of sight. The receive station must be in the
view of the transmit station. It is sometimes called space waves or tropospheric propagation. It is
limited by the curvature of the Earth for ground-based stations (100 km, from horizon to
horizon). Reflected waves can cause problems. Examples of line of sight propagation are: FM
radio, microwave and satellite.
The ground wave used for radio communications signal propagation on the long, and
medium wave bands for local radio communications
Ground wave propagation is particularly important on the LF and MF portion of the radio
spectrum. Ground wave radio propagation is used to provide relatively local radio
communications coverage, especially by radio broadcast stations that require to cover a
particular locality.
Ground wave radio signal propagation is ideal for relatively short distance propagation on these
frequencies during the daytime. Sky-wave ionospheric propagation is not possible during the day
because of the attenuation of the signals on these frequencies caused by the D region in the
ionosphere. In view of this, radio communications stations need to rely on the ground-wave
propagation to achieve their coverage.
A ground wave radio signal is made up from a number of constituents. If the antennas are in the
line of sight then there will be a direct wave as well as a reflected signal. As the names suggest
the direct signal is one that travels directly between the two antenna and is not affected by the
locality. There will also be a reflected signal as the transmission will be reflected by a number of
objects including the earth's surface and any hills, or large buildings. That may be present.
In addition to this there is surface wave. This tends to follow the curvature of the Earth and
enables coverage to be achieved beyond the horizon. It is the sum of all these components that is
known as the ground wave.
Beyond the horizon the direct and reflected waves are blocked by the curvature of the Earth, and
the signal is purely made up from the diffracted surface wave. It is for this reason that surface
wave is commonly called ground wave propagation.
Surface wave
The radio signal spreads out from the transmitter along the surface of the Earth. Instead of just
travelling in a straight line the radio signals tend to follow the curvature of the Earth. This is
because currents are induced in the surface of the earth and this action slows down the wave-
front in this region, causing the wave-front of the radio communications signal to tilt downwards
towards the Earth. With the wave-front tilted in this direction it is able to curve around the Earth
and be received well beyond the horizon.
Effect of frequency
As the wavefront of the ground wave travels along the Earth's surface it is attenuated. The degree
of attenuation is dependent upon a variety of factors. Frequency of the radio signal is one of the
major determining factor as losses rise with increasing frequency. As a result it makes this form
of propagation impracticable above the bottom end of the HF portion of the spectrum (3 MHz).
Typically a signal at 3.0 MHz will suffer an attenuation that may be in the region of 20 to 60 dB
more than one at 0.5 MHz dependent upon a variety of factors in the signal path including the
distance. In view of this it can be seen why even high power HF radio broadcast stations may
only be audible for a few miles from the transmitting site via the ground wave.
The surface wave is also very dependent upon the nature of the ground over which the signal
travels. Ground conductivity, terrain roughness and the dielectric constant all affect the signal
attenuation. In addition to this the ground penetration varies, becoming greater at lower
frequencies, and this means that it is not just the surface conductivity that is of interest. At the
higher frequencies this is not of great importance, but at lower frequencies penetration means
that ground strata down to 100 metres may have an effect.
Despite all these variables, it is found that terrain with good conductivity gives the best result.
Thus soil type and the moisture content are of importance. Salty sea water is the best, and rich
agricultural, or marshy land is also good. Dry sandy terrain and city centres are by far the worst.
This means sea paths are optimum, although even these are subject to variations due to the
roughness of the sea, resulting on path losses being slightly dependent upon the weather! It
should also be noted that in view of the fact that signal penetration has an effect, the water table
may have an effect dependent upon the frequency in use.
Effect of polarisation
The type of antenna has a major effect. Vertical polarisation is subject to considerably less
attenuation than horizontally polarised signals. In some cases the difference can amount to
several tens of decibels. It is for this reason that medium wave broadcast stations use vertical
antennas, even if they have to be made physically short by adding inductive loading. Ships
making use of the MF marine bands often use inverted L antennas as these are able to radiate a
significant proportion of the signal that is vertically polarised.
At distances that are typically towards the edge of the ground wave coverage area, some sky-
wave signal may also be present, especially at night when the D layer attenuation is reduced.
This may serve to reinforce or cancel the overall signal resulting in figures that will differ from
those that may be expected.
Space Waves, also known as direct waves, are radio waves that travel directly from the
transmitting antenna to the receiving antenna. In order for this to occur, the two antennas must be
able to see each other; that is there must be a line of sight path between them. The diagram on
the next page shows a typical line of sight. The maximum line of sight distance between two
antennas depends on the height of each antenna. If the heights are measured in feet, the
maximum line of sight, in miles, is given by:
Because a typical transmission path is filled with buildings, hills and other obstacles, it is
possible for radio waves to be reflected by these obstacles, resulting in radio waves that arrive at
the receive antenna from several different directions. Because the length of each path is different,
the waves will not arrive in phase. They may reinforce each other or cancel each other,
depending on the phase differences. This situation is known as multipath propagation. It can
cause major distortion to certain types of signals. Ghost images seen on broadcast TV signals are
the result of multipath one picture arrives slightly later than the other and is shifted in position
on the screen. Multipath is very troublesome for mobile communications. When the transmitter
and/or receiver are in motion, the path lengths are continuously changing and the signal
fluctuates wildly in amplitude. For this reason, NBFM is used almost exclusively for mobile
communications. Amplitude variations caused by multipath that make AM unreadable are
eliminated by the limiter stage in an NBFM receiver.
has a tremendous line of sight from its vantage point in space and many ground stations can
communicate through a single satellite.
Sky Waves
Radio waves in the LF and MF ranges may also propagate as ground waves, but
suffer significant losses, or are attenuated, particularly at higher frequencies. But
as the ground wave mode fades out, a new mode develops: the sky wave. Sky waves are
reflections from the ionosphere. While the wave is in the ionosphere, it is strongly
bent, or refracted, ultimately back to the ground. From a long distance away this
appears as a reflection. Long ranges are possible in this mode also, up to hundreds
of miles. Sky waves in this frequency band are usually only possible at night, when
the concentration of ions is not too great since the ionosphere also tends to attenuate
the signal. However, at night, there are just enough ions to reflect the wave but not
reduce its power too much.
Figure 14
The HF band operates almost exclusively with sky waves. The higher frequencies have less
attenuation and less refraction in the ionosphere as compared to MF. At the high end, the waves
completely penetrate the ionosphere and become space waves. At the low end, they are always
reflected. The HF band operates with both these effects almost all of the time. The characteristics
of the sky wave propagation depend on the conditions in the ionosphere which in turn are
dependent on the activity of the sun. The ionosphere has several well-defined regions in altitude.
Figure 15
D-region: about 75-95 km. Relatively weak ionization. Responsible for strong absorption of MF
during daylight E-region: 95-150 km. An important player in ionospheric scatter of VHF. F-
region: 150-400 km. Has separate F1 and F2 layers during the day. The strongest concentration
of ions. Responsible for reflection of HF radio waves. Since the propagation characteristics
depend on frequency, several key frequencies can de defined: Critical frequency: The minimum
frequency that will penetrate the ionosphere at vertical incidence. The critical frequency
increases during the daylight and decrease at night. At other angles, the wave will be reflected
back. At frequencies above the critical frequency, some range of waves from vertical incidence
and down will become space waves. This will cause a gap in coverage on the ground known as a
skip zone. In figure xx, the skip zone extends to about 1400 miles. The transmitted frequency
was 5 MHz and the critical frequency was 3 MHz in this example. Maximum Useable Frequency
(MUF): defined for two stations. The maximum frequency that will reflect back to the receiving
station from the transmitter. Beyond the MUF, the wave will become a space wave. At MUF the
skip zone extends to just short of the receiver. In figure xx, the MUF for a receiver at 1400 miles
is 5 MHz. Lowest Useable Frequency (LUF): again defined for two stations. At low frequencies,
the signal will be attenuated before it can be reflected. The LUF increases with sunlight and is a
maximum near noon. Optimum Frequency for Traffic (OFT): for two stations, taking into
account the exact conditions in the ionosphere, there will be the perfect frequency that gives the
strongest signal. This can be predicted by powerful modeling programs and is the best guarantee
of success in HF. The diurnal variation if HF propagation is characterized a simple rule-of-
thumb: the frequency follows the sun. At noon, the OFT is generally higher than at night.
Line of Sight
In the VHF band and up, the propagation tends to straighten out into line-of-sight(LOS)
waves. However the frequency is still low enough for some significant effects.
1. Ionospheric scatter. The signal is reflected by the E-region and scattered in all directions.
Some energy makes it back to the earth's surface. This seems to be most effective in the
range of 600-1000 miles.
Figure 16
1. Tropospheric scatter. Again, the wave is scattered, but this time, by the air itself. This can
be visualized like light scattering from fog. This is a strong function of the weather but
can produce good performance at ranges under 400 miles.
Figure 17
1. Tropospheric ducting. The wave travels slower in cold dense air than in warm air.
Whenever inversion conditions exist, the wave is naturally bent back to the ground.
When the refraction matches the curvature of the earth, long ranges can be achieved. This
ducting occurs to some extend always and improves the range over true the line-of-sight
by about 10 %.
1. Diffraction. When the wave is block by a large object, like a mountain, is can diffract
around the object and give coverage where no line-of-sight exists.
Ionospheric Storms: Solar activity such as flares and coronal mass ejections
produce large electromagnetic radiation incidents upon the earth and leads to
Solar Cycle
Every 11 years the sun undergoes a period of activity called the "solar maximum",
followed by a period of quiet called the "solar minimum". During the solar
maximum there are many sunspots, solar flares, and coronal mass ejections, all of
which can affect communications and weather here on Earth.
The Sun goes through a periodic rise and fall in activity which affects HF
communications; solar cycles vary in length from 9 to 14 years. At solar minimum,
only the lower frequencies of the HF band will be supported by the ionosphere,
while at solar maximum the higher frequencies will successfully propagate, figure
1.4. This is because there is more radiation being emitted from the Sun at solar
maximum, producing more electrons in the ionosphere which allows the use of
higher frequencies.
One way we track solar activity is by observing sunspots. Sunspots are relatively
cool areas that appear as dark blemishes on the face of the sun. They are formed
when magnetic field lines just below the sun's surface are twisted and poke though
the solar photosphere. The twisted magnetic field above sunspots are sites where
solar flares are observed to occur, and we are now beginning to understand the
connection between solar flares and sunspots.
During solar maximum there are many sunspots, and during solar minimum there
are few. The plot at right shows the number of sunspots observed during the last
two solar cycles. The last maximum occurred around 1989, and the next is predicted
to fall in the year 2000. This plot is updated monthly. Click here for a plot of
sunspot numbers from the year 1749 through the present.
The Earth is affected by both solar flares and sunspots. Solar flares emit high-speed
particles which cause auroras, known in the northern hemisphere as Northern
Lights. The image shown here is a real-time satellite image of the Earth's auroral
region above the North Pole. From the ground auroras appear as shimmering
curtains of red and green light in the sky.
Particles from solar flares can also disrupt radio communication, and the radiation
from the flares can give passengers in airplanes a dose of radiation equivalent to a
medical X-ray. Sunspots may have a long-term connection with the Earth's climate.
Scientists are currently debating whether ice ages on Earth are related to the Sun
having fewer sunspots than usual.
An HF signal transmitted from the earth may travel some way through the
ionosphere before being "bent" back down towards the ground. This occurs due to
the interaction between the HF signal and electrically charged particles in the
ionosphere. The signal can then "bounce" off the ground back into the ionosphere,
return to the earth again, and so on. The distance a given HF signal will travel
depends on the frequency, transmitter power, take-off angle relative to the ground
and the state of the ionosphere through which it is travelling
For any given distance and time, there will be a certain range of HF frequencies that
are most likely to provide successful communications; frequencies outside that
range will work poorly or not at all. Simply increasing the power of an HF signal
will not help if the frequency is too high for the distance required. Increasing the
power may help if the frequency is too low, but using a higher, more suitable
frequency is the best option. The highest frequency which may be used for reliable
HF communications is known as the Maximum Usable Frequency (MUF).
What Kind of Disturbances Can Degrade HF Communications?
Short-Wave Fadeouts - short lived (up to two hours) disturbances, in which solar
flare activity results in the absorption of lower frequency HF signals. These will only
affect signals passing through the daylight ionosphere
Ionospheric Storms - large scale changes in the chemical composition of the ionosphere
resulting in changes to the MUF. Decreased MUFs restrict the frequencies available for use
over a given distance. Ionospheric storms normally last for one to two day
Critical Frequency:
The highest frequency that will be returned to the earth when transmitted
vertically under given ionospheric conditions
Critical Angle:
The highest angle with respect to a vertical line at which a radio wave of a
specified frequency can be propagated and still be returned to the earth from
the ionosphere
The highest frequency that is returned to the earth from the ionosphere
between two specific points on earth
Optimum Working frequency:
The frequency that provides for the most consistent communication path via
sky waves
Tropospheric Scattering
Signals are aimed at the troposphere rather than the ionosphere
350 Mhz to 10GHz for paths up to 400 mi
Received signal = 10-6 th of the transmitted power
Fading a problem
Satellite communicatons
Intelsat III
LEO
GPS Systems
FDMA
TDMA
Time division multiplex access
Single satellite to service multiple earth stations simultaneously
All stations use the same carrier but transmit one or more traffic bursts in
nonoverlapping time frames
TDMA Advantages
CDMA
Code division multiple access
Allows use of one carrier
Each station uses a different binary sequence to modulate the carrier
Control uses a correlator that separates and distributes the signals to appropriate
downlink
VSAT
Amateur Satellites
OSCAR: Orbiting Satellite Carrying
Amateur Radio
Used on VHF (mainly)
Directional ant.s
are a must!
QUESTION BANK
PART-A ( 2 marks)
Waves that arrive at the receiver after reflection in the ionosphere is called sky wave.
Waves that arrive at the receiver after reflection from the troposphere region is called
Tropospheric wave.(ie 10 Km from Earth surface).
Waves propagated over other paths near the earth surface is called ground wave propagation.
i. Space wave
It is made up of direct wave and ground reflected wave. Also includes the portion of energy
received as a result of diffraction around the earth surface and the reflection from the upper
atmosphere.
Wave that is guided along the earth s surface like an EM wave is guided by a
transmission is called surface wave. Attenuation of this wave is directly affected by the
constant of earth along which it travels.
Variation of signal strength occur on line of sight paths as a result of the atmospheric
conditions and it is called .It can not be predicted properly.
Inverse bending may transform line of sight path into an obstructed one. Multi path fading is
caused by interference between the direct and ground reflected waves as well as interference
between two are more paths in the atmosphere.
To minimize the fading and to avoid the multi path interference the technique used are
diversity reception. It is obtained by two ways.
This method exploits the fact that signals received at different locations do not fade together.
It requires antenna spaced at least 100 l apart are referred and the antenna which high signal
strength at the moment dominates.
This method takes advantage of the fact that signals of slightly different frequencies do not
fade synchronously. This fact is utilized to minimize fading in radio telegraph circuits.
It is used in normally in microwave links, and it is found that signal transmitted over the
same path in two polarizations have independent fading patterns. In broad band dish
antenna system, Polarization diversity combined with frequency diversity reception
achieve excellent results.
Due to the earth s magnetic fields, the ionosphere medium becomes anisotropic and the
incident plane wave entering the ionosphere will split into ordinary and extra ordinary
waves/modes.
When these modes re-emerge from the ionosphere they recombine into a single plane
wave again.
Finally the plane of polarization will usually have changed, this phenomenon is known as
Faraday s rotation.
15. What are the factors that affect the propagation of radio waves?
i. Curvature of earth.
Frequency whose period is equal to the period of an electron in its orbit under the influence
of the earths magnetic flux density B.
For any layer , the highest frequency that will be reflected back for vertical incidence is fcr
=9
max
The phenomenon of splitting the wave into two different components (ordinary and extra-
ordinary) by the earths magnetic field is called Magneto-Ions Splitting.
The lowest useful HF for a given distance and transmitter power is defined as the lowest
frequency that will give satisfactory reception for that distance and power.
It depends on
ii. Absorption character of ionosphere for the paths between transmitter and receiver.
iii. The required field strength which in turn depends upon the radio noise at the receiving
location and type of service involved .
---- ------------------------------------
N=er
The maximum Frequency that can be reflected back for a given distance of transmission is
called the maximum usable frequency (MUF) for that distance.
MUF = fcr_secFi
The distance with in which a signal of given frequency fails to be reflected back is the
skip distance for that frequency.The higher the frequency the greater the skip distance.
Optimum frequency for transmitting between any two points is therefore selected as some
frequency lying between about 50 and 85 percent of the predicted maximum usable frequency
between those points.
h= h0/ 1-(fc/f)
PART B
Propagation? (10)
5. Describe the troposphere and explain how ducts can be used for
PART A( 2 Marks)
PART B
1. Derive the expression for the radiated field from a short dipole? (16)
2. Starting from first principles obtain the expression for the power
radiated by a half wave dipole? (16)
3. Derive the expression for power radiated and find the radiation
resistance of a half wave dipole? (16)
4. Derive the radian resistance, Directivity and effective aperture of a
half wave dipole? (10)
5. Derive the fields radiated from a quarter wave monopole antenna? (8)
6. Find the radiation resistance of elementary dipole with linear
current distribution? (8)
7. Derive the radiation resistance, Directivity and effective
aperture of a hertzian dipole? (10)
UNIT II
PART A( 2 Marks)
PART B
and also find the maximum direction minimum direction and half
power point direction? (16)
8. Write short notes a radiation resistance? (8)
9. Calculate the maximum effective aperture of a /2 antenna? (8)
10. Derive the maxima directions, minima directions, and half
power point direction for an array of two point sources with equal
amplitude and opposite phase? (16)
11. Explain the various types of amplitude distributions in details? (16)
UNIT III
PART A( 2 Marks)
PART B
UNIT IV
PART A( 2 Marks)
PART B
UNIT V
WAVE PROPAGATION
PART A ( 2 Marks)
PART B