Active Filter Design Using Operational

Transconductance Amplifiers: A Tutorial

Randall L. Geiger and Edgar Sânchez-Sinencio
Abstract are discussed. It is shown that these structures offer
Basic properties of the Operational Transconductance Amplifier (OTA)
improvements in design simplicity and program-
are discussed. Applications of the OTA in voltage-controlled amplifiers,
filters, and impedances are presented. A versatile family of voltage-
mability when compared to op amp based structures
controlled filter sections suitable for systematic design requirements is de­ as well as reduced component count.
scribed. The total number of components used in these circuits is small, Many of the basic OTA based structures use only
and the design equations and voltage-control characteristics are attractive. OTAs and capacitors and, hence, are attractive for in­
Limitations as well as practical considerations of OTA based filters using
tegration. Component count of these structures is
commercially available bipolar OTAs are discussed. Applications of OTAs
in continuous-time monolithic filters are considered. often very low (e.g., second-order biquadratic filters
can be constructed with two OTAs and two capaci­
tors) when compared to VCVS designs. Convenient
Introduction internal or external voltage or current control of filter
The conventional operational amplifier (op amp) is characteristics is attainable with these designs. They
used as the active device in the vast majority of the are attractive for frequency referenced (e.g., master/
active filter literature. For design purposes, the as­ slave) applications. Several groups have recently uti­
sumption that the op amp is ideal (Av = o°, Rin = oo, lized OTAs in continuous-time monolithic filter struc­
Ro = 0) is generally made, and large amounts of tures [28]-[40].
feedback are used to make the filter gain essentially
From a practical viewpoint, the high-frequency
independent of the gain of the op amp. A host of
performance of discrete bipolar OTAs, such as the
practical filter designs have evolved following this ap­
CA 3080, is quite good. The transconductance gain,
proach. It has also become apparent, however, that
g , can be varied over several decades by adjusting an
operational amplifier limitations preclude the use of
m

external dc bias current, J . The major limitation of

these filters at high frequencies, attempts to integrate
A B C

existing OTAs is the restricted differential input volt­

these filters have been unsuccessful (with the excep­
age swing required to maintain linearity [5]. For the
tion of a few nondemanding applications), and con­
CA 3080, it is limited to about 30 mV p-p to maintain a
venient voltage or current control schemes for exter­
reasonable degree of linearity.
nally adjusting the filter characteristics do not exist.
Although feedback structures in which the sen­
With the realization that the BJT and MOSFET are
sitivity of the filter parameters are reduced (as is the
inherently current and transconductance amplifiers,
goal in op amp based filter design) will be discussed,
respectively, the following question naturally arises.
major emphasis will be placed upon those structures
Can any improvements in filter characteristics, per­
in which the standard filter parameters of interest are
formance, or flexibility be obtained by using one of
directly proportional to g of the OTA. Thus, the g
the other basic types of amplifiers (e.g., transconduc­
m m

will be a design parameter much as are resistors and

tance, current, or transresistance) in place of a voltage
capacitors. Since the transconductance gain of the
amplifier (or specifically the operational amplifier) as
OTA is assumed proportional to an external dc bias
the basic active device in a filter structure?
current, external control of the filter parameters via
This question is currently difficult to answer for the bias current can be obtained.
two reasons. First, there is a near void in the literature
Most existing work on OTA based filter design ap­
of active filter structures employing the alternative
proached the problem by either concentrating upon
amplifier types. Second, the evolution of good inte­
applying feedback to make the filter characteristics
grated transresistance, transconductance, and cur­
independent of the transconductance gain or modify­
rent amplifiers has not kept pace with that of the
ing existing op amp structures by the inclusion of
voltage amplifiers, although a few devices in these al­
some additional passive components and OTAs. In ei­
ternate categories are commercially available (e.g.,
ther case, the circuits were typically component in­
transconductance amplifiers such as the CA 3080 and
tense and cumbersome to tune. Some of the earlier
LM 13600 and transresistance amplifiers such as the
works are listed in the Refs. [6]-[16]. Some of the
LM 3900) [ l ] - [ 5 ] . Comparisons of some characteris­
most practical circuits can be found in the manufac­
tics of these amplifiers were recently discussed by
turer's application notes [3]-[5].
Brugger et al. [40].
In this paper, basic first- and second-order struc­ OTA Model
tures using the transconductance amplifier (often The symbol used for the OTA is shown in Fig. 1,
termed the operational transconductance amplifier: OTA) along with the ideal small signal equivalent circuit.
20 8755-3996/85/0300-0020$!.00 © 1985 IEEE IEEE CIRCUITS A N D DEVICES MAGAZINE
 I ABC

Z 0" L R

(a) (b)

Fig. 1 OTA. (a) Symbol, (b) Equivalent circuit of ideal OTA.

The transconductance gain, g , is assumed propor­ m

tional* to 7 . The proportionality constant h is de­

ABC

pendent upon temperature, device geometry, and the

process [2].
z o'°
gm = W B C A (1) (d)

The output current is given by (c)

h = g (v
m + - ν-) (2)

As shown in the model, the input and output imped­ Re

ances in the model assume ideal values of infinity.
Current control of the transconductance gain can be
directly obtained with control of J . Since tech­ 1—
i
A B C
8m -oVo
niques abound for creating a current proportional to a
given voltage, voltage control of the OTA gain can T Ïabc
also be attained through the J input. Throughout
A B C
Vo.Q»(R|»*t>
this paper, when reference is made to either the cur­ V| i+gm R ι Vi i+g R, m

rent or voltage controllability of OTA based circuits, it 7

Ri+R 2 7 . R|+ 2 R

0
is assumed to be attained via control of g by 7 . m ABC *° l+QmRi
(e) (f)

Basic OTA Building Blocks

Some of the basic OTA building blocks [6] are intro­
duced in this section. A brief discussion about these
circuits follows.
Voltage amplifiers using OTAs are shown in Fig. 2,
along with voltage gain and output impedance ex­
pressions. The basic inverting and noninverting con­
figurations of Figs. 2a and 2b have a voltage gain
directly proportional to g which makes current (vol­
m /

tage) control of the gain via J straightforward. Fur­

A B C

thermore, observe that a differential amplifier can be

easily obtained by using both input terminals of the
OTA in Figs. 2a or 2b. The major limitation of these
circuits is the relatively high output impedance.

*A linear dependence on bias current is typically obtained for

bipolar OTAs and MOS configurations operating in weak inversion. Fig. 2 Voltage amplifiers, (a) Basic inverting, (b) Basic noninverting.
MOS structures operating in the saturation region typically exhibit (c) Feedback amplifier, (d) Noninverting feedback amplifier, (e) Buf­
a quadratic dependence on 7 . A B C
fered amplifier, (f) Buffered VCVC feedback, (g) All OTA amplifiers.

MARCH 1985 21
 A voltage buffer, such as used in Figs. 2c and 2d, is tained with either g or g . The total adjustment
ml ml

often useful for reducing output impedance.* Al­ range of the gain of this structure is double (in dB)
though the gain characteristics of these circuits are that attainable with the single OTA structures consid­
ideally identical, the performance of the two circuits ered in Figs. 2a and 2b. Furthermore, if both OTAs
is not the same. The performance differences are due are in the same chip, the variations with temperature
to differences in the effects of parasitics in the cir­ of the g 's are cancelled.
m

cuits. Specifically, the parasitic output capacitance of Several standard controlled impedance elements
the OTA in Fig. 2c, along with instrumentation para­ are shown in Fig. 3, along with the input impedance
sitics, parallel the resistor R in discrete component
L expression. These controlled impedances can be used
structures, thus causing a roll-off in the frequency re­ in place of passive counterparts (when applicable) in
sponse of the circuits. In the circuit of Fig. 2d, the active RC structures to attain voltage control of the
parasitic output capacitance of the OTA is connected filter characteristics or as building blocks in OTA
across the null port of an op amp and thus has negli­ structures.
gible effects when the op amp functions properly. The circuit of Fig. 3a is a grounded Voltage Variable
Likewise, instrumentation parasitics will typically ap­ Resistor ( W R ) . The circuit of Fig. 3b behaves as a
pear at the low impedance output of the op amp and floating W R , provided g and g are matched. If a
ml m2

thus not have a major effect on the performance. mismatch occurs, the structure can be modeled with
As with conventional amplifier design using resis­ a floating W R between terminals 1 and 2 of value g mV

tors and op amp's, the amplifier bandwidth of these along with a voltage dependent current source of
structures warrants consideration. For the circuits of value (g -g )
ml Vi driving node 2.
m2

Figs. 2c and 2d, the major factor limiting the band­ The circuit of Fig. 3c acts as a scaled W R . Higher
width is generally the finite gain bandwidth product impedances are possible than with the simple struc­
of the op amps. If the op amps are modeled by the ture of Fig. 3a, at the expense of the additional re­
popular single-pole roll-off model, A(s) = GB/s, and sistors.
the OTAs are assumed ideal, it follows that the band­ A voltage variable impedance inverter is shown in
width of the circuits of Fig. 2c and Fig. 2d is GB, Fig. 3d. Note the doubling of the adjustment range of
independent of the voltage gain of the amplifier. this circuit, as with the amplifier of Fig. 2g. Of special
This can be contrasted to the bandwidths of GBl Κ interest is the case where this circuit is loaded with a
and GB/1 + Κ for the basic single op amp noninvert­ capacitor. In this case, a synthetic inductor is ob­
ing and inverting amplifiers of gains Κ and -K, tained. The doubling of the adjustment range is par­
respectively. ticularly attractive for the synthetic inductor since
Note that the circuits of Figs. 2a and 2b differ only cutoff frequencies in active filter structures generally
in the labeling of the " + " and " —" terminals. In all involve inductor values raised to the 1/2 power. By
circuits presented in this paper, interchanging the making g = g and adjusting both simultaneously,
ml m2

" + " and " - " terminals of the OTA will result only in first-order rather than quadratic control of cutoff fre­
changing the sign of the g coefficient in any equation
m quencies is possible.
derived for the original circuit. Henceforth, it will be A floating impedance inverter is shown in Fig. 3e.
the reader's responsibility to determine when such an Note that it is necessary to match g and g for m2 m3

interchange provides a useful circuit. proper operation. The circuit of Fig. 3f serves as an
The circuits of Figs. 2e and 2f are feedback struc­ impedance multiplier. That of Fig. 3g behaves as a
tures. The circuit of Fig. 2e offers gains that can be super inductor and that of Fig. 3h as a FDNR.
continuously adjusted between positive and negative
values with the parameter g . By interchanging the
m
First-Order Filter Structures
+ and - terminals of the OTA, very large gains can be A voltage variable integrator structure with a differ­
obtained as g R approaches 1 (as Z approaches in­
m 1 0
ential input is shown in Fig. 4a. The integrator serves
finity). Gain is nonlinearly related to g . Control
m
as the basic building block in many filter structures.
range via g is reduced in these structures when com­
m
Two different lossy integrators (first-order lowpass fil­
pared to the amplifiers of Figs. 2a and 2b. If compo­ ters) are shown in Figs. 4b and 4c. The circuit of Fig.
nents are sized fitly, the gain of these structures can 4b has a loss that is fixed by the RC product and a
be made essentially independent of g (as in the con­
m
gain controllable by g . The circuit of Fig. 4c offers
m

ventional op amp inverting and noninverting config­ considerably more flexibility. The pole frequency can
urations) and the output impedances can be made be adjusted by g (interchanging the input terminals
m2

reasonably small. of OTA 2 actually allows the pole to enter the right
The amplifier of Fig. 2g is attractive since it contains half plane), and the dc gain can be subsequently ad­
no passive components. Gain adjustment can be at- justed by g . It should be noted that the structure of
ml

Fig. 4c contains no resistors and can be obtained from

*An alternative to the op amp buffer is the Darlington pair, the circuit of Fig. 4b by replacing the resistor R with
which is included in the same package in some commercial OTAs the controlled impedance of Fig. 3a. Another lossy
(e.g., CA 3094, LM 13600). integrator without adjustable gain but with adjustable
22 IEEE CIRCUITS AND DEVICES MAGAZINE
 9m
8m
gm
(α)
Zin©
(b)

•'HE
ι χι
*ΙΪΙ|

(c)
+

(d)

7 Z L m39m4
g

1 ΓΠ|

I Z :

ni| m L
r 3 m2
3 m3
'iti4

z L
9m h
2

(f)

Π z -
iB
s C| C
2 2 Z

ι ^SUPER
«n%2 C | C g
2 g
*
INDUCTOR
~2Τ Χ

(g) (h)

Fig. 3 Controlled impedance elements. (a) Single-ended voltage variable resistor (VVR). (b) Floating VVR. (c) Scaled VVR. id) Voltage variable
impedance inverter, (e) Voltage variable floating impedance, (f) Impedance multiplier, (g) Super inductor, (h) FDNR.

pole location and a very simple structure is shown in

Fig. 5a.
ViO-
When designing cascaded integrator-based filter
Vi, V
structures, it may be the case that the input imped­ 3 0

ance to some stages is not infinite. If that is the case, C Î fRo

a unity gain buffer would be required for coup­ is An.
V|,-V| sC
ling, since the output impedances of all integrators in
2

(o) Vq 9mR
Fig. 4 are nonzero. Note, however, that no buffer is
Vj " s R C + I
needed for the cascade of any of the integrators of (b)
Fig. 4, since the input impedance to each circuit is
ideally infinite.
First-order filters can be readily built using OTAs. Vi
Considerable flexibility in controlling those specific V 0 Qm,

filter characteristics that are usually of interest is pos­ V| sC + g m 2

sible with these structures. Several first-order voltage-

controlled filters are shown in Fig. 5, and a functional (c)
plot of the transfer characteristics as a function of the
transconductance gains is shown in Fig. 6. Fig. 4 Integrator structures, (a) Simple, (b) Lossy, (c) Adjustable.

MARCH 1985 23
 c 2

VjO-

»m 2

τ a Qm\ + s C 2
(h) V| 8 s(C +C )+g +g
| 2 m m 2

(i)

Fi'g. 5 First-order voltage-controlled filters, (a) Lowpass, fixed dc gain pole adjustable, (b) Lowpass fixed pole, adjustable dc gain, (c) Highpass, fixed high-
frequency gain, adjustable pole, (d) Shelving equalizer, fixed high-frequency gain, fixed pole, adjustable zero, (e) Shelving equalizer, fixed high-
frequency gain, fixed zero, adjustable pole, (f) Lowpass filter adjustable pole and zero, fixed ration, (g) Shelving equalizer, independently adjustable
pole and zero, (h) Lowpass or highpass filter, adjustable zero and pole, fixed ratio or independent adjustment, (i) Phase shifter, adjustable with g . m

24 IEEE CIRCUITS AND DEVICES MAGAZINE

 The 3dB cutoff frequency of the lowpass filter of equalizer. g can be used to adjust the phase shift.
ml

Fig. 5a is given by the expression For monolithic applications, the resistor R can be re­
placed with a third OTA, using the configuration of
hdB = Fig. 3a.

Linear adjustment of / with g is attainable with

Second-Order Structures
3 d B m

this circuit while maintaining a unity dc frequency

Second-order filter structures find widespread ap­
gain. The structure of Fig. 5b has a fixed pole location
plications directly and in the design of higher-order
and adjustable dc gain with the transconductance
filters. Although the emergence of practical voltage
gain g . If the resistor in this circuit is replaced with
or current-controlled first-order filters and ampli­
m

the controlled resistor of Fig. 3a, the circuit would

fiers has been slow, even fewer techniques exist for
have independently adjustable gain and break fre­
the design of controlled second- and higher-order
quency. The highpass structure of Fig. 5c also has a
structures. Switched-capacitor techniques have been
3dB cutoff frequency given by
successfully used to build voltage-controlled filter
structures by building a voltage-controlled oscillator
and using the output as the required clock for the
switching of the capacitors. Although useful in some
It can be observed that the characteristic networks applications, these structures are not continuous time
for the lowpass and highpass structures of Figs. 5a in nature, have limited dynamic range, and are lim­
and 5c are identical, and thus they have the same pole ited to reasonably low-frequency applications. Con­
structures. They differ only in where the excitation is centration here will be on continuous-time voltage-
applied. controlled structures.
The circuits of Figs. 5d and 5e act as shelving equal­ One common requirement in the design of voltage-
izers. The response of both circuits can be continu­ controlled filter structures is that the filter character­
ously changed from lowpass to allpass to highpass by istics be adjusted in a manner that essentially results
adjusting g , as can be seen from Fig. 6. The basic
m in frequency scaling. In all-pole applications, such
difference in the two circuits is that the former has a as the lowpass Butterworth and Chebyschev case,
fixed pole and adjustable zero, whereas the circuit of as well as the bandpass and highpass versions of
Fig. 5e has an adjustable pole and fixed zero. As for these approximations, the frequency scaling is tanta­
the circuit of Fig. 5b, additional flexibility can be ob­ mount to moving all poles a prescribed distance in a
tained if the grounded resistor in the circuit of Fig. 5d constant- Q manner. Those familiar with active filter
is replaced with the controlled resistor of Fig. 3a. structures will recall that pole movement in second-
The circuit of Fig. 5f acts as a lowpass filter with order structures through the adjustment of a single
high-frequency gain determined by the C :C ratio. 1 2
component is always on a circular path (constant ω ) 0

Both the pole and zero in this circuit are adjustable or on a straight line (constant bandwidth) parallel to
through the parameter g , but the ratio is held con­
m
the imaginary axis in the s-plane. The challenges as­
stant. This preserves the shape in the transfer charac­ sociated with constant-Q pole adjustment through
teristics and thus represents only a frequency shift in the simultaneous tuning of two or more components
the response, as shown in Fig. 6f. should be obvious.
The circuit of Fig. 5g utilizes an additional OTA and A seemingly more difficult situation exists when
offers considerable flexibility. If either g or g is ml m2
considering the design of the popular elliptic filters.
fixed, the circuit behaves much like the shelving To maintain the elliptic characteristics as the cutoff
equalizers discussed above. If g and g are ad­ ml m2
frequency is changed, all poles and all zeros of the
justed simultaneously, then a fixed pole-zero ratio approximating function must be moved simultane­
and, hence, shape preserving response is possible. In ously and with the appropriate ratio in a constant- Q
this case, the circuit can be lowpass, allpass, or high- manner.
pass, depending upon the g : g ratio. If the " + " and
ml m2 A group of second-order voltage-controlled filter
" - " terminals of g are interchanged and the trans­
ml structures are discussed in this section. Circuits with
conductance gains are adjusted so that g = g the ml m2f constant-Q pole adjustment, circuits with constant
circuit behaves as a phase equalizer. bandwidth ω adjustment, and circuits with indepen­
0

The circuit of Fig. 5h also preserves the shape of the dent pole and zero adjustment are presented. Some
transfer function, provided g and g are adjusted ml m2
circuits with simultaneous constant-Q adjustment of
in such a manner that their ratio remains constant. In both the poles and zeros are also presented along
this case, the shape of the response is determined by with a general biquadratic structure. These structures
the gmi'.gmi d C :C ratio. Depending upon these
a R
X 2
have immediate applications in voltage-controlled
ratios, the response is either lowpass or highpass in Butterworth, Chebyschev, and Elliptic designs.
nature, as indicated in Fig. 6h. A simple second-order filter structure is shown in
If g R = 1, the circuit of Fig. 5i behaves as a phase
m2
Fig. 7a [17], [19]. This structure is canonical in the

MARCH 1985 25
Fig. 6 Transfer characteristics for first-order structures of Fig. 5. (a) Circuit of Fig. 5a. (b) Circuit of Fig. 5b. (c) Circuit of Fig. 5c. (d) Circuit of Fig. 5d.
(e) Circuit of Fig. 5e. (f) Circuit of Fig. 5f. (g) Circuit of Fig. 5g. (h) Circuit of Fig. 5h. (i) Circuit of Fig. 5i.

Circuit Type Input Conditions Transfer Function Κ gml = gml = gm

«Ό Q

Vi = VA
w Adjustable gmlgml
0
V and V
B c Grounded
Lowpass S C ! C + SCig
2 2 m2 + gmlgml y Q c 2

Vi = v B

w Adjustable
0 V and V
A G Grounded SCigml gm
Bandpass S CiC
2 2 + SCig m2 + gmlgml y c , c 2
c,
Vi = V c

w Adjustable
0

Highpass
V and Vß Grounded
A S CiC
2 2 gm IL
S C i C + SCig
2 2 m2 + gmlgml y c , c 2 c,
Vi = v = v
A c

w Adjustable V Grounded S C ! C + gmlgml gm

0 B
2 2

Notch
S CiC
2 2 + SCxgml + gmlgml Ci

Table Transfer functions for biquadratic structure of Fig. 7a.

26 IEEE CIRCUITS A N D DEVICES MAGAZINE

 sense that only four components are needed to obtain
second-order transfer functions. The output voltage,
V , is given by the expression
0

s^QQVc + sC V lgm2 B +
(4)
s ^ Q + sC^ m 2 + gg ml m2

The transfer function for the specific excitations at

V V , d V are listed in the Table. Note that for
Af b a R c

8mi 8mi 8m> the lowpass, bandpass, highpass,

= =

and notch versions of this circuit all behave as ω ad­ 0

justable circuits with fixed pole Q's. The pole Q's are
determined by the capacitor ratio, which can be accu­
rately maintained in monolithic designs. It is interest­
ing to note that the zeros of the notch circuit also
move in a constant-Q (i.e. along the jw axis) manner
with the poles, as g is adjusted. m

Occasionally, it is desirable to have circuits in which

ω and Q of the poles can be independently adjusted.
0

Two circuits with these characteristics are shown in

-ο Vo2 Fig. 7b [18], [19], [24] and Fig. 7c [18], [24]. The out­
put voltages for these circuits are, respectively,
± c 2

*>v r
_ sCCV
2 x 2 c + sC V lgm2 B +
(5)
8τη\8ηΰΥA
s C,C
2 2 + sC^g^R + g
and
_ sCCV
2 + sC g V
x 2 c x m2 B + gg Vml m2 A

(6)
s ^ Q + sg^C, +g
The circuits of Figs. 7b and 7c can be also used to
implement lowpass, bandpass, highpass, and notch
transfer functions through the proper selection of the
inputs as for the circuit of Fig. 7a.
In the circuit of Fig. 7b, the expressions for ω and 0

Q of the poles of the circuit are given by

8m\8m2
(7)
QC 2

and
ι
(c)
Q = g R (8)
Sml m3 Vc l g

The poles can be moved in a constant- Q manner if g m3

is fixed, and if g = g = g is adjusted; whereas

ml m2 m

movement in a constant ω manner is attainable if g 0 m3

is adjusted when g and g remain constant. The ml m2

independent adjustment of ω and Q is apparent. 0

For the circuit of Fig. 7c, the expressions for ω and 0

Q of the poles become

8ml8m2
(9)
QC 2

ι / Q \J gmlgm2
n =

(10)
yg u / cJ m3

ω can be adjusted linearly with g = g = g and

0 ml m2 m

g constant. Such movement is often termed con­

m3

stant bandwidth movement. If g g , and g are mV m2 m3

adjusted simultaneously, constant-Q pole movement

Fig. 7 Second-order filter structures. is possible. Adjusting g (for Q > Vi) moves the poles m3

MARCH 1985 27
 ous circuits has been split to allow for adjusting the
pole-zero ratio. The transfer function of the circuit is
given by

Ci
c 2 +c 3

S + gml/ClCl
(14)
2

s + sg /(C
2 m2 2 + C ) + gmlgJC^C,
3 + C) 3

This circuit has applications in higher-order voltage-

controlled elliptic filters. For higher-order structures
obtained by cascading these second-order blocks, all
g 's would be made equal and adjusted simultane­
m

ously. Buffering between stages using a standard

unity gain buffer is required to prevent interstage
loading. Modifications of the other circuits in Fig. 7 to
obtain ω and Q adjustable features is also possible.
0

Although the ratio of the zero location to pole location

can be controlled with the C / C ratio in discrete de­ 2 3

Fig. 8 Elliptic filter structure. signs, this may pose some problems in monolithic
structures. One convenient way to control the pole-
zero ratio is to insert the voltage-controlled amplifier
of Fig. 2g between the points χ and x' in Fig. 8 and
along vertical lines parallel to the jw axis in the use the transconductance gain of either of these addi­
s-plane. tional OTAs as the control variable.
The circuit of Fig. 7d has an output given by The final second-order structure considered here is
the general biquad of Fig. 9. The output for this cir­
VçC^s + Vßg^sC, +
2 gg V
ml m2 A
cuit is given by
s ^ Q + sC^ + g m 3
(11)
ml$m2
The ω and Q of the poles are, respectively, _ s C ! C y + sC g ±v
2 2 c x m B + g g v
m2 m5 A
0
(15)
Smlgml
$ml$m2
(12)
The potential for tuning the ω and Q for both the
Qc 2
poles and zeros (when V, = V = V = V ) to any A
0

B c

Q = (13) desired value should be apparent. Although some­

8m3 what component intense, it can be argued that if there
Although the transfer function is similar to that above,
note that since the coefficient of the s term in the nu­
merator equals that in the denominator, adjustment of
the bandpass version of this circuit with g = g = g ml m2 m

will result in a constant bandwidth, constant gain

response.
For monolithic structures, it may prove useful to
replace the resistor in Fig. 7b with the OTA structure
of Fig. 3a. Likewise, if the bandwidth adjustment
with g is not needed, it may be desirable to replace
m3

the third OTA shown in Fig. 7c with a fixed resistor in

some applications.
Phase equalizers are also possible with the struc­
tures shown in Fig. 7. For example, interchanging the
" + " and " - " terminals of the first two OTAs in Fig.
7c, setting V = V = V = V , and making g = g
A B c t ml m2

= g = g results in a second-order g adjustable

m3 m m

phase equalizer.
The circuit of Fig. 8 has both poles and zeros that
can be adjusted simultaneously in a constant-Q man­
ner. The circuit is similar to those shown in Fig. 7a
with the exception that the capacitor C in the previ- 2
Fig. 9 General biquadratic structure.

28 IEEE CIRCUITS AND DEVICES MAGAZINE

is to be capability for completely arbitrary location of
a pair of poles and a pair of zeros via adjustment of
the transconductance gain of the OTA, then at least
4 degrees of freedom and, hence, 4 OTAs are re­
quired. This circuit uses only one more than the mini­
mum! The capability for various types of pole and/or
zero movement through the simultaneous adjustment
of two or more of the transconductance gains should
also be apparent. Many other biquadratic structures,
some of which offer more flexibility at the expense of
additional complexity, also exist but are not discussed Fig. 11 Macromodel of bias current port on bipolar OTA.
here.
Emphasis in this section has been placed entirely
upon second-order structures in which the desired major limiting factor with commercially available
filter characteristics depend directly upon the trans­ OTAs is the limited differential input voltage swing.
conductance gain of the OTA. Very simple structures Recent activity in the literature has concentrated
in which the filter characteristics are adjustable upon designing OTAs with improved input character­
through the parameter g resulted. As stated in the
m
istics [27]-[28]. Significant improvements in perfor­
introduction, g is readily controllable by a dc bias
m
mance over what is currently available with discrete
current over a wide range of values, thus making OTAs have been demonstrated. An alternative is to
these circuits directly applicable to voltage-controlled use voltage attenuators and buffers at the input of ex­
applications. Several of the more recent works on isting OTAs. This technique is often suggested in the
OTA applications [18]-[24] have followed this ap­ manufacturer's application notes and is illustrated in
proach. Most of the earlier works [7]-[16] and the cir­ Fig. 10. This technique can be used to obtain reason­
cuits presented in the manufacturer's application notes able signal swings with all circuits discussed up to
[3]-[5] concentrated upon topologies in which the fil­ this point. Although such circuits are useful, a rather
ter characteristics are independent or only mildly de­ high price is paid for this modification. First, the cir­
cuit requires many more components. Second, the
pendent upon the transconductance gain. Most of
finite bandwidth of the op amps will limit the fre­
these structures are very complicated, very compo­
quency response of the OTA structures. Finally, the
nent-intense, and require tuning algorithms that are
attenuation of the input signal to the OTA causes a
unwieldy. Alternatives to these earlier designs using
serious loss in dynamic range. From a topological
conventional operational amplifiers have proven to be
point of view, some OTA based structures are inher­
much better.
ently more susceptible to differential voltage limita­
tions than others. This parallels the concern for op
Practical Considerations amp based active RC and switched-capacitor struc­
Although all circuits presented up to this point in tures that the signal amplitudes at the output of inter­
this paper are practical with ideal operational trans­ nal op amps assume acceptable values. These consid­
conductance amplifiers, existing discrete OTAs are far erations become more serious for high Q and high
from ideal. As mentioned in the introduction, the dynamic range applications.
A macro model of the bias current ( J ) input port A B C

of a typical bipolar OTA is shown in Fig. 11. This actu­

ally forms part of an internal current mirror that is
discussed later. Several schemes for controlling the
current / and, thus, the g of the OTA by an exter­
A B C m

nal control voltage, V , are shown in Fig. 12. The first

c

circuit is the simplest but is very sensitive to small

changes of V as V approaches .6v + V . In the sec­
c c

ond circuit, the control voltage is referenced to zero,

but the small V is sensitive to mismatches between
c

the B-Ε voltage of the transistor and the forward di­

ode voltage drop. In the circuit of Fig. 12c, the control
voltage is also referenced to ground and is not depen­
dent upon the matching or cancellation of voltages
across external forward biased pn junctions. The
zener diode is used to maintain the common mode
voltage at a reasonable level. The frequency response
of the op amp is not of concern here since it is used
Fig. 10 Signal conditioner for OTAs. only in the dc control path. It should be noted that the
MARCH 1985 29
 Fig. 12 Schemes for obtaining voltage control with the OTA.

Iref J I ABC ι I I ABC? lABC n

<
6V-

I ABC.
i<4 lABCn
Rn Fig. Ï 4 Smg/e input-multiple output bias current generator for
monolithic applications.

(b)
1 1

l ÏABC ABC

'VRET

F/g. 23 Schemes for simultaneous g m adjustment. Fig. 15 g m attenuator.

amplifier bias current is proportional to V for all c ferences in the voltage V of Fig. 11a for small values
d

schemes shown in Fig. 12. Since 7 can typically beABC of 7 . The circuit of Fig. 13b again has V referenced
ABC c

adjusted over several decades, all schemes will be to ground and is essentially independent of the match-
very sensitive to small changes in V toward the low c ing of V for the individual OTAs. The scheme of Fig.
d

current end of the I range. Logarithmic amplifiers

A B C 13c is useful if an external single package pnp current
are often used to control I with an external control
A B C mirror with η outputs is available. A discrete compo­
voltage if the wide adjustment range of 7 is to be ABC nent version of this mirror would not be practical.
effectively utilized. For integrated circuit applications, the amplifier
Many of the filter circuits discussed in the previous bias currents of several OTAs are particularly easy
sections of this paper require the simultaneous ad- to match and control. For monolithic applications,
justment of matched g 's. Several schemes for achiev-
m the simultaneous adjustment of the gain of a large
ing this are shown in Fig. 13. In the first circuit, it is number of OTAs with a single dc bias current can be
easy to adjust the g 's by trimming the resistors for a
m easily attained by using a single input-multiple out­
fixed g . The circuit is quite sensitive to the slight dif-
m put current mirror such as is shown in Fig. 14. This
30 IEEE CIRCUITS AND DEVICES MAGAZINE
structure actually replaces the bias current mirrors on of components. Applications include amplifiers,
each of the OTAs. The transconductance gains can be controlled impedances, and filters. Higher-order
ratioed, if desired, by correspondingly ratioing the continuous-time voltage-controlled filters such as the
emitter areas (or width length ratio for MOS struc­ common Butterworth, Chebyschev, and Elliptic types
tures) in the outputs of the current mirror. can be obtained. In addition to the voltage-control
With conventional operational amplifiers, the slew characteristics, the OTA based circuits show promise
rate, input impedance, output impedance, and maxi­ for high-frequency applications where conventional
mum output current are essentially fixed at the de­ op amp based circuits become bandwidth limited.
sign stage. For OTAs, it is generally the case that these The major factor limiting the performance of OTA
parameters are either proportional or inversely pro­ based filters using commercially available OTAs is the
portional to I . Thus adjusting g via 7 causes all
A B C m ABC severely limited differential input voltage capability
of these parasitic parameters to change accordingly. inherent with conventional differential amplifier in­
Although the user should be cognizant of the changes put stages. Recent research results suggested signifi­
in these parameters, the problems they present cant improvements in the input characteristics of
are manageable. The output capacitance of an OTA OTAs can be attained [27]-[28].
does cause concern at low output currents and high
frequencies.
Much as in the design of conventional op amp References
based circuits, the designer must allow for a dc bias [1] H. A. Wittlinger, "Applications of the CA3080 and CA3080A
High Performance Operational Transconductance Ampli­
current path for both input terminals of the OTA. Al­ fiers," RCA Application Note 1CAN-6668.
though the amplifier of Fig. 15 serves as an effective [2] C. F. Wheatley and H. A. Wittlinger, "OTA Obsoletes OR
g attenuator, which will prove useful in some appli­
m AMP," P. Nat. Econ. Conf. pp. 1 5 2 - 1 5 7 , Dec. 1969.
cations, the circuit is useless since the required input [3] RCA Electronic Components, Linear Integrated Circuits, Model
CA3060: Data File 404, Mar. 1970.
bias current will cause an accumulation of charge on
[4] National Semiconductor, Linear Applications Handbook, 1980.
the capacitors and eventual saturation of the OTA. [5] RCA Solid-State Division, Data Book, Linear Integrated Circuits,
The reader should be cautioned that more compli­ File No. 475, Mar. 1975.
cated circuits with the same problem are suggested in [6] M. Bialko and R. W. Newcomb, "Generation of All Finite Lin­
the literature [17]. ear Circuits Using the Integrated DVCCS," IEEE Trans, on Cir­
cuit Theory, vol. CT-18, pp. 7 3 3 - 7 3 6 , Nov. 1971.
Numerous nonlinear applications of OTA struc­
[7] M. Bialko, W. Sienko, and R. W. Newcomb, "Active Synthesis
tures exist. Suffice it to say that since the amplifier Using the DVCCS/DVCVS," Int. J. of Circuits Theory App.,
bias current, 7 , can be considered as a third signal
ABC vol. 2, pp. 2 3 - 2 8 , 1974.
input, simple multipliers, modulators, and a host of [8] S. Franco, "Use Transconductance Amplifier to Make Pro­
other nonlinear circuits are possible. The reader is grammable Active Filters," Electronic Design, vol. 24, pp. 9 8 -
101, Sept. 1976.
referred to the application notes for a discussion
[9] F. Atiya, A. Soliman, and T. Saadawi, "Active RC Bandpass
of some of the nonlinear applications. Some of the and Lowpass Filters Using the DVCCS/DVCVS," Electron.
structures that use only OTAs and capacitors show Lett. vol. 12, pp. 3 6 0 - 3 6 1 , July 1976.
promise for monolithic applications in MOS or bi­ [10] F. Anday, "On the Analysis and Synthesis of Active Networks
polar processes. The circuits should offer high- Containing DVCCS/DVCVS," Proc. IEEE, pp. 3 7 5 - 3 7 6 , Mar.
1976.
frequency continuous-time capabilities. Either exter­
[11] F. S. Atiya, A. M. Soliman, and T. N. Saadawi, "Active RC
nal voltage-control or an internal reference circuit to Nominum Phase Network Using the DVCCS/DVCVS," Proc.
compensate for process and temperature variations IEEE, vol. 65, pp. 1606-1607, Nov. 1977.
will be necessary to make these circuits practical in [12] A. M. Soliman, "A Grounded Inductance Simulation Using
demanding applications. the DVCCS/DVCVS," Proc. IEEE, vol. 66, pp. 1089-1091,
Sept. 1978.
Finally, it should be noted that some of the filter
[13] R. Nandi, "New Ideal Active Inductance and Frequency-
structures presented earlier in this paper have a non- Dependent Negative Resistance Using DVCCS/DVCVS: Ap­
infinite input impedance, and that the output imped­ plications in Sinusoidal-Oscillator Realization," Electron. Lett.,
ance is generally quite high. Cascading of such struc­ vol. 14, pp. 5 5 1 - 5 5 3 , Aug. 1978.
tures will require interstage buffer amplifiers, which [14] I. M. Filanovsky and K. A. Stromsmoe, "More Active-RC Fil­
ters Using DVCCS/DVCVS," Electron. Lett., vol. 15 pp. 4 6 6 -
will tend to degrade the bandwidth of the overall filter
467, Aug. 1979.
structures. Output buffers are also generally required [15] D. Patranabis and A. Paul, "Floating Ideal Inductor with One
to drive external loads. DVCCS," Electron. Lett., vol. 15, pp. 5 4 5 - 5 4 6 , Aug. 1979.
[16] T. Deliyannis, "Active RC Filters Using an Operational Trans­
conductance Amplifier and an Operational Amplifier," Int. J.
Conclusions Circuit Theory Appl, vol. 8, pp. 3 9 - 5 4 , Jan. 1980.
A group of voltage-controlled circuits using the [17] A. Urbas and J. Osiwski, "High-Frequency Realization of
C-OTA Second-Order Active Filters," Proc. IEEEIISCAS,
OTA as the basic active element have been presented.
pp. 1 1 0 6 - 1 1 0 9 , 1982.
The characteristics of these circuits are adjusted with [18] H. S. Malvar, "Electronically Controlled Active Filters with
the externally accessible dc amplifier bias current. Operational Transconductance Amplifiers," IEEE Trans. Cir­
Most of these circuits utilize a very small number cuits Syst., vol. CAS-29, pp. 3 3 3 - 3 3 6 , May 1982.

MARCH 1985 31
[19] R. L. Geiger and J. Ferrell, "Voltage Controlled Filter Design
Using Operational Transconductance Amplifiers," Proc. IEEE/
ISCAS, pp. 5 9 4 - 5 9 7 , May 1983.
[20] A. R. Saha and R. Nandi, "Integrable Tunable Sinusoid Os­
cillator Using DVCCS," Electron. Lett., vol. 19, pp. 7 4 5 - 7 4 6 ,
Sept. 1983.
[21] G. M. Wierzba and S. Esmelioglu, 'Techniques for Designing
Enhanced-Gain-Bandwidth-Product Circuits," Proc. 26th Mid­
west Symp. on Circuits and Syst., pp. 6 0 2 - 6 0 6 . Aug. 1983.
[22] R. W. Newcomb and S. T. Liu, "A Voltage Tunable Active-R
Filter," Proc. IEEE/ISCAS, pp. 4 0 9 - 4 1 2 , May 1984.
[23] J. Hoyle and E. Sânchez-Sinencio, "Sinusoidal Quadrature
OTA Oscillators," Proc. 27th Midwest Symp. on Circuits and
Syst., June 1984.
Randall L. Geiger Edgar Sânchez-Sinencio
[24] H. S. Malvar, "Electronically Controlled Active Active-C Fil­
ters and Equalizers with Operational Transconductance Am­
plifiers," IEEE Trans. Circuits Syst., vol. CAS-31, pp. 6 4 5 - 6 4 9 , Randall L. Geiger was born in Lexington, Nebraska, on May 17,
July 1984. 1949. He received the B.S. degree in electrical engineering and the
[25] F. Krummenacher, "High-Voltage Gain CMOS OTA for Micro- M.S. degree in mathematics from the University of Nebraska, Lin­
power SC Filters," Electron. Lett., vol. 17, pp. 1 6 0 - 1 6 2 , Feb. coln, in 1972 and 1973, respectively. He received the Ph.D. degree
1981. in electrical engineering from Colorado State University, Fort Col­
[26] M. G. Degrauwe, J. Rijmenants, E. A. Vittoz, and J. deMan, lins, in 1977.
"Adaptive Biasing CMOS Amplifiers," IEEE J. Solid-State Cir­ In 1977, Dr. Geiger joined the Faculty of the Department of Elec­
cuits, vol. SC-17, pp. 5 2 2 - 5 2 8 , June 1982. trical Engineering at Texas A&M University, College Station, and
[27] K. D. Peterson and R. L. Geiger, "CMOS OTA Structures with currently holds the rank of Associate Professor. His present re­
Improved Linearity," Proc. 27th Midwest Symp. on Circuits and search is in the areas of integrated circuit design and active circuits.
Syst., June 1984. He received the Meril B. Reed Best Paper Award at the 1982 Mid­
[28] A. Nedungadi and T. R. Viswanathan, "Design of Linear west Symposium on Circuits and Systems, served as Conference
CMOS Transconductance Elements," IEEE Trans, on Circuits Chairman at the 1983 UGIM conference, and is currently serving as
and Syst., vol. CAS-31, pp. 8 9 1 - 8 9 4 , Oct. 1984. an Associate Editor for the IEEE Transactions for Circuits and Systems.
Dr. Geiger is a member of Eta Kappa Nu, Sigma Xi, Pi Mu Ep­
[29] K. S. Tan and P. R. Gray, "Fully Integrated Analog Filters Using
silon, and Sigma Tau; he is also a Senior Member of the IEEE.
Bipolar-JFET Technology," IEEE J. Solid-State Circuits, vol.
SC-12, pp. 8 1 4 - 8 2 1 , Dec. 1978.
[30] A. P. Nedungadi and P. E. Allen, "A CMOS Integrator for
Edgar Sânchez-Sinencio was born in Mexico City, Mexico, on
Continuous-Time Monolithic Filters," Proc. IEEE/ISCAS, vol.
October 27, 1944. He received the degree in communications and
2, pp. 9 3 2 - 9 3 5 , May 1984.
electronic engineering (professional degree) from the National
[31] G. Tröster and W. Langheinrich, "Monolithic Continuous-
Polytechnic Institute of Mexico, Mexico City; the M.S.E.E. degree
Time Analogue Filters in NMOS Technology," Proc. IEEE/
from Stanford University, California; and the Ph.D. degree from
ISCAS, vol. 2, pp. 9 2 4 - 9 2 7 , May 1984.
the University of Illinois at Champaign-Urbana, in 1966, 1970, and
[32] H. Khorramabadi and P. R. Gray "High-Frequency CMOS
1973, respectively.
Continuous-Time Filters," Proc. IEEE/ISCAS, vol. 3, pp. 1 4 9 8 -
During his graduate studies, he was awarded with fellowships
1501, May 1984.
from the United Nations Educational, Scientific, and Cultural Or­
[33] K. Fukahori, "A Bipolar Voltage-Controlled Turnable Filter,"
ganization; the Mexican Atomic Energy Commission; and the Con-
IEEE ] . Solid-State Circuits, vol. SC-16, pp. 7 2 9 - 7 3 7 , Dec. 1981.
sejo Nacional de Ciencia y Tecnologia of Mexico. From January
[34] K. W. Moulding, "Fully Integrated Selectivity at High Fre­ 1965 to March 1967, he worked with the Mexican Atomic Energy
quency Using Gyrators," IEEE Trans. Broadcast. Telev. Reg., vol. Commission, as a Design Engineer. In April 1967, he joined the
BTR-19, pp. 1 7 6 - 1 7 9 , Aug. 1973. Petroleum Institute of Mexico, where he was associated with
[35] H. O. Voorman and A. Biesheuvel, "An Electronic Gyrator," the design of instrumentation equipment until August 1967. He
IEEE ] . Solid-State Circuits, vol. SC-7 pp. 4 6 9 - 4 7 4 , Dec. 1972. worked as a Research Assistant at the Coordinated Science Labora­
[36] K. W. Moulding and P. J. Ranking, "Experience with High- tory at the University of Illinois from September 1971 to August
Frequency Gyrator Filters Including a New Video Delay-Line 1973.
IC," Proc. 6th European Conf. on Circuit Theory and Design, In 1974, Dr. Sânchez-Sinencio held an industrial post-doctoral
pp. 9 5 - 9 8 , Sept. 1983. position with the Central Research Laboratories, Nippon Electric
[37] J. O. Voorman, W. H. A. Bruls, and J. P. Barth, "Bipolar Integra­ Company, Ltd., Kawasaki, Japan. From 1976 to 1983, he was the
tion of Analog Gyrator and Laguerre Type Filters (Transcon- Head of the Department of Electronics at the Instituto Nacional
ductance-Capacitor Filters)," Proc. 6th European Conf. on Cir­ de Astrofisica, Optica y Electronica (INAOE), Puebla, Mexico.
cuit Theory and Design, pp. 1 0 8 - 1 1 3 , Sept. 1983. Dr. Sânchez-Sinencio was a Visiting Professor in the Department of
[38] J. O. Voorman, W. H. A. Brüls, and P. J. Barth, "Integration of Electrical Engineering at Texas A&M University during the aca­
Analog Filters in a Bipolar Process," IEEE ] . of Solid-State Cir­ demic years of 1 9 7 9 - 1 9 8 0 and 1 9 8 3 - 1 9 8 4 , where he is currently a
cuits, vol. SC-17, pp. 7 1 3 - 7 2 2 , Aug. 1982. Professor. He was the General Chairman of the 1983 26th Midwest
[39] K. W. Moulding, J. R. Quarterly, P. J. Rankin, R. S. Thompson, Symposium on Circuits and Systems. He is the coauthor of the
and G. A. Wilson, "Gyrator Video Filter IC with Automatic book Switched-Capacitor Circuits (Van Nostrand-Reinhold, 1984). Dr.
Tuning," IEEE J. of Solid-State Circuits, vol. SC-15, pp. 9 6 3 - 9 6 8 , Sânchez-Sinencio's present interests are in the areas of active filter
Dec. 1980. design, solid-state circuits, and computer-aided circuit design. He
[40] V. W. Burgger, B. J. Hosticka, and G. S. Moschytz, "A Com­ is a Senior Member of the IEEE.
parison of Semiconductor Controlled Sources for the Design
of Active RC Impedances," Int. ] . of Circuit Theory and Appl.,
vol. 10, pp. 2 7 - 4 2 , 1982.

32 IEEE CIRCUITS AND DEVICES MAGAZINE