Trade-offs in the OTA-based analog filter design

Bilgin Metin Æ Kirat Pal Æ Shahram Minaei ÆOguzhan Cicekoglu

Received: 4 August 2008 / Revised: 1 December 2008 / Accepted: 8 December 2008 / Published online: 5 February 2009

� Springer Science+Business Media, LLC 2009

Abstract Operational transconductance amplifiers (OTAs)

are widely used in the design of electronically tunable

circuits. However, electronic tunability ranges of the OTA

based filters are restricted by the limited bandwidth of the

transconductance gain of the OTA. Furthermore, stability

conditions and the linearity of the OTA which depends on

the control current restrict the tunability. In this paper, some

trade-offs in the electronically tunable filters are investi-

gated. In addition, the tunability ranges of some first and

second order OTA-C and OTA-RC filters are comparatively

examined. Moreover, an OTA-C all-pass filter circuit is

presented. SPICE simulations are performed and stability

analyses are given for both of the OTA-C and OTA-RC

filters. Operation of the presented all-pass filter is verified

experimentally.

Keywords OTA � All-pass filter � Trade-off � Stability �Tunability

1 Introduction

There has been a great interest to resistorless designs in the

literature due to the difficulties in the implementation of

resistors in integrated circuits (IC), such as high tolerance,

parasitic components, and large chip area.

Electronically tunable circuits attracted considerable

attention in the design of analog integrated circuits because

tolerances of the electronic components in IC realization in

practice are unacceptably high and thus fine-tuning is

necessary. Tunability is more important in advanced

technologies where component sizes are strongly reduced

and the designers control on absolute element values is

weak. The OTA-based circuits are most suitable from

electronic tunability point of view as their transconduc-

tance gain (gm) can be varied for several decades through

its bias current.

The OTA is a commercially available active component

which has been used widely in many applications. More-

over, the OTA is a simpler element compared to similar

active elements such as current conveyors, because OTA

consists of a voltage controlled current source whereas the

current conveyor is composed of one voltage controlled

voltage source and one current controlled current source.

Therefore the OTA internal structure is expected to use

fewer transistors compared to current conveyor.

In 1976, Franco proposed first example of electronically

tunable filter circuit using OTA [1]. Then, the OTA circuits

have been shown to be potentially advantageous for the

design of many first and second order high-frequency analog

filters [2–8]. For example, first order all-pass filter circuit in

[2] includes two OTAs, one resistor and one capacitor. The

circuit in [3] uses a single OTA, a grounded capacitor and

four resistors. However, the all-pass filters in [2, 3] are not in

the scope of this paper due to their voltage-mode operation.

B. Metin (&)

Department of Management Information Systems, Bogazici

University, Hisar Campus, 34342 Bebek-Istanbul, Turkey

e-mail: bilgin.metin@boun.edu.tr

K. Pal

Department of Earthquake Engineering, Indian Institute

of Technology, Roorkee 247667, Uttaranchal, India

e-mail: kiratfeq@iitr.ernet.in

S. Minaei

Department of Electronics and Communications Engineering,

Dogus University, Acibadem, Kadikoy 34722, Istanbul, Turkey

e-mail: sminaei@dogus.edu.tr

O. Cicekoglu

Department of Electrical and Electronic Engineering,

Bogazici University, 34342 Bebek-Istanbul, Turkey

e-mail: cicekogl@boun.edu.tr

123

Analog Integr Circ Sig Process (2009) 60:205–213

DOI 10.1007/s10470-008-9270-x

The current-mode OTA-based circuit in [4] uses one

capacitor and two OTAs, which are one dual output OTA

and one triple output OTA. Also, many second order OTA-

based filters have been presented in the literature [5–8]. The

circuit proposed in [5] uses two OTAs to realize a biquad

filter. In addition, the filter circuit reported in [6] employs

three OTAs. In [7] both resistor and OTA are used in the

same circuit. Many other examples can be found in [5] and

[8]. In addition, OTA has been an inspiration for developing

new active elements. Recently, an element called current

difference transconductance amplifier (CDTA) which is

based on OTA has been developed and used in some first

order and second order filter designs [9, 10].

In this work, an OTA-C first-order all-pass filter is

proposed and compared with another OTA-C first order all-

pass filter [4] by discussing some trade-offs [11] in analog

filter design as follows:

Grounded or floating capacitor trade-off: A trade-off

exists between convenience for IC implementation and

high-frequency operation. A floating capacitor between the

input and the output of the circuit bypasses a section of the

circuit and introduces a feed-forward path, which cancels

the effect of non-idealities of the active element and pre-

vents roll-off at high frequencies. On the other hand a

grounded capacitor has less parasitic elements compared to

the floating one in the IC implementation. In addition, it is

more convenient for fabrication than floating one which

requires an IC process with two poly layers.

The tunability range and output current amplitude

trade-off: The control current of the active element is

usually also used for biasing purpose. Therefore, keeping

the control current sufficiently higher than the amplitude of

the output signal current may improve the linearity of the

active element. This may limit the tunability range of the

filter.

The tunability range and stability/frequency response

trade-off: The electronically tunable circuits have attrac-

ted increasing attention in the design of analog integrated

circuits, because the absolute tolerances of the electronic

components can exceed 20% and thus fine-tuning is a

must. On the other hand, due to the parasitic capacitances

of the active element parasitic poles are created and this

alters the ideal transfer function of the circuit. This may

lead to two important problems. First, the order of the

filter may be increased and some undesired high order

terms appear in the transfer function. The effects of these

undesired terms may be very strong on the tunability

range for the filter. Secondly, the altered transfer function

can be unstable for some passive element values. There-

fore, the stability conditions may also decrease the

tunability range of the circuit.

The cascadability and power consumption/chip area

trade-off: The easiest approach to decrease power

consumption of the analog filters is to reduce the number of

active elements. However, many circuits require additional

active element for cascading which increases the power

consumption and chip area.

Active element noise and resistor thermal noise trade

off: Analog design without resistors obviously removes

resistor thermal noise but introduces transistor flicker

noise. However the amount of the noise is also dependent

on the circuit topology and needs further investigation

which is kept out of the scope of this study.

2 The presented OTA-C first order all-pass filter and

effects of the frequency limitation

The proposed first-order all-pass filter is shown in

Fig. 1(a). An ideal OTA is an infinite bandwidth voltage-

controlled current source, with an infinite input and output

impedance. The output current of an OTA is given as

IOUT = agm(V?-V-), where the IOUT current is defined as

flowing out from the OTA and a is the gain of the current

mirror at the output stage of the OTA that is ideally equal

to unity. The transconductance parameter (gm) of the OTA

can be controlled by an external current. Assuming ideal

OTA, routine analysis of the proposed circuit gives the

following current transfer function

Io

Ii¼ sC � gm

sC þ gmð1Þ

Fig. 1 a The proposed OTA-based first order all-pass filter. b The

OTA based all-pass filter presented by by Al-Hashimi et al. [4]

206 Analog Integr Circ Sig Process (2009) 60:205–213

123

It is well known that the gm of the OTA is a frequency-

dependent parameter [12] and can be expressed as,

gmðsÞ ¼gmo

1þ sxp

ð2Þ

where gmo is the zero-frequency transconductance gain

and xp is the parasitic pole frequency. Furthermore,

since the proposed circuit uses a dual-output OTA, both

of the output currents of the OTA should be equal for

proper filtering operation [13]. In other words the gains

of the current mirrors employed at the output stage of

the OTA should be unity. However, in practice due to

the mismatches between current-mirror transistors, these

gains can be considered as a1 and a2 instead of unity for

the first and second outputs of the OTA, respectively.

Taking into account these non-ideal current gains and

frequency-dependent characteristic of the gm, the

following transfer function is obtained.

Io

Ii¼

s2 C=xp

� �þ sC � a2gmoð Þ

s2 C=xp

� �þ sC þ a1gmoð Þ

ð3Þ

The transfer function of the circuit in Fig. 1(b) can be

given as follows [4],

Io

Ii¼ � sC � gm2

sC þ gm2

ð4Þ

Substituting frequency dependent gm model of (2) and

considering the non-ideal current mirror gains of the OTAs

into (4) the following transfer function is obtained.

Io

Ii¼�a12s2 C=xp2

� �þ �a12sC þ a13a22gm20ð Þ

a11s2 C=xp2

� �þ a11sC þ a13a21gm20ð Þ

ð5Þ

where gm20 is the zero-frequency transconductance gain of

the OTA2, aij (i = 1,2 and j = 1,2,3) are the non-ideal

current gain of the ith OTA at its jth output. Comparison of

(5) and (3) shows that the circuit in 1(b) has a more

complicated dependency to the output current mismatches

of the OTA. In order to operate the circuits in Fig. 1(a, b)

as a first-order filters, the following conditions should be

satisfied respectively.

gm0 �x2C

aixpi ¼ 1; 2 ð6aÞ

gm20 �a11x2C

a13a21xp2

gm20 �a12x2C

a13a22xp2

9>>>=

>>>;

ð6bÞ

It is obvious that the restrictions given in (6a) and (6b)

decrease the tunability range of the filters. On the other

hand, the limited bandwidth of the gm does not impose any

restriction with respect to the stability as it is seen in (3)

and (5).

3 Effects of frequency limitation of the gm

on tunability in second order filters

In the second order filters, the effects of the frequency

limitation of the gm are expected to be more complicated in

comparison to the first order filters. Therefore, these effects

are examined under two different subsections. In this sec-

tion, two OTA-based second-order filters reported in [5]

and [7] are chosen for investigation.

3.1 Frequency response and tunability range trade-off

in second order filters

The OTA-RC circuit in [7] is shown in Fig. 2(a). Routine

analysis of the circuit gives the transfer function as

Vo

Vi¼ gm

s2R1C1C2 þ sðC1 þ C2Þ þ gmð7Þ

From (7) the angular pole frequency and quality factor of

the filter (Q) can be respectively found as

x0 ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

gm

C1C2R1

rð8aÞ

Q ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffigmC1C2Rp

1

C1 þ C2

ð8bÞ

Considering the frequency limitation of gm as given in Eq. 2

and reanalysis of the OTA-RC circuit in Fig. 2(a) yields

Fig. 2 Second order OTA based filters derived from (a) [7], (b) [5]

Analog Integr Circ Sig Process (2009) 60:205–213 207

123

Vo

Vi¼ gmo

D sð Þð9Þ

where,

DðsÞ ¼ s3 R1C1C2

xpþ s2 R1C1C2 þ

C1 þ C2

xp

� �

þ sðC1 þ C2Þ þ gmo ð10Þ

From Eq. 10 it can be seen that due to the limited

bandwidth of the transconductance gain of the OTA we

obtain a third order filter response with undesirable terms

in their transfer functions. In order to operate the circuit as

a second-order filter the following conditions should be

satisfied:

x2 � xpðC1 þ C2ÞR1C1C2

R1 �C1 þ C2

C1C2xp

9>>=

>>;ð11Þ

As a second example, the OTA-based filter shown in

Fig. 2(b) [5] is considered. The transfer function of this

filter can be given as follows:

Vo

Vi¼ gm1gm2

s2C1C2 þ sC1gm2 þ gm1gm2

ð12Þ

From Eq. 12 the angular pole frequency and quality

factor of the filter (Q) can be respectively found as:

x0 ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffigm1gm2

C1C2

rð13aÞ

Q ¼ffiffiffiffiffiffiffiffiffiffiffiffiC2gm1

pffiffiffiffiffiffiffiffiffiffiffiffiC1gm2

p ð13bÞ

Considering frequency-dependent OTA model in Eq. 2, the

transfer function in (12) yields

In (14), ignoring the fourth-order effects, the circuit can

operate as ideal case by satisfying the following conditions

gm02 � C2xp1

x 2 � gm02xp1xp2

C2ðxp1 þ xp2Þ

9=

;ð15Þ

In this way, we can compare tunability range of the circuit

in Fig. 2(a) with the OTA-C low-pass filter circuit shown in

Fig. 2(b). From (11) and (15) it can be seen that, while the

OTA-RC filter in Fig. 2(a) does not impose any constraint

on the transconductance gain of the OTA, the workability of

the OTA-C filter in Fig. 2(b) depends on the value of the

transconductance gain of the second OTA (gm02). There-

fore, in the latter case gm02 cannot be used freely as a tool

for tuning the pole frequency or the quality factor.

3.2 Stability conditions and tunability range trade-off

in second order filters

The frequency limitations of the gm cause difficulties in the

stability of the OTA-based filters. In addition, employing a

large number of OTAs in a circuit increases the number of

feedback loops and may deteriorate the circuit stability.

Consequently, the tunability range of the filters is expected

to decrease. In other words, the circuit can be stable in a

narrow range of tuning. It should be noted that the cut-off

frequency of the gm depends on the OTA design and

transistor technology used in the realization of OTA.

Limited bandwidth of the transconductance gain of the

OTAs affects stability of the circuits. Therefore, we apply

Routh-Hurwitz stability criteria on Eq. 10 and denominator

of Eq. 14 to examine stability of the circuits in Fig. 2(a, b).

In this way, we can find minimum allowable value of xp.

Applying Routh-Hurwitz criterion on Eq. 10 for the

circuit of Fig. 2(a) gives the following condition:

C1 þ C2 �C1C2gm0R1

C1 þ C2 þ C1C2R1xp[ 0 ð16Þ

From (16), we can find the minimum value of xp such that

the circuit remains stable as

xp [gm0

C1 þ C2

� C1 þ C2

C1C2R1

ð17Þ

Therefore, the designer must take particular care to ensure

the above inequality. It is observed that that the OTA-RC

circuit can be stable even for OTAs having small xp val-

ues. For example, if we select the passive component

values as R1 = 2.5 kX, C1 = C2 = 50 pF and the trans-

conductance gm0 = 1.6 mS to realize a filter with Q = 1

and x0 = 16 Mrad/s, we find the condition of (17) as

xp [ 0. This means that the frequency-dependent gm of the

OTA does not cause stability problem in the circuit for the

selected element values.

The stability is also investigated for the OTA-C filter

in Fig. 2(b). Applying the Routh-Hurwitz stability

Vo

Vi¼ gm01gm02

s4 C1C2

xp1xp2þ s3 C1C2ðxp1þxp2Þ

xp1xp2þ s2 C1gm02

xp1þ C1C2

� �þ sC1gm02 þ gm01gm02

ð14Þ

208 Analog Integr Circ Sig Process (2009) 60:205–213

123

criteria on (14) and assuming xp2 ¼ kxp1; where xp1

and xp2 are the cut-off frequencies of the gms of the

OTA1 and OTA2 respectively, the following stability

condition is found:

�ð1þ kÞ2C2gm01 þ k2C1gm02 þ C1C2xpkð1þ kÞ[ 0

ð18Þ

So,

xp [ð1þ kÞgm01

kC1

� kgm02

ð1þ kÞC2

ð19Þ

For the passive component values of C1 = C2 = 60 pF and

identical OTAs with transconductances gm01 = gm02 =

1 mS and xp1 ¼ xp2 ¼ xp (k ¼ 1Þ; it is found that xp

should be greater than 25 Mrad/s. Therefore using of OTAs

with xp values lower than 25 Mrad/s will limit the tun-

ability range of the filter. By choosing C1 [ C2 for OTAs

with low xp values, the stability can be obtained more

easily, but it restricts the Q value of the OTA-C filter in

(13b).

To examine the stability of the OTA-RC circuit of

Fig. 2(a) in detail, root-locus diagrams are given in Fig. 3

for the passive component values of R1 = 2.5 kX,

C1 = C2 = 50 pF and gm0 = 1.6 mS while xp value is

changing between 0.5 Mrad/s and 30 Mrad/s. As it can be

seen from Fig. 3, the circuit is still stable for such a low xp

value as 0.5 Mrad/s. In addition, for the OTA-C filter

circuit of Fig. 2(b), we select C1 = C2 = 60 pF, gm01 =

gm02 = 1 mS and change xp values between 10 Mrad/s

and 30 Mrad/s. The result is shown in Fig. 4. It can be

seen that the value of xp must be greater than 25 Mrad/s

for the OTA-C filter circuit to be stable under given

assumptions.

4 Simulation and experimental results

To verify the theoretical analyses, we simulated the circuit

in Fig. 1(a) using the SPICE circuit simulation program.

A BJT-based OTA implementation shown in Fig. 5 is used

in the simulations [14]. Here the transconductance gain

of the OTA can be calculated as gm = IC/2VT, where

VT = kT/q & 26 mV at 27�C is the thermal voltage. The

supply voltages are VCC = 2.5 V and VEE = -2.5 V.

The OTA is implemented by using AT&T ALA400 BJT

transistors [15]. SPICE parameters of the transistors are

tabulated in Table 1. The capacitor value of C = 60 pF is

employed in the simulations. In order to show the tunability

of the presented circuit, the pole frequency of the proposed

-15 -10 -5 0

x 106

-1.5

-1

-0.5

0

0.5

1

x 107

real

ima

gina

ry

ωp=30Mrad/s

ωp=30Mrad/s

ωp=0.5Mrad/s

ωp=0.5Mrad/s

-8 -6 -4 -2 0

x 104

-1

-0.5

0

0.5

1

1.5x 10

7

real

imag

inar

y

Fig. 3 Root-locus graphic of

the OTA-RC circuit with one-

pole OTA model for

0.5 Mrad/s \xp \ 30 Mrad/s

-3 -2.5 -2 -1.5 -1 -0.5 0 0.5

x 107

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2x 107

real

ωP=10MHz

ωP=10MHz

ω P =30MHz

ωP =30MHz

ωP =25MHz

ωP =25MHz

imag

inar

y

Fig. 4 Root-locus graphic of the OTA-C circuit with one-pole OTA

model for 10 Mrad/s \xp \ 30 Mrad/s

Analog Integr Circ Sig Process (2009) 60:205–213 209

123

filter varied between f0 % 1.3 MHz and f0 % 5.2 MHz for

IC = 26 lA and IC = 104 lA. The SPICE simulation

results of the gain and phase responses are depicted in Fig. 6.

The presented circuit in Fig. 1(a) is compared with OTA

based circuit in Fig. 1(b). Dual and triple output OTAs of

Fig. 1(b) are obtained modifying the OTA implementation

in Fig. 5. The phase and the gain responses are depicted in

Fig. 7 for gm = 0.5 mS (IC = 26 lA) and C = 60 pF for a

pole frequency of f = 1.32 MHz. Note that the phase

response of the Al-Hashimi’s circuit of Fig. 1(b) is shifted

by 180� for a fair comparison with the proposed one. After

10 MHz, the gain response of the Al-Hashimi’s circuit

starts rolling off while the proposed circuit provides quite

flat phase and gain responses even above hundreds of MHz

frequency ranges.

To illustrate the time-domain performance and linearity

of the circuits in Fig. 1(a, b), the THD analyses are per-

formed using SPICE. The filters are constructed with

capacitor value of 60 pF and control current of 52 lA. The

THD values for linearity comparison for various input

amplitudes at 20 kHz are given in Fig. 8. Figure 8 shows

that the input signal amplitude of the examined circuits

should be chosen sufficiently lower than the control current

to obtain better linearity.

In addition, the all-pass filter in Fig. 1(a) is experi-

mentally constructed using LM13700 from National

Semiconductor which includes two OTAs to obtain a dual

output OTA. The power supply voltages are chosen as

±12 V. The transconductance gain of the LM13700 is

calculated as gm = 19.2 IC at 27�C. The two OTAs in

LM13700 are biased for a transconductance value of

2 mS with IC = 105 lA. Figure 9 shows the experimental

results of the phase and the gain responses with

C = 0.9 nF. A sinusoidal input signal peak to peak

100 lA is applied to the filter and at the output a 100 kXresistor is used as a load. Experimental results are close to

ideal values.

Fig. 5 A simple bipolar OTA design [14]

Table 1 ALA400 bipolar real process parameters from AT&T

.MODEL PR100 N PNP (IS = 73.5E-018 BF = 110 VAF = 51.8

IKF = 2.359E-3 ISE = 25.1E-16 NE = 1.650 BR = 0.4745

VAR = 9.96 IKR = 6.478E-3 RE = 3 RB = 327 RBM = 24.55

RC = 50 CJE = 0.180E-12 VJE = 0.5 MJE = 0.28

CJC = 0.164E-12 VJC = 0.8 MJC = 0.4 XCJC = 0.037

CJS = 1.03E-12 VJS = 0.55 MJS = 0.35 FC = 0.5

TF = 0.610E-9 TR = 0.610E-8 EG = 1.206 XTB = 1.866

XTI = 1.7)

.MODEL NR100 N NPN (IS = 121E-018 BF = 137.5

VAF = 159.4 IKF = 6.974E-3 ISE = 36E-16 NE = 1.713

BR = 0.7258 VAR = 10.73 IKR = 2.198E-3 RE = 1

RB = 524.6 RBM = 25 RC = 50 CJE = 0.214E-12 VJE = 0.5

MJE = 0.28 CJC = 0.983E-13 VJC = 0.5 MJC = 0.3

XCJC = 0.034 CJS = 0.913E-12 VJS = 0.64 MJS = 0.4

FC = 0.5 TF = 0.425E-9 TR = 0.425E-8 EG = 1.206

XTB = 1.538 XTI = 2)

1.0KHz 10KHz 100KHz 1.0MHz 10MHz 100MHz 1.0GHz

0d

100d

-30d

180d

-10

-5

0

5

10

_.._.._ IC=52µA........ IC=104µA_ _ _ IC=26µA ____ Ideal

Phase [degree]

Gain[dB]

Phase

Gain

Fig. 6 Illustrating tunability of

the proposed circuit varying its

pole frequency between

f0 % 1.3 MHz and

f0 % 5.2 MHz with control

current

210 Analog Integr Circ Sig Process (2009) 60:205–213

123

5 Discussions

It seems that the presented circuit can be useful when the

trade-offs above are taken into consideration. In contrast to

grounded capacitor advantage of the circuit in [4], the

floating capacitor of the presented filter bypasses a section of

the proposed circuit, so it prevents roll-off at high frequen-

cies. Therefore, the proposed filter is expected to be more

suitable for high frequency applications than the circuit in

[4]. The frequency-response comparison in Fig. 7 supports

this idea. The circuit in Fig. 1(b) is cascadable, but it requires

two OTAs. In addition, Fig. 8 point outs another trade-off

between the control current value and output signal ampli-

tude. The control current should be kept sufficiently higher

than the output signal amplitude for this circuit, restricting

the tunability range. The limited bandwidth of the gm

decreases tunability range of these first order filters as given

in Eq. 6, but it does not restrict the tunability range due to the

stability problem. On the other hand, the limited bandwidth

of the gm brings stability restrictions to the second order

filters in Fig. 2(a, b) as shown in Eqs. (17) and (19). This

situation is clarified by the root-locus graphics in Figs. 3 and

4. Considering the frequency limitation of the gm gives us

important clues for better tunability. For example (15) shows

that the transconductance gain of the second OTA (gm02)

cannot be used freely as a tool for tunability. On the other

hand, the effect of the frequency limitation of the gm on the

Frequency

1.0KHz 10KHz 100KHz 1.0MHz 10MHz 100MHz 1.0GHz

0d

100d

-30d

180d

-20

-10

0

10

20

. . . . . Al-Hashimi+180o___ Ideal _ _ _ _ _ Proposed

Gain [dB]

Phase [degree]

Phase

Gain

Fig. 7 The phase and gain

response comparison with the

Al-Hashimi’s circuit (The phase

response of Al-Hashimi’s

circuit is shifted by 180� for

comparison with the proposed

circuit)

Fig. 8 THD comparison of output currents with 20 kHz sinusoidal

input signal when C = 60 pF and IC = 52 lA

1 10 100 1,000

Frequency [kHz]

0

40

80

120

160

200

Pha

se [D

egre

e]

-4

-2

0

2

4

Gai

n [d

B]

Ideal

Measured Phase

Measured Gain

Fig. 9 Experimental and ideal gain and phase responses of the

proposed circuit

Analog Integr Circ Sig Process (2009) 60:205–213 211

123

stability can be diminished by choosing appropriate values

for passive elements such as C1 [ C2 for the circuit in

Fig. 2(b) considering (19). Consequently considering the

trade offs between different parameters in advance, the

desired features of the filters can be improved.

The model given by Eq. 2 used in the above calculations

provides a first order approach to the stability problem in

conjunction with frequency dependency of gm. Therefore, it

does not guarantee stability. However it gives an idea to the

designer.

6 Conclusions

This study discusses some design trade-offs of analog fil-

ters by presenting a current mode-mode all-pass filter with

a single dual output OTA. The proposed circuit has been

compared with another OTA based all-pass filter realiza-

tion. The presented circuit utilizes all the benefits of the

OTA-C filters such as integrability, easy tunability and

being resistorless. Moreover, it is most suitable for high

frequency applications, because the floating capacitor

bypasses a portion of the circuit. The circuit is verified with

SPICE simulations and the results confirm the theory.

Acknowledgement This work was supported by Bogazici Univer-

sity Research Fund with the project code 05A201D.

References

1. Franco, S. (1976). Use transconductance amplifier to make pro-

grammable active filters. Electronic Design, 24(19), 98–101.

2. Geiger, R. L., & Sanchez-Sinencio, E. (1985). Active filter design

using operational transconductance amplifiers: A tutorial. IEEECircuits and Devices Magazine, 1, 20–32.

3. Shah, N. A., & Ahmad, S. N. (1990). Electronically tunable

OTA-based all-pass circuit. International Journal of Electronics,68(6), 963–966. doi:10.1080/00207219008921237.

4. Al-Hashimi, M., Dudek, F., & Sun, Y. (2000). Current-mode

delay equalizer design using multiple output OTAs. AnalogIntegrated Circuits and Signal Processing, 24, 163–169. doi:

10.1023/A:1008388106946.

5. Urbas, A., & Osiowski, J. (1982). High-frequency realization of

C-OTA second-order active filters. Proceedings of InternationalSymposium on Circuits and Systems (ISCAS), 3, 1106–1109.

6. Malvar, H. (1982). Electronically controlled active filters with

operational transconductance amplifiers. IEEE Transactions onCircuits and Systems, CAS-29, 333–336. doi:10.1109/TCS.1982.

1085147.

7. Deliyannis, T., Sun, Y., & Fidler, J. K. (1999). Continuous timeactive filter design (392 pp). Boca Raton: CRC Press.

8. Sun, Y. (2002). Design of high frequency integrated analoguefilters. London: IEE Press.

9. Keskin, A. U., & Biolek, D. (2006). Current mode quadrature

oscillator using current differencing transconductance amplifier.

IEE Proceedings. Circuits, Devices and Systems, 153, 214–218.

doi:10.1049/ip-cds:20050304.

10. Chaiya Tanaphatsiri Winai Jaikla and Montree Siripruchyanun.

(2008). An electronically controllable voltage-mode first-order

all-pass filter using only single CCCDTA. In InternationalSymposium on Communications and Information Technologies(ISCIT 2008), 21–23 Oct 2008, Vientiane, Lao PDR.

11. Toumazou, C., Moschytz, G., & Gilbert, B. (2002). Trade-offs inanalog circuit design: The designer’s companion. Boston: Klu-

wer Academic Publishers.

12. Peterson, K. D., Nedungadi, A., & Geiger, R. L. (1987).

Amplifier design consideration for high frequency monolithic

filters. In Proceedings of European Conference on Circuit Theoryand Design (ECCTD) (pp. 321–326).

13. Biolek, D., Biolkova, V., & Kolka, Z. (2007). Universal current-

mode OTA-C KHN biquad. In The International Conference onElectronics, Circuits, and Systems (ICECS 2007), Venice, pp.

289–292.

14. Fabre, A., Saaid, O., Wiest, F., & Boucheron, C. (1996). High

frequency applications based on a new current controlled con-

veyor. IEEE Transaction Circuits Systems I, 43(2), 82–91. doi:

10.1109/81.486430.

15. Frey, D. R. (1993). Log-domain filtering: an approach to current-

mode filtering. IEE Proceedings. Part G. Circuits, Devices andSystems, 140(6), 406–416.

Bilgin Metin received the B.Sc.

degree in Electronics and Com-

munication Engineering from

Istanbul Technical University,

Istanbul, Turkey in 1996 and

the M.Sc. and Ph.D. degrees

in Electrical and Electronics

Engineering from Bogazici Uni-

versity, Istanbul, Turkey in 2001

and 2007, respectively. He is

currently an Assistant Professor

in the Management Information

Systems Department of Bogazici

University. His research interests

include continuous time filters, analog signal processing applications,

current-mode circuits, computer networks and network security. He

was given the best student paper award of ELECO’ 2002 conference in

Turkey. He has over 25 publications in scientific journals or conference

proceedings.

Kirat Pal was born in Aligarh,

India on 20th August 1950. He

received the B.Sc & M.Sc

degree from the Aligarh Muslim

University in 1972 and 1977

respectively and the Ph.D.

degree from the University of

Roorkee (Presently Indian Insti-

tute of Technology, Roorkee) in

1982. He joined University of

Roorkee as a scientific officer in

1979 and worked in various

capacities as lecturer, reader and

at present holds the post of

Associate Professor in Earthquake Engineering Dept. of Indian Insti-

tute of Technology Roorkee. His main research interests are analog

circuits and signal processing, transducers, seismological instrumen-

tation and digital image processing. He has authored more than 100

research papers in the above areas in national, international journals

and conferences.

212 Analog Integr Circ Sig Process (2009) 60:205–213

123

Shahram Minaei received the

B.Sc. degree in Electrical and

Electronics Engineering from

Iran University of Science and

Technology, Tehran, Iran, in

1993 and the M.Sc. and Ph.D.

degrees in electronics and com-

munication engineering from

Istanbul Technical University,

Istanbul, Turkey, in 1997 and

2001, respectively. He is cur-

rently an Associate Professor in

the Department of Electronics

and Communication Engineering, Dogus University, Istanbul, Tur-

key. He has over 80 publications in scientific journals or conference

proceedings. His current field of research concerns current-mode

circuits and analog signal processing. Dr. Minaei served as a reviewer

for a number of international journals and conferences. He is a senior

member of the IEEE.

Oguzhan Cicekoglu received

the B.Sc. and M.Sc. degrees

from Bogazici University and

the PhD. degree from Istanbul

Technical University all in

Electrical and Electronics Engi-

neering in 1985, 1988 and 1996,

respectively. He served as lec-

turer at the School of Advanced

Vocational Studies Electronics

Prog. of Bogazici University

where he held various adminis-

trative positions between 1993

and 1999. He has also given

lectures at the Turkish Air Force

Academy. He was with the Biomedical Engineering Institute of

the Bogazici University between 1999 and 2001. He is currently a

professor at the Electrical and Electronics Engineering Department of

the same University. Oguzhan Cicekoglu served in organizing and

technical committees of many national and international conferences.

He was the Guest co-editor of a Special Issue of the Journal Analog

Integrated Circuits and Signal Processing Journal. One of the publi-

cations he co-authored in IEEE Transactions on Circuits and Sytems

II-Analog and Digital Signal Processing is among the top cited papers

listed in IEEE Circuits and Systems Society web page. He received the

Research Excellence Award of Bogazici University Foundation in

2004. Currently he serves as the Associate Editor of International

Journal of Electronics (SCI). His current research interests include

analog circuits, active filters, analog signal processing applications and

current-mode circuits. He is the author or co-author of about 150

papers published in scientific journals or conference proceedings and

conducts review in numerous journals including Analog Integrated

Circuits and Signal Processing, IEEE CAS-I, IEEE CAS-II, Interna-

tional Journal of Electronics, ETRI Journal, IEE Proceedings Pt.G and

others. Oguzhan Cicekoglu is a member of the IEEE.

Analog Integr Circ Sig Process (2009) 60:205–213 213

123