RADIOENGINEERING, VOL. 17, NO.

4, DECEMBER 2008 33
A Simple Current-Mode Quadrature Oscillator Using
Single CDTA
Winai JAIKLA
1
, Montree SIRIPRUCHYANUN
2
, Josef BAJER
3
, Dalibor BIOLEK
3,4
1
Dept. of Electric and Electronic, Faculty of Industrial Technology, Suan Sunandha Rajabhat University, Dusit, Bangkok,
10300, Thailand
2
Dept. of Teacher Training in Electrical Engineering, Faculty of Technical Education, King Mongkuts University of
Technology North Bangkok, Bangsue, Bangkok, 10800, Thailand
3
Dept. of Electrical Engineering, Faculty of Military Technology, University of Defense Brno, Kounicova 65,
612 00 Brno, Czech Republic
4
Inst. of Microelectronics, Faculty of Electrical Engineering and Communications, Brno University of Technology,
Udolni 53, 602 00 Brno, Czech Republic

[email protected], [email protected], [email protected], [email protected]

Abstract. This article presents a simple current-mode
quadrature oscillator using a single Current Differencing
Transconductance Amplifier (CDTA) as the active element.
The oscillation condition and oscillation frequency can be
electronically controlled. The circuit structure is very
simple, consisting of merely one CDTA, one resistor and
two capacitors. The proposed circuit is suitable for IC
architecture. The PSpice simulation and experimental
results are shown, and the results agree well with the
theoretical assumptions.
Keywords
Oscillator, CDTA, current-mode.
1. Introduction
Oscillator is an important basic building block, which
is frequently employed in electrical engineering applica-
tions. Among several kinds of oscillator, the quadrature
oscillator is widely used because it can offer sinusoidal
signals with 90 phase difference, for example, in
telecommunications for quadrature mixers and single-
sideband modulators [1]. Presently, the current-mode
technique has been more popular than the voltage-mode
technique. This is due to the requirements in low-voltage
environments such as portable and battery-powered
equipment. The currentmode technique is ideally suited
for this purpose. Today, there is a growing interest in
synthesizing current-mode circuits because of their
potential advantages such as larger dynamic range, higher
signal bandwidth, greater linearity, simpler circuitry, and
lower power consumption [2].
A reported 5-terminal active element, namely Current
Differencing Transconductance Amplifier (CDTA) [3],
seems to be a versatile component in the realization of
a class of analog signal processing circuits, particularly
analog frequency filters [3], [4]. It is actually a current-
mode element whose input and output signals are currents.
In addition, the output current of the CDTA can be
electronically adjusted. Besides, a modified version of
CDTA, whose parasitic resistances at two current input
ports can be electronically controlled, has been proposed in
[5]. This circuit is called Current Controlled Current
Differencing Transconductance Amplifier (CCCDTA).
Another CDTA modification, called ZC-CDTA (Z Copy
CDTA) is proposed in [6], providing a copy of the current
flowing to the z terminal. This copy can be used as
an output signal for driving an independent load.
Several implementations of oscillator employing
CDTAs or CCCDTAs have been reported in the literature
[7-13]. Unfortunately, these circuits suffer from one or
more of the following weaknesses: They use more than one
CDTA or CCCDTA and an excessive number of passive
elements, which is not convenient for IC fabrication. In
addition, some reported circuits use a multiple-output
CDTA or CCCDTA. Consequently, the circuits become
more complicated.
The purpose of this paper is to introduce a current-
mode quadrature oscillator, based on a single CDTA. The
oscillation condition and oscillation frequency can be ad-
justed electronically. The circuit construction consists of
one CDTA, one resistor, and two capacitors. Finally, this
oscillator has been built by means of a CDTA chip which
has been fabricated in the CMOS technology [14]. The
PSpice simulation and the experimental results correspond
to the theoretical analyses.
34 W. JAIKLA, M. SIRIPRUCHYANUN, J. BAJER, D. BIOLEK, A SIMPLE CURRENT-MODE QUADRATURE OSCILLATOR
2. Principle of Operation
2.1 Current Differencing Transconductance
Amplifier (CDTA)
Since the proposed circuit is based on the CDTA,
a brief review of CDTA is given in this Section. The
characteristics of the ideal CDTA are represented by the
following hybrid matrix:

z
x
n
p
m x
z
n
p
V
V
I
I
g I
I
V
V
0 0 0
0 0 1 1
0 0 0 0
0 0 0 0
. (1)
In general, CDTA can contain an arbitrary number of x
terminals, providing currents I
x
of both directions. As an
example, the symbol and the equivalent circuit of the CDTA
with a pair of x+ and x- terminals are illustrated in Fig. 1(a)
and (b), respectively.

n p z
I I I =


z m x
V g I =

z
V

CDTA
n
p
n
I
p
I
z
CDTA
p
n x-
z
x+
x+
x-
p
V
n
V

(a) (b)
Fig. 1. CDTA (a) schematic symbol, (b) equivalent circuit.
2.2 Proposed Circuit
The proposed quadrature oscillator is designed by
cascading a first-order all-pass filter and a non-inverting
lossless integrator as shown in Fig. 2. Based on this block
diagram, the single-CDTA quadrature oscillator can be
implemented according to Fig. 3. In order to utilize the
current through the capacitor C
2
, an auxiliary zc terminal is
used. The internal current mirror provides a copy of the
current flowing out of the z terminal to the zc terminal [6].
The characteristic equation of the proposed oscillator
in Fig. 3 can be expressed as follows:
0 ) (
1 2 2 1
2
= + +
m m
g R g C C s R C C s . (2)
From Eq. (2), it can be seen that the proposed circuit can
produce oscillations if the oscillation condition is fulfilled:
1
2
C
C
R g
m
=
. (3)
For example, this condition can be achieved by
setting
2 1
C C = and
. (4) R g
m
/ 1 =

1
1
1
1
+

R sC
R sC
2
sC
g
m
I
o1

I
o2


Fig. 2. Block diagram of the quadrature oscillator.

CDTA
p
n x-
z
x+
zc
C
1
C
2
R
I
o1
I
o1
I
o2
I
o2
all-pass filter lossless integrator

Fig. 3. Proposed CDTA-based quadrature oscillator.
Then the characteristic equation of the system becomes
0
2 1
2
= +
m
g R C C s . (5)
From (5), the oscillation frequency is as follows:
R C C
g
m
osc
2 1
=
. (6)
From the circuit in Fig. 3, the current transfer function
from I
o1
to I
o2
is
2 1
2
) (
) (
sC
g
s I
s I
m
o
o
=
. (7)
For sinusoidal steady state, equation (7) becomes

=
90
2 1
2
) (
) (
j m
o
o
e
C
g
j I
j I

. (8)
The phase difference between I
o1
and I
o2
is
= -90, (9)
ensuring that the currents I
o2
and I
o1
are in quadrature.
For the oscillation frequency, with regard to (6),
equation (8) gives

=
90
2
1
1
2
) (
) (
j
m
osc o
osc o
e
C
C
R g
j I
j I

. (10)
Taking into account oscillation condition (3), one can
conclude that the oscillator will provide quadrature signals
of equal magnitudes.
All the active and passive sensitivities of the oscillator
are low as shown in (11):
2
1
,
2
1
, ,
2 1
= =
osc
m
osc
g R C C
S S

. (11)
RADIOENGINEERING, VOL. 17, NO. 4, DECEMBER 2008 35
3. Experimental Results
To prove the performance of the proposed circuit, the
oscillator was constructed using the CDTA chip. The
CDTA was designed and manufactured in the 0.7m
CMOS technology [14]. Its basic small-signal parameters
are as follows:
Input resistances of p and n terminals R
p
= 260 ,
R
n
= 250 .
Resistance of z terminal R
z
= 3.48 M.
Resistance of x+, x- terminals R
x
= 2.2 M.
Current gains from p to z terminal and from n to z
terminal are 1 and -1, respectively, with a 3-dB cutoff
frequency of 10 MHz.
Transconductance of the OTA amplifier is 1.22 mA/V
with a 3-dB cutoff frequency of 16 MHz.
The CDTA chip was manufactured with the fixed
transconductance, with x+ and x- terminals, and without
the auxiliary zc terminal. During the experiment, the cur-
rent through the x terminal was sensed via a I/V converter,
formed by an by operational amplifier OPA 2650 with
1 k feedback resistance. That is why the output voltage in
millivolts corresponded to the input current in micro-
ampers.
The quadrature oscillator was designed for an oscilla-
tion frequency of about 1MHz. The values of the external
elements were designed as follows:
R = 1/g
m
= 819 , C
1
= C
2
= 150 pF. (12)
According to (6), the theoretical value of the
oscillation frequency is 1.294 MHz.
Actually, the measured values (12), as used in the
experiment, were
R = 829 , C
1
= 153 pF, C
2
= 156 pF. (13)
Then the corresponding theoretical value of the
oscillation frequency is 1.257 MHz.
After substituting the values from (13) to (3), we can
see that the left-side value is greater than the right-side one.
In other words, the current loop-gain is greater than unity,
which is the condition of self-starting the oscillations.
The waveform measured, namely the output of the
I/V converter measuring the current I
o2
in Fig. 3, is shown
in Fig. 4. The magnitude is approximately 75 A. The
measured frequency of 962 kHz is less than the theoretical
value 1.257 MHz. In the following Section, an analysis of
the real effects leads to an explanation.
Fig. 5 shows the measured spectrum components of
the generated waveform. The magnitude of the second
harmonics is more than 56 dB below the magnitude of the
fundamental harmonics. The corresponding THD is
0.16 %. Increasing the value of R results in increasing the
loop-gain and thus increasing the magnitude and also the
distortion of generated signals.

Fig. 4. Measured output waveform in millivolts, corresponding
to the current I
o2
in microampers.
V [dB]
0
-20
-40
-60
-80
-100
0 1 2 3 4 5 6 7 8 9 1
f [MHz]
0

Fig. 5. Fourier analysis of generated waveform.
4. Analysis of Real Effects
The circuit operation is affected by linear and non-
linear influences. In this Section, an analysis of real effects
is done in the following structure:
a) Analysis of linear effects. The oscillation
frequency and the condition of oscillation are affected by
parasitic impedances of CDTA terminals, by DC current
transfers from p to z and from n to z terminals, which are
not generally equal to one in magnitude, and by their
frequency dependence. Also, the frequency characteristic
of the transconductance may have an effect.
The analysis of the above linear model of the oscilla-
tor can be performed in fully symbolic form in order to
obtain formulas for the oscillation condition and for the
frequency of oscillation. However, more complex high-
order models do not provide such results in a closed form.
That is why the symbolic analysis will be evaluated only
for several main influences, and the remaining one will be
analyzed separately via computer simulation.
36 W. JAIKLA, M. SIRIPRUCHYANUN, J. BAJER, D. BIOLEK, A SIMPLE CURRENT-MODE QUADRATURE OSCILLATOR
b) Analysis of nonlinear effects. There are two types
of important nonlinearities of manufactured CDTA:
nonlinear curves V
p
= f
p
(I
p
) and V
n
= f
n
(I
n
), and nonlinear
OTA characteristic I
x
=f
x
(V
z
). These nonlinearities are re-
sponsible for properly starting - up the oscillations, for
automatically achieving the oscillation condition in the
steady state, and they also affect the value of the oscillating
frequency.
A hand-and-paper analysis of such nonlinear effects
is practically impossible. Therefore, all these characteris-
tics were measured on the chip and transferred to PSpice as
nonlinear behavioral models, built via controlled sources
and the look-up table method. These models were then
completed with several small-signal parameters which
PSpice cannot extract from the above nonlinear curves.
Such a final model has been tested and its behavior has
been compared with real experiments. The coincidence was
excellent. This PSpice model of the CDTA chip helped us
with the analysis of nonlinear effects in the quadrature
oscillator. The simulations based on this model also pro-
vide waveforms which are not easily accessible via practi-
cal measurements.
4.1 Linear Effects
The most important small-signal real influences
which affect the circuit operation are modeled in Fig. 6.

CDTA
p
n x-
z
x+
zc
C
1
C
2
R
I
o1
I
o2
Rp
R
n
Rz
Rzc
Rx-
Rx+
RLz
RLx
Cz Czc
Cx+
Cx-

n
R'
C'
2

Fig. 6. Modeling linear effects in the oscillator.
In addition to the working passive elements R, C
1
, and C
2

(in thick lines), several parasitic impedances show up here.
The input Current Differencing Unit (CDU) is now
described by the equation
z p p n n
I I I = , (14)
where the parameters
p
and
n
are current transfer values,
deviating from one, depending on the internal circuit con-
struction. R
Lz
and R
Lx
are loading impedances for sensing
the output currents I
o1
and I
o2
. In conjunction with the R
zc
,
C
zc
, or R
x-
, C
x-
pair they form a frequency-dependent cur-
rent divider of terminal current I
zc
or I
x-
. Making the influ-
ence of these parasitic impedances negligible, one should
choose the load resistances as low as possible. Throughout
the following test, we follow this assumption. Then the
remaining non-ideal factors will be as follows:
General values of
p
and
n.
. We will take them into
account for the symbolic analysis in order to evaluate
this influence. However, for the manufactured CDTA
chip, these values are of unity-value with sufficiently
high precision.
Nonzero R
p
and R
n
resistances. R
n
is in series to R and
thus it does not change the character of this working
impedance. The R
n
influence can be compensated by
decreasing the R value. This is not the case of R
p
,
which acts in series to C
1
.
Parasitic impedance of the z terminal. C
z
can be con-
sidered a part of C
2
. However, R
z
changes the type of
the impedance, which should normally be of purely
capacitive character.
Parasitic impedance of the x+ terminal. The x+ termi-
nal is connected to low-impedance p and n terminals
through C
1
and R elements. For low frequencies, the
corresponding impedance level is given by R (kilo-
Ohms), which is much smaller than R
x+
. For the high-
frequency region, this impedance level is determined
by R
p
(hundreds of Ohms), but C
x+
can cause a leak-
age of the I
x+
current to the ground. This capacitive
part of parasitic impedance increases the order of the
oscillator model from the original value 2 to 3. This
fact considerably complicates the evaluation of the
symbolic analysis. That is why we will consider this
part of the model in the second stage of the analysis.
Frequency dependence of
p
and
n
transfer coeffi-
cients and of OTA g
m
. According to the information
in Section 3, the corresponding cutoff frequencies are
10 MHz and 16 MHz. For the oscillation frequencies
near these values, the effect can be considerable.
However, the order of circuit model is also increased.
That is why we will analyze these factors separately,
after the symbolic analysis.
Considering the above factors and using the notation
z z x x z n
R G R G C C C R R R / 1 , / 1 , ,
2 2
= = + = + =
+
,
the symbolic analysis of the model in Fig. 6 leads to the
following equations:
The modified oscillation frequency
) (
) 1 (
2 1 x p p
z x m n
osc
G R R R R C C
G G R g
+ +
+ +
=

. (15)
The modified condition of oscillation
z x p p x n p p m
G G R R R R G R
C
C
a R R g ) ( ) 1 ( ) (
1
2
+ + + + =
(16)
RADIOENGINEERING, VOL. 17, NO. 4, DECEMBER 2008 37
Setting the numerical values of the parameters from
Section 3 and a short re-arrangement lead to the conclusion
that several terms in (15) and (16) can be neglected. The
simplified versions of (15) and (16) are as follows:

+ +
=
+

) ( ) (
2 1 2 1 n p
m n
p
m n
osc
R R R C C
g
R R C C
g

R
R R
a
a
n p
osc
osc
+
+ = 1 ,

, (17)

1
2
) (
C
C
a R R g
n p p m


1
2
) (
C
C
a R R g R g
n p p n m p m
+
. (18)
Equation (17) shows that the oscillation frequency
decreases below its theoretical value due to nonzero para-
sitic resistances R
p
and R
n
. For R
p
= 260 , R
n
= 250 , and
R = 829 (see Eq. 13), the frequency should decrease
1.172 times, i.e. from the theoretical value 1.257 MHz to
1.072 MHz. However, the real experiments yielded the
value 962 kHz (see Section 3). Obviously, there are some
additional influences that decrease the oscillation fre-
quency and which were not taken into account throughout
the above analysis. To find these influences in the frame of
the linear models, the SNAP program [15] for symbolic
analysis was used for additional modeling of the parasitic
impedance of x+ terminal, and for modeling the cutoff
frequencies of
p
,
n
, and g
m
. The analysis has confirmed
that all these factors decrease the oscillation frequency. For
example, parasitic capacitance C
x+
of only 1 pF decreases
the frequency to 980 kHz, and the cutoff frequencies of
p
,

n
, and g
m
decrease it by another 15 kHz. We can conclude
that such additional modifications of oscillation frequency
are approximately by one order less relevant than the
changes described by (17).
Equation (18) demonstrates that for the identical
values of parasitic resistances R
p
and R
n
and for
p
=
n
=1,
the oscillation condition is practically not affected by the
above effects.
Another analysis of the model in Fig. 6 under the
above simplifications leads to the conclusion that the ratio
of current I
o2
and current I
o1
is still described by (8), thus
their phase shift is 90 degrees. However, their magnitudes
will not be identical any more. Utilizing (8), (17) and (18)
and a short re-arrangement yield the following formula:


+

90
1
2
) ( ) (
) (
j
n p p n
p
osc o
osc o
e
R R
R R
j I
j I

. (19)
For ideal values of
p
=
n
=1, the magnitude ratio
b = abs(I
o2
/I
o1
) is as follows:
279 . 1 =
+
+ +
= &
p n
p n
R R R
R R R
b . (20)

The above results can be summarized in a practical
formula: If the oscillation frequency decreases a times due
to parasitic linear effects, where a is given by (17), then the
magnitudes of generated quadrature signals will no longer
be equal but their ratio will be b, where b is given by (19)
or (20). This formula will be confirmed in the following
Subsection via PSpice simulation.
4.2 Nonlinear Effects
Nonlinear effects were studied by means of PSpice
simulation, utilizing nonlinear models of the CDTA chip,
briefly described in Section 4. Two CDTA models were
used subsequently, first model No. 1 and then model No. 2.
Model No. 1: A simplified model of the CDU con-
sists of linear R
p
and R
n
resistances, controlled source for
modeling the difference of input currents, the z-terminal
parasitic resistance, and single-pole frequency response
with 3dB cutoff frequency. For the internal OTA, a full
nonlinear model of the I
x
= f
x
(V
z
) from Fig. 7, resulting
from the data measured [16], was used. In addition, the
output resistances of x-terminals and the 3dB cutoff fre-
quency were also modeled.
500uA
0A
-500uA
0V
Vz
-1V -0.5V 0.5V 1V
Ix
Ix+ Ix-

Fig. 7. Measured OTA characteristics, transferred to PSpice.

200mV
0V
-200mV
-400mV
-350uA -200uA 0A 200uA
Vp
Vp, Vn
Ip, In
Vn

Fig. 8. Measured input characteristics of Current Differencing
Unit, transferred to PSpice.
38 W. JAIKLA, M. SIRIPRUCHYANUN, J. BAJER, D. BIOLEK, A SIMPLE CURRENT-MODE QUADRATURE OSCILLATOR
Model No. 2: A full nonlinear model of the CDU is
implemented here, utilizing the measured nonlinear
characteristics V
p
= f
p
(I
p
) and V
n
= f
n
(I
n
) [16], [17], see
Fig. 8. The other parts of the model are identical with
model No. 1.
The results of the steady-state analysis by means of
model No. 1 are shown in Fig. 9. Starting up the oscilla-
tions was attained for R = 864 , which is in good agree-
ment with real measurements (829 ). For such a resis-
tance value, the theoretical oscillation frequency is 1.224
MHz. The simulated value is 960.3 kHz, i.e. 1.275 times
smaller. It is obvious that the generated waveforms are not
of equal magnitudes. Their ratio was measured as 1.128,
which roughly corresponds to the ratio 1.173, computed
from (20).
As shown in Fig. 9, the peak value of the voltage at z
terminal is below 200 mV. It is obvious from Fig. 7 that the
transconductance amplifier works in nearly linear regime
for such a voltage swing. Its equivalent transconductance is
decreasing when increasing the voltage swing. The equilib-
rium, i.e. fulfilling the oscillation condition, is established
for a concrete voltage swing. That is why the automated
adjustment of the oscillation condition is due to the nonlin-
earity of the OTA stage.
Fig. 9 illustrates an interesting phenomenon: in spite
of different magnitudes of output currents I
o1
and I
o2
, the
magnitudes of currents I
p
and I
n
are equal. It can be easily
proved that I
p
and I
n
are also orthogonal. Their magnitudes
are approximately 150 A.
Ip
In
Io2
Io1
Vz
200mV
-200mV
0
200uA
0
-200uA
200uA
0
-200uA
0 time 10us 5us

Fig. 9. Simulated steady-state oscillations, CDTA model No. 1,
f
osc
=960 kHz.
The simulation results for CDTA model No. 2 is given in
Fig. 10. The parameters of the passive elements are the
same as for the previous simulation. The oscillation fre-
quency is now 964 kHz, which is in very good agreement
with the real measurement (962 kHz). Also, the magnitude
of I
o2
(73 A) now corresponds to the reality. Note that the
signal levels are now lower than without detailed modeling
of the CDU. This is due to the nonlinearity of p and n in-
puts. The R
p
and R
n
equivalent resistances are now depend-
ent on signal levels, and the nonlinearity of CDU provides
the automatic adjustment of the oscillation condition. It
should be also noted that a DC offset appears in the wave-
forms, particularly in I
o2
, which is caused by the offset of
DC characteristics of the CDU (see Fig. 8).

Vz
100mV
-100mV
0
Ip
In
50uA
0
-50uA
Io1
100uA
0
-100uA
Io2
0 time 10us 5us

Fig. 10. Simulated steady-state oscillations, CDTA model No. 2,
f
osc
= 964 kHz.
4.3 Elimination of Real Effects
It is obvious from (17) and (20) that both real effects,
i.e. decreasing the oscillation frequency and different mag-
nitudes of generated quadrature signals, can be eliminated
via increasing the working resistance R to be at least by
one order higher than the parasitic resistances R
p
and R
n
.
Let us suppose that R will be increased k times. In order to
preserve the designed value of the oscillation frequency (6)
and also to fulfill the condition of oscillation (3), several
possible operations can accompany the R increase. Two
relevant operations are summarized below:
decreasing C
1
k times, or
increasing g
m
k times, increasing C
2
k times, decreas-
ing C
1
k times.
The second method cannot be applied to the oscillator
with CDTA chip, which has a fixed transconductance. Let
us increase R ten times, to 8.64 k and decrease C
1
ten
times, to 15.3 pF. The results of PSpice simulation for
model No. 2 is in Fig. 11. Note that the oscillation
frequency is now near the theoretical value, namely
1.171 MHz, and the quadrature signals are of almost equal
magnitudes.
RADIOENGINEERING, VOL. 17, NO. 4, DECEMBER 2008 39
Vz
200mV
-200mV
0
Ip
In
200uA
0
-200uA
Io1
200uA
0
-250uA
Io2
0 time 10us 5us

Fig. 11. Simulated steady-state oscillations, CDTA model No. 2,
with elimination of real effects, f
osc
= 1.171 MHz.
5. Conclusion
A simple quadrature oscillator based on one Current
Differencing Transconductance Amplifier (CDTA), two
capacitors and one resistor is described. The proposed
oscillator, designed for the megahertz range, was con-
structed using a CDTA chip, which was fabricated in the
CMOS technology. The measurements confirmed the
functionality of the principle. Two problems were dis-
cussed: 1) The oscillation frequency is lower than the theo-
retical value. 2) The quadrature output currents are of dif-
ferent magnitudes although the oscillator was designed for
equal magnitudes. Theoretical analysis shows the common
reason for both problems: the finite values of R
p
and R
n

resistances of the CDTA. A quantitative description of
these phenomena is given in equations (17) and (20).
Based on this knowledge, a simple procedure is designed
how to overcome the above difficulties.
For a practical implementation of the proposed to-
pology of quadrature oscillator, it is useful if the CDTA
has at least two x-terminals and one so-called zc terminal,
which provides a copy of the z-terminal current. For an
easy elimination of the above real influences, the current
differencing unit of the CDTA should be designed with the
R
p
and R
n
parasitic resistances as low as possible. The
proposed value is at least one order lower than the value of
the resistance of the working resistor connected in series to
the n terminal.
Acknowledgements
The research described in the paper was supported by
the Czech Grant Agency under grant No. 102/08/0784, and
by the research programmes of BUT No.
MSM0021630503 and UD Brno No. MO FVT0000403.
References
[1] KHAN, I. A., KHAWAJA, S. An integrable gm-C quadrature
oscillator. International Journal of Electronics, 2000, vol. 87, no. 1,
p. 1353-1357.
[2] TOUMAZOU, C., LIDGEY, F. J., HAIGH, D. G. Analogue IC
Design: The Current-Mode Approach. London: Peter Peregrinus,
1990.
[3] BIOLEK, D. CDTA Building block for current-mode analog signal
processing. In Proceedings of the European Conference on Circuit
Theory and Design 2003. Krakow (Poland), 2003, p. 397400.
[4] BIOLEK, D. BIOLKOVA, V. Universal biquads using CDTA
elements for cascade filter design. In Proceeding of the 13
th

International Multiconference on Circuits, Systems, Communications
& Computers (CSCC2003). Corfu (Greece), 2003, p. 812.
[5] SIRIPRUCHYANUN, M., JAIKLA, W. Realization of current
controlled current differencing transconductance amplifier
(CCCDTA) and its applications. ECTI Trans. on Electrical
Engineering, Electronics, and Communications. 2007, vol. 5, no. 1,
p. 41-50.
[6] BIOLEK, D., BIOLKOVA, V., KOLKA, Z. Single-CDTA (Current
Differencing Transconductance Amplifoer) current-mode biquad
revisited. WSEAS Trans. on Electronics, Special Issue Modern
Circuit Components for Analogue Signal Processing and their
Applications. To be issued in 2008.
[7] JAIKLA, W., SIRIPRUCHYANUN, M. A versatile quadrature
oscillator and universal biquad filter using dual-output current
controlled current differencing transconductance amplifier. In Proc.
of the International Symposium on Communications and Information
Technologies 2006. Bangkok (Thailand), 2006, p. 1072-1075.
[8] BIOLEK, D., KESKIN, A. U., BIOLKOVA, V. Quadrature oscillator
using CDTA-based integrators. WSEAS Transactions on Electronics,
2006, vol. 3, no. 9, p. 463-469.
[9] JAIKLA, W., SIRIPRUCHYANUN, M. CCCDTAs-based versatile
quadrature oscillator and universal biquad filter. In Proceedings of
ECTI conference 2007, Chieng Rai (Thailand), 2007, p. 1065-1068.
[10] KESKIN, A. U., BIOLEK, D. Current mode quadrature oscillator
using current differencing transconductance amplifiers (CDTA). IEE
Proceedings-Circuits, Devices and Systems. 2003, vol. 153, no. 3,
p. 214218.
[11] BIOLEK, D. BIOLKOVA, V., KESKIN, A. U. Current mode
quadrature oscillator using two CDTAs and two grounded capacitors.
In Proceedings of the 5
th
WSEAS International Conference on System
Science and Simulation in Engineering, Tenerife, Canary Islands
(Spain), 2006, p. 368-370.
[12] TANGSRIRAT, W., TANJAROEN, W. Current-mode multiphase
sinusoidal oscillator using current differencing transconductance
amplifiers. Circuits Systems and Signal Processing, 2008, vol. 27,
p. 8193.
[13] TANGSRIRAT, W., TANJAROEN, W., PUKKALANUN, T.
Current-mode multiphase sinusoidal oscillator using CDTA-based
allpass sections. International Journal of Electronics and
Communications (AEU). 2008, online first.
[14] PROKOP, R., MUSIL, V. New modular current devices for true
current mode signal processing. Electronics, 2007, vol. 16, no. 4, p.
36-42.
[15] KOLKA, Z., BIOLEK, D. BIOLKOVA, V. Symbolic analysis of
linear circuits with modern active elements. WSEAS Trans. on
Electronics, Special Issue Modern Circuit Components for Analogue
Signal Processing and their Applications. To be issued in 2008.
[16] BAJER, J. CDTA applications for analog signal processing. Diploma
work, Faculty of Military Technologies, University of Defense,
Czech Republic, 2008, p. 65 (in Czech).
40 W. JAIKLA, M. SIRIPRUCHYANUN, J. BAJER, D. BIOLEK, A SIMPLE CURRENT-MODE QUADRATURE OSCILLATOR
[17] FOREJTEK, J. Measuring the CDTA and CCTA chip. Research
report for the Grant Agency of the Czech Republic, grant No.
102/05/0277, 2007 (in Czech).
About Authors...
Winai JAIKLA received the B. Tech. Ed. degree in Tele-
communication Engineering from King Mongkuts Insti-
tute of Technology Ladkrabang, Thailand, in 2002 and M.
Tech. Ed. in Electrical Technology from King Mongkuts
Institute of Technology, North Bangkok (KMITNB), in
2004. He has been with the Electric and Electronic Pro-
gram, Faculty of Industrial Technology, Suan Sunandha
Rajabhat University, Bangkok, Thailand since 2004. His
research interests include electronic communications, ana-
log signal processing and analog integrated circuits. He is
a member of ECTI (Thailand).
Montree SIRIPRUCHYANUN received the B. Tech. Ed.
degree in Electrical Engineering from King Mongkuts
Institute of Technology, North Bangkok (KMITNB), the
M.Eng. and D. Eng. degrees in Electrical Engineering from
King Mongkuts Institute of Technology Ladkrabang
(KMITL), Bangkok, Thailand, in 1994, 2000 and 2004,
respectively. He has been with the Faculty of Technical
Education, King Mongkuts University of Technology,
North Bangkok (KMUTNB), since 1994. Presently, he is
with the Department of Teacher Training in Electrical
Engineering as Associate Professor, KMUTNB. His re-
search interests include analog-digital communications,
analog signal processing, and analog integrated circuits. He
is a member of IEEE (USA), IEICE (Japan), and ECTI
(Thailand).
Josef BAJER was born in 1982. In 2005 and 2008, he
received the B.Sc. degree in Electrical and Special Aircraft
Equipments and the M.Sc. degree in Avionic Systems from
the University of Defense Brno (UDB), Czech Republic.
He is currently working towards the Ph.D. degree in Elec-
tronic and Weapons Systems at the Faculty of Military
Technologies UDB. His interests include analog and digital
signal processing and applications of modern active ele-
ments working in current and hybrid modes.
Dalibor BIOLEK received the M.Sc. degree in Electrical
Engineering from the Brno University of Technology,
Czech Republic, in 1983, and the Ph.D. degree in Elec-
tronics from the Military Academy Brno, Czech Republic,
in 1989. He is currently with the Department of EE, Uni-
versity of Defense Brno (UDB), and with the Department
of Microelectronics, Brno University of Technology
(BUT), Czech Republic. His scientific activity is directed
to the areas of general circuit theory, frequency filters, and
computer simulation of electronic systems. For years, he
has been engaged in algorithms of the symbolic and nu-
merical computer analysis of electronic circuits with a view
to the linear continuous-time and switched filters. He has
published over 250 papers and is the author of a book on
circuit analysis and simulation. At present, he is Professor
at BUT and UDB in the field of Theoretical Electrical
Engineering. Prof. Biolek is a member of the CAS/COM
Czech National Group of IEEE. He is also the President of
Commission C of the URSI National Committee for the
Czech Republic.