2008 IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 58, NO.

6, JUNE 2009
Bandwidth of Current Transformers
Nisha Kondrath, Student Member, IEEE, and Marian K. Kazimierczuk, Fellow, IEEE
AbstractBandwidth is one of the important performance pa-
rameters of current transformers. Transformers are required to
have adequate bandwidth to regenerate the measured current,
apart from uncertainties that are caused by the nonsinusoidal
waveforms along with harmonics. An expression for the transresis-
tance of the current transformer in terms of parasitic components
is derived. Approximate expressions for lower cutoff frequency,
upper cutoff frequency, and bandwidth are derived from the
equivalent models of the current probe in the corresponding
frequency range. The expressions are validated by experimental
results for a current transformer using a Ferroxcube ferrite core
528T500-4C4. An overview of the effects of different parasitic
components, such as magnetizing inductance, leakage inductance,
and stray capacitance, on the bandwidth of a current transformer
is presented.
Index TermsBandwidth, current probe, current transformer,
cutoff frequencies, frequency response, transformer circuit model,
transformer transfer function, transresistance.
I. INTRODUCTION
D
EVICES for measuring current waveforms or peak cur-
rent values are called current probes, current transform-
ers, current measuring transformers, or current-monitoring
transformers [1][14]. They are simple and inexpensive and
provide conductive electrical isolation from the rest of the cir-
cuit, particularly for high-voltage applications. Current probes
can be classied into dc current probes and ac current probes.
The dc current probes are based on the Hall effect, and the ac
current probes are based on the magnetic coupling principle.
This paper deals with ac current probes in the linear region.
Current waveforms that are observed by the current trans-
formers can be sinusoidal, as in resonant power converters,
or nonsinusoidal, as encountered in pulsewidth-modulated
(PWM) power converters, where semiconductor devices are
operated as switches. These nonsinusoidal waveforms, along
with the signicant amount of harmonics, introduce large un-
certainties resulting in signicant measurement errors or mal-
functioning of the devices that are connected to the waveforms.
Therefore, a current transformer is desired to have adequate
bandwidth so that it can regenerate the measured current, apart
from the uncertainties that are caused by the nonsinusoidal
waveforms along with the harmonics. Hence, one of the im-
portant performance parameters of the current transformer is its
Manuscript received October 12, 2007; revised March 20, 2008. First
published October 21, 2008; current version published May 13, 2009.
The Associate Editor coordinating the review process for this paper was
Dr. Gerd Vandersteen.
The authors are with the Department of Electrical Engineering, Wright
State University, Dayton, OH 45435 USA (e-mail: [email protected];
[email protected]).
Color versions of one or more of the gures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identier 10.1109/TIM.2008.2006134
Fig. 1. Physical representation of a current probe.
bandwidth [3][6]. Models of current transformers are required
for simulations of circuits that contain current probes, e.g.,
for simulations of PWM converters with current-mode control.
Circuit models are useful for simulators such as Spice, and
transfer function models over a wide frequency range are used
in simulators such as MATLAB.
The objectives of this paper are to derive expressions for the
transfer function, the lower and upper cutoff frequencies, and
the bandwidth of an ac current transformer and to discuss the
effects of parasitic components on the bandwidth. Section II
describes the construction of a current probe. The principle of
operation of the current transformer is given in Section III.
Expressions for transresistance, lower cutoff frequency, upper
cutoff frequency, and bandwidth of a current probe are derived
in Section IV. Validation of the derived equations with exper-
imental results and the effects of the parasitic components on
the bandwidth are presented in Sections V and VI, respectively.
Conclusions are given in Section VII.
II. CONSTRUCTION OF THE CURRENT TRANSFORMER
A typical ac current transformer arrangement is shown in
Fig. 1. AC current transformers are based on the magnetic cou-
pling principle. A typical ac current transformer (current probe)
consists of a toroidal ferromagnetic core, on which a copper
wire of n turns is wound [2]. This winding is terminated with
a low-inductance sense resistor or burden resistor R. Thin-lm
chip resistors exhibit low inductive and capacitive reactance.
The conductor carrying the measured time-varying current acts
as the primary of the current transformer. The toroid can be
clamped around the current-carrying conductor. The winding
wound on the toroid acts as the secondary of the current
transformer. The burden resistance is selected depending on the
sensitivity that is required.
For better performance, current transformer cores are de-
sired to have high permeability, high resistivity, low hysteresis
and eddy current losses, and, thereby, wide bandwidth. To
have a low value for the lower cutoff frequency, cores with
0018-9456/$25.00 2008 IEEE
KONDRATH AND KAZIMIERCZUK: BANDWIDTH OF CURRENT TRANSFORMERS 2009
high permeability are needed. The most commonly used are
nanocrystalline cores and molypermalloy cores.
For experimentation purposes, ferrite toroidal core 528T500-
4C4, which is manufactured by Ferroxcube, was chosen. The
construction parameters of the chosen core were as follows:
external diameter D
ext
= 38.9 mm, internal diameter D
int
=
19.6 mm, height H = 12 mm, toroid mean length l
c
= 8.49 cm,
core cross-sectional area A
c
= 1.17 cm
2
,
r
= 125, and induc-
tance per 1000 turns A
L
= 218 mH/1000 turns. The number
of turns in the secondary winding was chosen to be 40, and
a copper AWG22 wire of radius 0.322 mm, which has a total
diameter of 1 mmwith the insulation, was used as the secondary
winding. A sense resistor with a measured value of 10.5 was
chosen.
III. PRINCIPLE OF OPERATION OF
AC CURRENT TRANSFORMERS
The principle of operation of ac current transformers is
based on the combination of Amperes and Faradays laws [2].
According to Amperes law, the measured current i through the
primary conductor produces a circular time-varying magnetic
eld H

. The purpose of the ferromagnetic toroidal core is to

contain the majority of the magnetic eld within it. The mag-
netic eld passes through the secondary winding and, according
to Faradays law, induces a voltage v
s
across the terminals of the
secondary winding. Voltage v
s
causes current i
s
through sense
resistor R. The measured current is given by
i =
n
R
v
s
. (1)
The current probe can be calibrated by passing a known value
of current through the primary conductor and measuring the
voltage across the sense resistor. Sensitivity S
p
of the current
probe is expressed by
S
p
=
v
s
i
=
R
n
(V/A). (2)
For the chosen example, the sensitivity was
S
p
=
R
n
=
10.5
40
= 0.2625 (V/A). (3)
The current owing through the sense resistor is much lower
than the measured current i.
IV. BANDWIDTH
One of the most important performance parameters of
the current transformer is its bandwidth. An expression for
the transresistance over a wide frequency range is needed
for the simulation of circuits containing current probes. The
transresistance is derived from the transformer model. Approx-
imate expressions for lower cutoff frequency f
L
, upper cutoff
frequency f
H
, and bandwidth BW are derived using the equiv-
alent models of the current transformer for the corresponding
frequency range.
Fig. 2. Current transformer. (a) Physical circuit. (b) Transformer model with
parasitic components. (c) Transformer model with the current source and the
primary-side capacitance reected to secondary.
A. Current Transformer Model
As already mentioned, most often, the current involved will
be nonsinusoidal and may contain a wide range of harmon-
ics. To reproduce the waveform at the secondary side with a
proportional magnitude and the same shape as in the primary,
the device must have adequate bandwidth. The physical circuit,
as well as the equivalent circuit, of the current transformer is
shown in Fig. 2, where C
/
p
is the capacitance of the primary
winding, L
m
is the magnetizing inductance on the secondary
side, R
c
is the core parallel equivalent resistance, L
l
is the leak-
age inductance of the secondary, R
s
is the secondary winding
resistance, C is the stray capacitance, R is the sense resistor,
and L
R
is the inductance of the sense resistor. The models
shown in Fig. 2(b) and (c) are valid for a wide frequency range,
i.e., a low-frequency range, a mid-frequency range, and a high-
frequency range. It is assumed that the circuit is linear and is
operated in the nonsaturation region.
The magnetizing inductance of the secondary winding is
given by
L
m
=

r

0
n
2
A
c
l
c
=

r

0
n
2
A
c
2r
(4)
where r is the mean radius of the toroidal core. Using the deni-
tion given in the data sheet of the chosen core, the magnetizing
inductance is found as
L
m
=
n
2
A
L
10
6
= 0.349 mH. (5)
2010 IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 58, NO. 6, JUNE 2009
B. Transresistance
The transresistance of the current transformer is given by the
relation
R
m
(s) =
V
R
(s)
I(s)
(6)
where V
R
(s) and I(s) are the Laplace transforms of the voltage
across the sense resistor and the measured current, respectively.
Primary capacitance C
/
p
is very small in a current transformer
since the primary winding consists of only one turn. In addition,
it is reected to the secondary side, where it is reduced by a
factor of n
2
. Therefore, the effect of C
/
p
on the bandwidth is
negligible. Thus, after neglecting C
p
= C
/
p
/n
2
in Fig. 2(c) and
assuming that burden resistor R is noninductive
V
R
= I
Z2
Z
1
(7)
where
Z
1
=
1
sC
_
_
_
_
_
R =
R
sC
R +
1
sC
=
1
C
s +
1
RC
Z
2
= sL
l
+ R
s
. (8)
The current through impedance Z
2
is found as that given in
(9), shown at the bottom of the page.
Substituting (8) and (9) into (7), we obtain
V
R
=
I(s)
n

s
R
c
C
sR
c
_
s+
1
RC
_
+(sL
l
+R
s
)
_
s+
R
c
L
m
_
_
s+
1
RC
_
+
1
C
_
s+
R
c
L
m
_.
(10)
Therefore, the transresistance of a current transformer for a
wide frequency range is given by
R
m
(s) =
V
R
(s)
I(s)
=
R
c
nC
s
Den
vR
(11)
where
Den
vR
=L
l
_
s
3
+s
2
_
L
m
(L
l
+R
s
RC)+R
c
RC(L
m
+L
l
)
RCL
m
L
l
_
+ s
_
L
m
(R
s
+R+R
c
)+R
c
(L
l
+R
s
RC)
RCL
m
L
l
_
+
_
R
c
(R
s
+R)
RCL
m
L
l
__
.
(12)
Fig. 3. Equivalent circuit for determining the lower cutoff frequency. (a) Low-
frequency equivalent circuit. (b) Low-frequency dead circuit.
This equation can be used in the simulations of circuits con-
taining current transformers with simulators such as MATLAB.
An expression for transresistance, including C
p
, is given in the
Appendix.
From (11) and (12), it is seen that the expression for trans-
resistance is so complex that it is difcult to extract the
expressions for the lower cutoff frequency, the higher cut-
off frequency, or the bandwidth. Hence, simplied models of
the current transformer for low-frequency and high-frequency
ranges are obtained from the general model shown in Fig. 2(c),
and approximate expressions for the lower and upper cutoff
frequencies of the current transformer are derived.
C. Low-Frequency Response
The lower cutoff frequency is determined by the relative
values of the transformers secondary winding impedance and
its sense resistance. The low-frequency equivalent circuit [6]
is shown in Fig. 3(a). This equivalent circuit is derived from
the general model shown in Fig. 2(c) by neglecting leakage
inductance L
l
, sense resistor inductance L
R
, and stray capac-
itance C since L
l
and L
R
present very low reactances, and C
presents very high reactance at low frequencies. The reactance
I
Z2
=
I(s)
n
(sL
m
|R
c
)
(sL
m
|R
c
) + Z
2
+ Z
1
=
I(s)
n
sR
c
s+
R
c
L
m
sR
c
s+
R
c
L
m
+ (sL
l
+ R
s
) +
1
C
s+
1
RC
=
I(s)
n
sR
c
_
s +
1
RC
_
sR
c
_
s +
1
RC
_
+ (sL
l
+ R
s
)
_
s +
R
c
L
m
_
_
s +
1
RC
_
+
1
C
_
s +
R
c
L
m
_ (9)
KONDRATH AND KAZIMIERCZUK: BANDWIDTH OF CURRENT TRANSFORMERS 2011
of magnetizing inductance L
m
and core parallel resistance R
c
must be very large, so that the current through sense resistor
R is approximately equal to i/n. However, at low frequencies,
the reactance of the magnetizing inductance is comparable with
resistance R
Lm
= R
c
|(R + R
s
), which is seen by magnetizing
inductance L
m
. As a result, more current will ow through
L
m
as the frequency decreases. Hence, the current transformer
behaves like a rst-order high-pass lter in the low-frequency
range. From Fig. 3(a), the voltage across the sense resistance R
is given by
V
R
(s) =R
I(s)
n
sL
m
|R
c
(sL
m
|R
c
) + (R + R
s
)
=
I(s)R
n
R
c
R
c
+ R + R
s
s
s +
L
(13)
where

L
=
1

=
R
c
|(R + R
s
)
L
m
=
L
m
R
Lm
=
L
m
R
c
|(R + R
s
)
. (14)
Thus, the transfer function of the current transformer for low
frequencies is given by the transresistance
R
m
(s) =
V
R
(s)
I(s)
=
R
n
R
c
R
c
+ R + R
s
s
s +
L
. (15)
Therefore, the lower 3-dB cutoff frequency of the current
transformer is given by
f
L
=

L
2
=
1
2
=
R
c
|(R + R
s
)
2L
m
. (16)
This expression is valid only if the core does not enter satura-
tion. If R
c
R + R
s
, then the transresistance of the current
transformer can be approximated by
R
m
(s) =
V
R
(s)
I(s)
=
R
n
s
s +
L
. (17)
For the selected core, core constant L
c
/A
c
= 7.24 cm
1
, and
the resistivity of the ferrite material is
c
= 1000 cm.
Therefore
R
c
=
c
L
c
A
c
= 1000 7.24 = 7.24 k.
For the selected core with measured dimensions, as external
diameter D
ext
= 39.9 mm, internal diameter D
int
= 18.6 mm,
and height H = 14.7 mm, the length of one turn of the
secondary winding is obtained as
l
t
= 2
_
14.7 +
39.9 18.6
2
_
= 5.07 cm. (18)
Thus, the total length of the secondary winding is given by
l
w
= nl
t
= 40 5.07 = 202.8 cm. (19)
Fig. 4. Magnitude Bode plot of a current transformer for low frequencies.
Therefore, the dc and low-frequency secondary winding
resistance for a copper AWG22 wire with a radius of 0.322 mm
and resistivity of 1.724 10
6
cm is obtained as
R
s
=
Cu
l
w
A
w
=
1.724 10
8
202.8 10
2
(0.322 10
3
)
2
= 0.1073 .
(20)
Thus, with L
m
= 0.349 mH, R = 10.5 , R
s
= 0.1073 , and
R
c
= 7.24 k, we obtain
R
Lm
=
R
c
(R+R
s
)
R
c
+R+R
s
=
7.24 10
3
(10.5+0.1073)
7.2410
3
+10.5+0.1073
=10.59
(21)
and
f
L
=
R
Lm
2L
m
=
10.59
20.34910
3
=4.829 kHz. (22)
If the sense resistor is replaced by an active op-amp circuit
[7], the burden resistance can be neglected in (21), and as a
result, the lower cutoff frequency obtained is much lower.
Setting s = j, the transresistance of the current transformer
at low frequencies becomes
R
m
=
v
R
i
=
R
n
R
c
R
c
+ R + R
s
1
1 j
f
L
f
= [R
m
[e
jR
m
(23)
where
[R
m
[ =
R
n
R
c
R
c
+ R + R
s
1
_
1 +
_
f
L
f
_
2

R
m
= arctan
_
f
L
f
_
. (24)
Figs. 4 and 5 show the Bode plots for the designed current
transformer for low frequencies with L
m
= 0.349 mH, R
c
=
7.24 k, R
s
= 0.1073 , and R = 10.5 .
2012 IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 58, NO. 6, JUNE 2009
Fig. 5. Phase Bode plot of a current transformer for low frequencies.
Let us consider the step response of the current transformer
for low frequencies. The step change in the measured current is
given by
i(t) = Iu(t) (25)
which, in the s-domain, becomes
I(s) =
I
s
. (26)
Hence, the voltage across the sense resistor in the s-domain
is expressed by
V
R
(s) = R
m
(s)I(s) =
R
n
R
c
R
c
+ R + R
s
I
s +
1

(27)
which gives the step response in the time domain as
v
R
(t) = L
1
V
R
(s) =
R
n
R
c
I
R
c
+ R + R
s
e

= v
R
(0)e

(28)
where
v
R
(0) =
R
n
R
c
I
R
c
+ R + R
s
. (29)
Fig. 6 shows the step response of the current transformer
for low frequencies. The step response in (28) can be approxi-
mated as
v
R
(t) = v
R
(0)
_
1
t

_
, for t (30)
from which
v
R
(t) = v
R
(0) v
R
= v
R
(0) v
R
(0)
t

. (31)
Therefore, the differentiation of voltage v
R
leads to
Sag =
v
R
v
R
(0)
=
t

= t
R
c
|(R + R
s
)
L
m
. (32)
Fig. 6. Step response of a current transformer for low frequencies.
Fig. 7. High-frequency equivalent circuit.
D. High-Frequency Response
Unlike the lower cutoff frequency, higher cutoff frequency
f
H
is determined by the effects of leakage inductance L
l
and
stray capacitance C. Fig. 7 shows the high-frequency equivalent
circuit for the current transformer. This model is obtained from
the equivalent circuit shown in Fig. 2(c) by neglecting L
m
, R
s
,
and L
R
. The reactance of magnetizing inductance L
m
is much
higher than core resistance R
c
at high frequencies. Secondary
winding resistance R
s
is much lower than the reactance of
leakage inductance L
l
. It is assumed that a high-quality sense
resistor is used, so that its inductance L
R
is negligibly small.
Thus, the equivalent circuit of a current transformer resembles
a second-order, low-pass lter.
From Fig. 7, the current through the leakage inductance is
given by
I
Ll
(s) =
I(s)R
c
n
_
R
c
+ sL
l
+
_
1
sC
|R
__
=
I(s)R
c
C
_
s +
1
RC
_
n
_
CL
l
_
s +
1
RC
_
_
s +
R
c
L
l
_
+ 1
_. (33)
Therefore, the voltage across sense resistor R is given by
V
R
(s)=
_
R
_
_
_
_
1
sC
_
I
Ll
(s)=
I(s)R
c
n
_
CL
l
_
s+
1
RC
_
_
s+
R
c
L
l
_
+1
_.
(34)
KONDRATH AND KAZIMIERCZUK: BANDWIDTH OF CURRENT TRANSFORMERS 2013
Fig. 8. Magnitude Bode plot of a current transformer for high frequencies.
Fig. 9. Phase Bode plot of a current transformer for high frequencies.
Hence, the transresistance of the current transformer for high
frequencies is obtained as
R
m
(s) =
V
R
(s)
I(s)
=
R
c
nCL
l
1
s
2
+
_
R
c
L
l
+
1
RC
_
s+
1
CL
l
_
R
c
R
+1
_
=
R
c
nCL
l
1
s
2
+ 2
0
s +
2
0
(35)
where
f
0
=

0
2
=
1
2

1
CL
l
_
R
c
R
+ 1
_
=
R
c
L
l
+
1
RC
2
_
1
CL
l
_
R
c
R
+ 1
_
. (36)
Figs. 8 and 9 show the Bode plots for the designed current
transformer for high frequencies with R
c
= 7.24 k, R =
10.5 , measured values of L
l
= 0.126 H, and C = 0.81 nF.
For < 1, a second-order system has two imaginary conju-
gate poles and the upper 3-dB frequency f
H
kf
0
, where k
depends on the value of . For 1, the two poles p
1
and p
2
are real, and the denominator of the transresistance R
m
in the
high-frequency range can be represented as
s
2
+
_
R
c
L
l
+
1
RC
_
s +
1
CL
l
_
R
c
R
+ 1
_
= (s +
p1
)(s +
p2
) = s
2
+ (
p1
+
p2
)s +
p1

p2
. (37)
Hence

p1
+
p2
=
R
c
L
l
+
1
RC

p1

p2
=
R
c
RCL
l
+
1
CL
l

R
c
RCL
l
(38)
if
1
CL
l

R
c
RCL
l
, which simplies to
R
c
R
1. (39)
This inequality is well satised in practice. Thus, the frequen-
cies of the real poles can be approximated by
f
p1
=

p1
2
=
1
2RC
f
p2
=

p2
2
=
R
c
2L
l
. (40)
If f
p2
4f
p1
, then pole p
1
is dominant, and f
H
f
p1
. On
the other hand, if f
p1
4f
p2
, then pole p
2
is dominant, and
f
H
f
p2
. When f
p1
and f
p2
are comparable, the upper 3-dB
frequency of the current transformer is given by
f
H

1
_
1
f
2
p1
+
1
f
2
p2
=
1
2
_
(RC)
2
+
_
L
l
R
c
_
2
. (41)
For example, for R = 10.5 , R
c
= 7.24 k, L
l
=
0.126 H, and C = 0.81 nF
f
p1
=
1
2RC
=
1
210.50.8110
9
=18.71 MHz (42)
f
p2
=
R
c
2L
l
=
7.24 10
3
20.12610
6
=9.145 GHz (43)
f
H

1
_
1
(18.7110
6
)
2
+
1
(9.14510
9
)
2
=18.71 MHz. (44)
Fig. 10 illustrates the step response of the current transformer
for high frequencies. Rise time t
r
= 2.62 ns can be observed
from the plot.
Although R
s
is considered to be a constant, in practice,
the value of secondary winding resistance increases with the
increase in frequency. The skin depth in the secondary winding
at f
H
is given by

w
=
_

Cu

0
f
H
=
_
1.72410
8
1410
7
18.7110
6
=15.28 m. (45)
2014 IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 58, NO. 6, JUNE 2009
Fig. 10. Step response of a current transformer for high frequencies.
The winding diameter with isolation is given by
p =
d
0.8
=
0.644
0.8
= 0.805 mm. (46)
Hence
np = 40 0.805 = 3.22 cm < l = 8.49 cm. (47)
This implies that a single-layer winding N
l
= 1 of 40 turns can
be made on a core of mean length 8.49 cm. Since d/
w
1,
the secondary winding resistance at f
H
is given by
R
s
= 0.8343
d

d
p
_
2N
2
l
+1
_
3
(48)
= 0.8343
0.64410
3
15.2810
6
_
0.64410
3
0.80510
3
(21+1)
3
= 31.45 . (49)
E. Bandwidth
Using (18) and (48), the bandwidth of a current transformer
can be given as
BW = f
H
f
L
=
1
2
_
_
L
l
R
c
_
2
+ (RC)
2

R
Lm
2L
m
. (50)
For the given example, the bandwidth is given by
BW=f
H
f
L
=18.7110
6
4.82910
3
18.705 MHz.
(51)
V. EXPERIMENTAL RESULTS
The designed current transformer had the following spec-
ications: n = 40, L
m
= 0.349 mH, R
c
= 7.24 k, L
l
=
0.126 H, R
s
= 0.1073 , C = 0.81 nF, and R = 10.5 .
Fig. 11. Magnitude Bode plot of the current transformer using experimental
results and MATLAB simulations.
Fig. 12. Phase Bode plot of the current transformer using experimental results
and MATLAB simulations.
An experiment using the designed current transformer was
done to sense the current in a simple circuit consisting of a
sinusoidal voltage source connected to a series combination
of two resistors, i.e., 10.4 and 10.6 . The amplitude of the
sinusoidal voltage source was set to 5 V, and the frequency was
varied from 1 kHz to 20 MHz. The frequencies below 1 kHz
resulted in a distorted output from the current transformer. The
signal generator used had a maximum frequency of 20 MHz.
The frequency responses of the circuit obtained using
MATLAB simulations, as well as the experiments, are illus-
trated in Figs. 11 and 12. As seen fromthe MATLABsimulation
results in Figs. 11 and 12, the Bode plots corresponding to
KONDRATH AND KAZIMIERCZUK: BANDWIDTH OF CURRENT TRANSFORMERS 2015
TABLE I
COMPARISON OF CUTOFF FREQUENCIES OBTAINED FROM APPROXIMATE
EQUATIONS, MATLAB SIMULATION, AND EXPERIMENTAL RESULTS
TABLE II
EFFECTS OF PARASITIC COMPONENTS ON BANDWIDTH
the approximate equations in (15) and (35) follow the Bode
plots corresponding to the wideband transresistance given in
(11). Therefore, considering the simplicity of the expressions,
values corresponding to the approximate equations are used in
any further discussions. From Figs. 11 and 12, it is also seen
that the Bode plots obtained from the experiment matched very
well with the predicted Bode plots. The approximate values of
f
L
and f
H
are calculated as given in (22) and (45). The values
of f
L
and f
H
are also obtained from MATLAB simulation
results, as well as from the experiment results. These values
that are obtained from the approximate equations, MATLAB
simulations, and the experiment are given in Table I.
VI. EFFECTS OF PARASITIC COMPONENTS ON THE
BANDWIDTH OF CURRENT TRANSFORMERS
The effects of variations in the parasitic component values
on the cutoff frequencies and, thereby, on the bandwidth are
summarized in Table II. Thus, to have a low lower cutoff
frequency, the core of the current transformer is desired to have
high permeability and, thereby, high magnetizing inductance
L
m
. For a high value of the higher cutoff frequency, low values
of leakage inductance and stray capacitance are required. The
leakage inductance and the stray capacitance depend on the
number of turns in the transformer, as well as the construction
of the transformer, and are inversely related to each other.
Selection of a sense resistor with a high value of parasitic
inductance results in peaking at high frequencies.
VII. CONCLUSION
The transfer function of the current transformer has been
derived for a wide frequency range. Approximate expressions
for lower cutoff and higher cutoff frequencies of a current
transformer have also been derived. The approximated equa-
tions for cutoff frequencies have been validated by MATLAB
simulations and experimentation. The conclusions on the ef-
fects of various parasitic components on the current transformer
frequency response are as follows.
1) As magnetizing inductance L
m
is increased, lower cutoff
frequency f
L
decreases.
2) As leakage inductance L
l
is decreased, upper cutoff fre-
quency f
H
increases.
3) As stray capacitance C is decreased, upper cutoff fre-
quency f
H
increases.
For a broadband current transformer, magnetizing inductance
L
m
should be very high, and leakage inductance L
l
and stray
capacitance C should be very low. As the leakage inductance
and the stray capacitance are inversely related to each other,
the major challenge encountered in the design of a broadband
current transformer is achieving low values of L
l
and C.
The expressions for the cutoff frequencies and the bandwidth
derived in this paper can be used for the construction of
wideband current transformers. Also, the circuit model of the
current transformer can be used for various simulation purposes
using Spice, e.g., for current measurements or observations in
PWM converters. The transfer function model of the current
transformer over a wide frequency range for simulations with
MATLAB can be used in applications such as current-mode
controlled PWM converters. Some of the results of this paper
can be adopted for regular transformers since the analysis has
been done using the model of the current transformer that is
similar to the transformer model.
APPENDIX
The transresistance of the current transformer, including the
capacitance of the primary winding, is given by
R
m
=
v
R
i
=
1
nC
p
C
s
Den
vR
(A1)
where Den
vR
is given in (A2), shown at the bottom of the page.
For the input capacitance C
/
p
in the range from 0 to 0.03 pF,
the Bode plots obtained from (11) and (A1) are nearly identical.
Den
vR
=L
l
_
s
4
+ s
3
_
L
l
(RC + C
p
R
c
) + RCC
p
R
c
R
s
RCC
p
R
c
L
l
_
+ s
2
_
RCR
c
L
m
+ RC
p
R
c
L
m
+ L
l
(RCR
c
+ L
m
) + R
s
(RCL
m
+ C
p
R
c
L
m
)
RCC
p
R
c
L
m
L
l
_
+s
_
L
m
(R
c
+ R + R
s
) + L
l
R
c
+ RCR
c
R
s
RCC
p
R
c
L
m
L
l
_
+
R + R
s
RCC
p
L
m
L
l
_
(A2)
2016 IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 58, NO. 6, JUNE 2009
REFERENCES
[1] A. Wright, Current Transformers, Their Transient and Steady State
Performance. London, U.K.: Chapman & Hall, 1968.
[2] R. Paul, Electromagnetics for Engineers With Applications. New York:
Wiley, pp. 186188.
[3] C. W. T. McLyman, Transformer and Inductor Design Handbook, 3rd ed.
New York: Marcel Dekker, 2004.
[4] N. Locci and C. Muscas, Hysteresis and eddy current compensation
in current transformers, IEEE Trans. Power Del., vol. 16, no. 2, pp. 154
159, Apr. 2001.
[5] G. Cerri, R. De Leo, V. M. Primiani, S. Pennesi, and P. Russo, Wide band
characterization of current probes, IEEE Trans. Electromagn. Compat.,
vol. 45, no. 4, pp. 616625, Nov. 2003.
[6] B. V. Cordingley and D. J. Chamund, Some observations on the per-
formance of modern wideband current transformers in pulse current
measurement applications, in Proc. 5th Modular Klyston Workshop
Future Linear Colliders, Geneva, Switzerland, Apr. 2627, 2001.
[7] J. G. Kassakian, M. F. Schlecht, and G. C. Verghese, Principles of Power
Electronics. Reading, MA: Addison-Wesley, pp. 597599.
[8] A. Ahmad, P. Auriol, and C. Kieny, Shell-form power transformer
modelling at high frequencies, IEEE Trans. Magn., vol. 30, no. 5,
pp. 37293732, Sep. 1994.
[9] P. N. Miljanie, Mutual inductance highly independent of primary
winding position and ambient elds, IEEE Trans. Instrum. Meas.,
vol. 46, no. 2, pp. 471473, Apr. 1997.
[10] J. Faiz, B. Abed-Ashtiani, and M. R. Byat, Lumped complete equivalent
circuit of a coreless high-frequency transformer, IEEE Trans. Magn.,
vol. 33, no. 1, pp. 703707, Jan. 1997.
[11] E. Labour, F. Costa, and F. Forest, Current measurement in static
converters and realization of a high frequency passive current probe
(50 A-300 MHz), in Proc. EPE, 1993, vol. 4, pp. 478483.
[12] G. Grandi, M. K. Kazimierczuk, A. Massarini, and U. Reggiani, Stray
capacitances of single-layer solenoid air-core inductors, IEEE Trans. Ind.
Appl., vol. 35, no. 5, pp. 11621168, Sep./Oct. 1999.
[13] F. Costa, P. Poulichet, F. Mazaleyrat, and E. Labour, The current sensors
in power electronics, a review, EPE J., vol. 11, no. 1, pp. 718, Oct. 2001.
[14] N. Das and M. K. Kazimierczuk, An overview of technical challenges
in the design of current transformers, in Proc. Elect. Insul. Conf. Elect.
Manuf. Expo, 2005, pp. 369377.
Nisha Kondrath (S06) received the B.Tech. degree
from Mahatma Gandhi University, Kerala, India, in
2002 and the M.S. degree from Wright State Uni-
versity (WSU), Dayton, OH, where she is currently
working toward the Ph.D. degree, both in electrical
engineering.
She is currently a Research Assistant with the
Department of Electrical Engineering, WSU. Her
research interests include power electronics, high-
frequency PWM dcdc converters, power semicon-
ductors, and magnetics.
Ms. Kondrath is a member of Tau Beta Pi and the Electrical Manufacturing
and Coil Winding Association.
Marian K. Kazimierczuk (M91SM91F04)
received the M.S., Ph.D., and D.Sci. degrees in elec-
tronics engineering from the Technical University of
Warsaw, Warsaw, Poland, in 1971, 1978, and 1984,
respectively.
From 1978 to 1984, he was an Assistant Profes-
sor with the Department of Electronics, Institute of
Radio Electronics, Technical University of Warsaw.
In 1984, he was a Project Engineer with Design Au-
tomation, Inc., Lexington, MA. During 19841985,
he was a Visiting Professor with the Department
of Electrical Engineering, Virginia Polytechnic Institute and State Univer-
sity, Blacksburg. Since 1985, he has been with the Department of Electrical
Engineering, Wright State University, Dayton, OH, where he is currently a
Professor. He is the coauthor of the books Resonant Power Converters (Wiley,
1995), Electronic Devices, A Design Approach (PrenticeHall, 2004), and Lab
Manual (PrenticeHall, 2004). He has published over 250 technical papers,
95 of which have appeared in IEEE TRANSACTIONS AND JOURNALS. He
served as an Associate Editor for the Journal of Circuits, Systems, and Com-
puters. He is the holder of eight patents. His research interests include high-
frequency, high-efciency RF power ampliers, resonant and PWM dcdc
power converters, dcac inverters, high-frequency rectiers, modeling and
control of converters, high-frequency magnetics, power semiconductor devices,
electronic ballasts, RF transmitters, sensors, and superconductivity.
Prof. Kazimierczuk is a member of Tau Beta Pi. He was an Associate
Editor of the IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMSI:
FUNDAMENTAL THEORY AND APPLICATIONS, during 19931995,
19972001, and 20022003, and the IEEE TRANSACTIONS ON CIRCUITS
AND SYSTEMSI: REGULAR PAPERS in 2004 and 2005. He serves
as an Associate Editor for the IEEE TRANSACTIONS OF INDUSTRIAL
ELECTRONICS. He was the Chair of the Circuits and Systems Technical
Committee of Power Systems and Power Electronics Circuits during
20012002 and 20042005. He was the recipient of the 1991 IEEE Harrell
V. Noble Award for his contributions to the elds of aerospace, industrial,
and power electronics; the 1991 Presidential Award for Faculty Excellence in
Research; the 1993 and 2000 College of Engineering and Computer Science
Teaching Awards; the 1995 Presidential Award for Outstanding Faculty
Member; the 19961999 Brage Golding Distinguished Professor of Research
Award from Wright State University; the 1997 and 2002 College Professional
Service Awards; and the 2004 Wright State University Board of Trustees
Award.