MIRAFZAL et al.: INDUCTION MOTOR MODEL WITH LOW-TO-HIGH FREQUENCY-RESPONSE CHARACTERISTICS 1235
versus frequency of both 5- and 20-hp motors for a 460-V stator
winding connection only.
Fig. 1 demonstrates that the motor stray capacitors are mainly
a function of motor geometry. Low-frequency motor impedance
values below the first resonance are mainly L–R equivalent
circuits. In this region, the impedance of the series 460-V
connection has four times the value of the parallel 230-V
connection since a winding inductance varies as N
2
(see Fig. 1
at 100 Hz). Fig. 1 also shows that the first resonance frequency
of the 5-hp series connection is almost half the resonance
frequency (f
r
) of the parallel motor connections.
The f
r
of the motor impedance is defined as where the
impedance has its maximum magnitude and is calculated by
f
r
=1/(2π
√
LC), where L and C are the effective inductance
and capacitance values, respectively. The effective inductance
L for a series connection is four times the value in a parallel
connection, and f
r
is half that of the parallel connection.
Accordingly, the conclusion is that the motor stray capacitance
does not significantly vary with changes in stator winding
configuration. That is, in this case, the effective capacitance
value C is mainly a function of the motor geometry or the
motor frame size. It will be shown in Section IV that Y- and
∆-connections have impacts on the equivalent capacitance
value at high frequencies.
The second objective of this section is to introduce the low-
impedance antiresonance phenomenon observed in the case of
the 20-hp motor at 4.1 MHz in Fig. 2. The motor impedance
at this resonance frequency is 10 Ω. This DM low impedance
at 4.1 MHz will be approximately equivalent to 2.2 Ω in the
CM circuit. This provides a very low impedance path for the
CM leakage current at the low-impedance antiresonance. Thus,
modeling of this phenomenon is of significance for EMI analy-
sis in long cable motor drive systems. It is shown later that the
so-called antiresonance is actually caused by the inductance of
the first few turns of the winding. These observations are further
discussed in Section IV, where the proposed motor model is
derived based on the IEEE Standard 112 (T-equivalent) circuit.
B. Testing of DM Transfer Function
DM testing was performed with the phase-A motor lead
as one probe and the phase-B and phase-C motor leads tied
together to form the second probe. The measurements were
taken using HP4284 (20 Hz to 1.0 MHz), HP4285 (75 kHz to
30 MHz), and AP200 (20 Hz to 15 MHz) LCR meters in Z–θ
(impedance magnitude and phase angle) mode [33], [34]. The
recommended test procedure is to use an ungrounded motor
frame. It was observed that grounding the motor frame resulted
in significant measurement errors at high frequency. This is
attributed to errors in the instrument’s impedance calculation
procedure due to the LCR meter’s two-lead source voltage
driving capacitive current to ground that does not return on the
two leads [35].
C. Testing of CM Transfer Function
CM testing was performed with the motor frame as one probe
and the phase-A,-B, and -C motor leads tied together to form
Fig. 3. IEEE 112 recommended per-phase low-frequency equivalent circuit.
the second probe. Impedance measurements versus calculated
DM and CM transfer functions using the proposed induction
motor model are demonstrated later in Section IV-E.
III. M
ODELING INDUCTION MOTOR
FREQUENCY RESPONSE
In this section, existing lumped- and distributed-parameter
models for the analysis of the low- and high-frequency behavior
of induction motors are described.
A. Low-Frequency Response of Induction Motor
The low-frequency response of the polyphase induction mo-
tor has historically been studied since its origin. The IEEE
Standard 112 was established to obtain test methods and pa-
rameters for a low frequency (50, 60, and 400 Hz) per phase
T-equivalent induction motor model in Fig. 3 [36]. The rotor
circuit and neutral are actually physically isolated from the line
side stator neutral by the air gap in Fig. 3. Also, the rotor fre-
quency (slip frequency) is different from the stator (or line) fre-
quency. Users desire this low-frequency transfer function model
to obtain the line side electrical quantities (frequency, current,
power factor, and efficiency) for the given mechanical output
quantities (power, horsepower, and torque). For convenience of
analysis, rotor electrical quantities have thus been referred by
stator-to-rotor turns ratio to the stator with an implied electrical
connection to the stator neutral. In the low-frequency range,
the applied voltage is uniformly distributed through the stator
winding coils. Substantial literature exists on obtaining these
model parameters from mathematical calculations, heuristic
equations, and finite-element analysis (FEA). IEEE Standard
112 is a universally accepted test method to verify the ac-
curacy of these model parameters that are readily obtainable
by all users.
B. Mid-Frequency Response of Induction Motor
An accurate mid-frequency response (400 Hz to 20 kHz), as
related to PWM ac drives, would be correlated with the carrier
frequency motor ripple current and its associated heating in the
motor copper and core laminations.
C. High-Frequency Response of Induction Motor
The high-frequency response of the polyphase induc-
tion motor has recently been studied. High-frequency motor