EMC Model of Low Voltage DC Motor

I. Oganezova1,2, R. Kado3, B. Khvitia1, Z. Kuchadze1, A. Gheonjian1,2, R. Jobava1,2

1
EMCoS Ltd., 0160 Tbilisi, Georgia
2
Electrical and Electronics Engineering Department, Tbilisi State University, Tbilisi, Georgia
3
Electromagnetic Compatibility, Chrysler, 800 E. Chrysler Drive, CIMS 481-47-20, Auburn Hills, MI 48326

Abstract—This paper proposes a high-frequency dynamic performed. A behavioral circuit model allows various DC
circuit network model of a DC motor for predicting conductive motors to be analyzed at the beginning of the design stage in
emissions in low-voltage automotive applications, and discusses a the time and frequency domains. The model will facilitate
study in which this model was examined. The proposed model, recommendations about motor component design in order to
which is based on a behavioral approach, introduces some avoid the critical EMI limits imposed by the automotive
physical features and phenomena of motor armature windings. manufacturer specifications.
Together with impedance characterization current model gives
possibility to perform accurate transient analysis and describe After presenting the proposed DC motor model and test
complex commutation processes in motor during rotation. benches for motor validation, we describe a real automotive
application as a case example and demonstrate the
Keywords—EMC; EMI; DC motor; commutaion model; stray effectiveness of the proposed approach for predicting the
elements; circuit network model; conducted emissions. conducted EMI generated by the motor according to CISPR 25
[14] and ISO 7637 [15] standards.
I. INTRODUCTION
In the automotive industry, low-voltage DC motors are II. EQUIVALENT MODEL AND VALIDATION
used widely for various applications. However, motors are a Two-speed DC motors are used widely in the automotive
common source of electromagnetic interference (EMI), and to industry. For example, they are used in front window wiper
suppress electromagnetic noise, DC motors are frequently systems in order to obtain two-speed operation. Typically, one
equipped with filters. Motor rotation involves arcing at the high-speed motor terminal is set at some angle from the
motor brushes and commutator bars connected to armature standard low-speed positive terminal position so that when
windings. The discontinuous commutation process produces voltage is applied to this brush, the motor rotates at a higher
electromagnetic broadband noise with frequencies ranging speed. However, adding the high-speed brush in an asymmetric
from tens of kHz to tens of MHz. These disturbances can position causes the wiper motor to produce more noise.
conduct to other components or radiate and generate EMI.
Fig. 1 depicts the two-speed 12 V DC motor.
To optimize DC motors, it is of great value to study the
causes of noise generation processes in rotating direct current
machines and to attempt to reduce the noise by making changes
to the design of the motor. In contrast to induction models, DC
motors cannot be described exclusively by their complex
frequency-dependent impedance; EMC studies that
characterize the switching between the brushes and the
commutating bars also are required. [1]-[6] discussed
interesting ways to generate behavior models of the frequency-
dependent impedance of DC motors. Various models of
armature coils and other motor components are presented; they (a)
achieved excellent agreement with measured Z(f) values.
Previous studies also have characterized the switching
processes that occur during the operation of DC motors and
have examined different commutation models [7]-[13].
However, to obtain a generic model of the motor that can be
used to predict electromagnetic disturbances, the impedance
model of the motor and the commutation model of arcing at the
motor brushes must be properly interconnected.
Using the proposed generic model of a rotating DC motor, (b)
the instantaneous current and voltage of each of the armature
coils of the motor can be computed. Both the impedance and Fig. 1. Illustration of the motor armature used for the model: (a) wiper
motor with chassis; (b) wiper motor with no chassis
commutation models are described by means of simple circuit
elements. To obtain the output circuit, a calculation must be

978-1-4799-5545-9/14/$31.00 ©2014 IEEE 81

 Inside the motor, the rotor armature is composed of 12 A suitable HF equivalent circuit can be represented by two
groups of coil windings, each wound on a laminated iron core circuits connected in series, as shown in Fig 4. A three-terminal
and connected to commutation bars. The rotor is supplied by coil model contains various stray elements of armature
three brushes, a positive low-speed brush, a positive high-speed windings, such as the inductance of the coil, turn-to-turn
brush and a negative brush. The stator is composed of two capacitance, iron loss, etc.
permanent magnets, which create a magnetic field. Fig. 2
depicts a schematic of the motor. RP1 RP2

C1 RC1 C2 RC2
L12
Armature coils
L L1 RL1 L2 RL2
A B

L1 L2 L3 L4 L5 L6 L7 L8 L9 L10 L11 Rg Rg

Ctg Ctg

Commutator
1 2 3 4 5 6 7 8 9 10 11 12
bars
N L H
Brushes

Motor terminals Fig. 4. Electrical schematic of a two-speed DC motor

Fig. 2. Electrical schematic of a two-speed DC motor
The proposed HF impedance model, implemented in the
The equivalent electrical model consists of a network EMC Studio software package, consisted of 12 three-terminal
representation of the motor that changes with time as the coil blocks [16]. The model is available for circuit simulations.
armature rotates. During operation, brushes slide on the surface
The parameters of the investigated DC motor (Fig. 1) were
of the rotating collector and transfer from one commutation bar
adjusted to simulate the frequency behavior of armature coils
to another. During commutation, switching occurs between
for the L brush terminal-to-ground and H brush terminal-to-
different sections of the armature winding and the current
ground cases. Table I lists the parameters and their descriptions
reverses when the armature winding is shortened by the brush.
for the HF equivalent circuit.
Due to the magnetic field, back electromagnetic field is
induced in the shortened coil. Generated energy stored in the TABLE I. THREE-TERMINAL MODEL PARAMETERS
coil causes additional current on the brush surface, thus
generating a voltage strike between commutation ends. Parameter Description Values
L1, L2 Inductance of coil 28 uH, 1.2 uH
The arcing that occurs during the commutation process and
the complex windings of the motor armature can be described RL1, RL2 DC resistance of coil wire 0.2 Ohm, 0.2 Ohm
accurately by means of circuit elements. The methodology for C1, C2
Turn-to-turn coil
400 pF, 250 pF
generating equivalent circuits can be separated into two parts, capacitance
which are described in the following subsections. Turn-to-turn capacitive
RC1, RC2 1 mOhm, 1 mOhm
current losses
A. Impedance model of DC motor RP1, RP2 AC iron losses 480 Ohm, 400 Ohm
To determinate a suitable high frequency (HF) equivalent Ctg
Coil-to-ground
15 pF
model of winding, it is necessary to perform measurements of capacitance
the coil impedance at different frequencies and based on the Rg Coil-to-ground losses 40 Ohm
data measured construct equivalent circuit.
The low positive brush terminal-to-ground and high Fig. 5 illustrates a comparison of the simulated and
positive brush terminal-to-ground impedance values are measured impedance values of the H brush terminal.
measured using a network analyzer in the frequency range
from 300 kHz to 108 MHz. For the purpose of these
1200
measurements, the motor terminals are connected to an SMA Measured Rs
connector, as shown in Fig. 3. 900 Measured Xs
Impedance [Ohm]

Simulation model
600

N L 1MOhm 300
DC Motor
0
H -300

SMA Network Analyser -600 0 1 2

10 10 10
HP 8752A Frequency [MHz]
Fig. 5. Motor H terminal to ground
Fig. 3. Motor impedance measurements
The results indicate good agreement between the simulated
and measured impedance values. Similar accuracy was

82
achieved during the analysis of L brush terminal. Therefore, the L Brush to segment contact [1/Ohm]
validated HF equivalent model of armature windings can be
used to effectively analyze the impedance of a two-speed wiper Bar 1
motor in the RF frequency range. Bar 2
Bar 3
B. Commutation model Bar 4
To analyze voltages and currents generated during the Bar 5
operation of a DC motor, another equivalent circuit model is
proposed. In an equivalent circuit, each brush exhibits time-
dependent resistance to each armature coil, as shown in Fig. 6. Bar 12
The resistance value depends on the position of the brush
0 2 4 6 8 10
inside the motor. Time [msec]

Fig. 7. Brush-to-commutator bar conductance

BH1 BH2 BH3 BH4 BH5 BH6

BL1 BL2 BL3 BL4 BL5 BL6

C. Measurement setup and validation
BN1 BN2 BN3 BN4 BN5 BN6
The proposed EMC model of the DC motor was
implemented as a new element in the EMC Studio software
L1 L2 L3 L4 L5 L6
package [16] and used for transient circuit analysis. To validate
the EMC model the simulated results of the voltages generated
L12 L11 L10 L9 L8 L7 during motor operation were compared with data measured
using a Lecroy oscilloscope. The measurement setup consisted
of a two-speed wiper motor fed by a 12V battery.
BN12 BN11 BN10 BN9 BN8 BN7
Measurements were taken during high and low operational
BL12 BL11 BL10 BL9 BL8 BL7 stages of the motor. Two 1-m cables were used for the
BH12 BH11 BH10 BH9 BH8 BH7
connection to the battery. The motor was isolated from the
metallic table, and 30 cm of wire was used for remote
Fig. 6. Equivalent circuit model of commutation process grounding. Fig. 8 shows a picture of the complete measurement
setup.
The value of the brush-to-commutator bar resistance is
represented by a periodical pulse function with the parameters
listed in Table 2. The period of the pulse function depends on
the rotation speed of the motor. The duration of the
commutation depends on brush width. The motor brush
considered in this study covered less than two collector
segments.
Due to the material properties of both the brush and the
commutator bars, we used finite values of the turn-on and turn-
off resistances for the simulations. The exact characteristics of Fig. 8. Measurement setup
the motor components were not available, so we adjusted the
resistance values to fit the measured results during the To measure the voltages at the output terminals of the DC
validation of the model. The resistance between the brush and motor, additional short wires were used to connect the scope
the commutator bar varied from Rc+Ron to Rc+Roff. probes directly to the motor’s output pins.
Fig. 9 illustrates the equivalent circuit model of the setup,
TABLE II. COMMUTATION MODEL PARAMETERS which consisted of three main blocks, a wiper motor equipped
Low- with an LF filter, a wiper grounding bolt, and a 12V battery. In
High-speed
Parameter Description speed the simulation model, the system’s supply and ground cables
model
model were modeled using lumped circuit elements, as described in
S Motor nominal speed 30 50 [17].
K Reduction coefficient 80 80
The DC motor was grounded remotely, so we applied an
b Brush width 48 degree 48 degree additional chassis-to-ground capacitance of the motor in the
simulations. The value of Cgnd was 9pF. The simulation model
Ron Brush on resistance 0.1 mOhm 0.1 mOhm
of the battery was represented by a DC voltage source having
Roff Brush off resistance 10 kOhm 10 kOhm an internal resistance and capacitance between the battery’s
pins and ground plane. The capacitance between battery’s pins
Rc Brush contact resistance 10 mOhm 10 mOhm
and ground was 122 pF.
Fig. 7 depicts the conductance of the brush-to-commutator
segment depending on the time and rotation angle.

83
 Wiper motor with filter 12V battery III. AUTOMATIVE APPLICATION: ESTIMATION OF
N L 5.6uH
O1
CONDUCTIVE EMI
DC Motor 1uF 2000pF
2mOhm

GND point GND

H
5.6uH
1uF 2000pF
Plus line
A. Conducted emmision test according to CISPR 25 standard
O3 O2
12V
The applicability of the EMC model of the two-speed wiper
Ground wire Minus line motor for predicting conductive emission testing was checked.
Cgnd C_battery The test bench described in the CISPR 25 Standards [14] was
used to measure the conducted EMI from the DC motor in the
Fig. 9. Circuit representation of measurement setup (low-speed mode) RF band.
Fig. 12 depicts the setup, which consisted of a DC motor,
Fig. 10 depicts the measured and simulated results of the 12V battery, two artificial networks (LISN) and supply cables.
voltages at the motor’s output terminals in low-speed mode.
The measured results were obtained using FFT transform of the Wiper motor with filter 12V battery

transient signals recorded by an oscilloscope. For simulation, N

DC Motor
L 5.6uH
DUT+ Supply+ O1
1uF 2000pF
transient analysis of the commutation processes also was H 1uF 2000pF
DUT- LISN Supply-
Plus line 2mOhm

performed, and frequency responses up to 108 MHz were GND 5.6uH Probe

12V

investigated. DUT+

DUT- LISN
Supply+

Supply-
O2

Minus line

The simulation model can be used to describe the frequency Cgnd

Probe
C_battery

behavior and level of noise produced by the DC motor. Despite Reciever

operating in low-speed mode, due to the complex interaction of

the armature coils inside the motor, noise can be measured at Fig. 12. Simulation model of CISPR 25 test setup (low-speed mode)
both positive terminals of a two-speed motor.
The conductive EMI were obtained for both the high-speed
Low speed brush; Low speed mode and low-speed operating modes of the DC motor. Simulated
values at the LISN’s 50 Ohm ports were compared with
Measurements
Simulations
standard defined limits for the frequency range under
100
consideration. Simulated voltages produced by the DC motor
Voltage [dBV]

equipped with an LC filter, consisted of non-ideal elements,

50 were below the required levels for both the low-speed and
high-speed modes.

0 80

60
500Hz 10kHz 1MHz 100MHz
Voltage [dBV]

Frequency 40
Fig. 10. Generated noise at motor’s low-speed brush
20
High speed brush; Low speed mode
0
Measurements
-20 Wiper motor at low speed
100 Simulations
Wiper motor at high speed
Voltage [dBV]

-40
150kHz 1MHz 10MHz 100MHz
Frequency [MHz]
50
Fig. 13. Predicted conductive emission compared against standard limits

0 B. Transient conducted emmision test according ISO 7637-2

standard
500Hz 10kHz 1MHz 100MHz The EMC model of the wiper motor was checked to
Frequency
evaluate its predicted transient conductive emission along
Fig. 11. Generated noise at motor’s high-speed brush battery-fed supply lines. The ISO 7637-2 standard specifies test
The results indicate good agreement between the simulated benches for testing the compatibility of equipment with
results and the measured data in the RF frequency range, thus conducted electrical transients. For fast pulses (nanosecond to
suggesting that the proposed model can be used to predict microsecond range), the disturbance source (DC motor) was
disturbances and analyze motor performance in industrial connected to a switch, an artificial network (LISN), and a 12V
automotive applications at different design stages. battery, as shown in Fig. 14.
The disturbance voltage caused by a supply line
disconnection was measured at the motor terminals using a
voltage probe and oscilloscope. The measurements began at the
moment the switch was opened.

84
 characterization,” IEEE Transactions on Electromagnetic
Wiper motor with filter PULSE(0 1 2.5ms 300ns 300ns 2ms 5ms)
V1
12V battery
Compatibility, vol. 52, no. 4, pp. 888–901, 2010.
N
DC Motor
L
5.6uH
DUT+ Supply+ O1 [2] R. Kahoul, P. Marcha, Y. Azzouz, and B. Mazari, "HF model of DC
1uF 2000pF
DUT- LISN Supply- 2mOhm motor impedance EMC problems in automotive applications," EMC

S1

SW
Plus line
H
GND 5.6uH
1uF 2000pF
Probe 2008 IEEE International Symposium on, Detroit, USA, pp.782-786,
R1
50
O2
12V August 2008.
Minus line [3] J. Benecke, A. Lined, and S. Dickmann, “Automatic HF model
Cgnd C_battery generation and impedance optimization for low voltage DC motors,”
Oscilloscope
Proceedings of the 2008 International Conference on Electrical
Machines, Vilamoura, 6-9 September 2008, pp. 1-6.
Fig. 14. Simulation model of ISO 7637-2 test setup (low-speed mode)
[4] J. Benecke, and S. Dickmann,"Analytical HF model of al voltage DC
Conductive transients were obtained for both the high- motor armature including parasitic properties," Proc. IEEE Int. Symp.
speed and low-speed operating modes of the DC motor. Fig. 15 Electromagn. Compat., pp.1-4 2007.
depicts the simulated values at the motor terminals. The [5] J. Benecke, and S. Dickmann, "Analytical high frequency model for
maximum negative and positive burst levels were -49.6 V and multipole DC motors, in M urich Munich 00 , 1 th
nternational urich Symposium on lectromagnetic ompati ility,
42.9 V, respectively, which are below the requirements M nchen, Sept. 2007, pp. 201-204.
(generally +/-80V). [6] G. Grandi , D. Casadei, and U. Reggiani, "Equivalent circuit of mush
wound AC windings for high frequency analysis," Proceedings of the
Switch ON to OFF condition ISIE\'97 Symposium, Guimaraes, Portugal, pp. SS201 -SS206 1997.
40
[7] T. D. Batzel, N. C. Becker, and M. Comanescu, "Analysis of brushed
DC machinery faults with coupled finite element method and
20 equivalent circuit model," Proceedings of the 2011 IAJC-ASEE
International Conference, University of Hartford, 200 Bloomfield
Voltage [V]

Avenue, West Hartford, CT 06117, United States, 2011.

0
[8] J. Liu, X. Wang, and J. Nan, "Application of ant colony algorithm to
the analysis of common mode EMI model of DC motor," Energy and
-20 Power Engineering, vol. 3 no. 2, 2011, pp. 96-106.
Low speed mode [9] H. R. Choi, “Arc reduction of high-voltage DC-motor for automotive
High speed mode application by targetwise parameter optimization (TPO),” IEEE
-40 Transactions on Magnetics, vol. 45, issue 3, pp. 1840-1842, 03/2009.
0 100 200 300 400
Time [s] [10] G. C. R. Sincero, J. Cros, and P. Viarouge, "Arc models for simulation
Fig. 15. Predicted transient conductive emission (switch from on to off of brush motor commutations," IEEE Trans. on Mag., vol. 44, no. 6,
condition) June 2008, pp. 1518-1521.
[11] C. R. Suriano, J. R. Suriano, G. Thiele, and T. W. Holmes, "Prediction
of radiated emissions from DC motors," IEEE Trans. Electromagn.
IV. CONCLUSIONS Compat., p. 790, 1998.
The proposed EMC model of low-voltage DC motors is [12] J. Suriano, and C. M. Ong, "Modeling of electromechanical and
based on behavioral time-varying circuit topology. It takes into electromagnetic disturbances in DC motors," Conf. Rec. IEEE National
Symposium on Electromagnetic Compatibility, 1989, pp. 258 - 262.
account all main components of the motor (brushes, coils,
commutator segments, iron core and chassis). The impedance [13] D. P. Motter, "Commutation of D-C machines and Its effects on radio
influence voltage generation," AIEE Transactions, vol. 68, 1949, pp.
model, which contains the main stray elements present in the 491-496.
DC motor, applies to the applicable RF frequency range. The [14] CISPR25, Radio disturbance characteristics for the protection of
simulated EMC model is represented as an active device that receivers used on board vehicles, boats and on devices - limits and
generates EMI due to the commutation process. The described methods of measurement. International Electrotechnical Commission
model can be used to analyze the transient and frequency (CISPR), August 2002.
responses of a two-speed wiper motor. Predicting the noise [15] ISO 7637-2, Road vehicles – electrical disturbances from conduction
produced by the DC motor allows for an analysis of a and coupling – Part 2: Electrical transient conduction along supply
lines only, Third Edition, November, 2011.
prototype’s design in terms of its conductive disturbance at
[16] M Studio User’s Manual. v. 4.1, M oS Tech. Rep. S010 0 ,
various design stages of the motor, along with embedded Tbilisi, 2008.
filters.
[17] C. R. Paul, Analysis of Multiconductor Transmission Lines, Wiley,
New York, 1994.
REFERENCES
[1] R. Kahoul, Y. Azzouz, P. Marchal, and B. Mazari, “New behavioral
modeling for DC motor armatures applied to automotive EMC

85