THE DYNAMIC SIMULATION OF THE THREE-PHASE BRUSHLESS

PERMANENT MAGNET AC MOTOR DRIVES WITH LabVIEW

Li Ying * and Nesimi Ertugrul **
The University of Adelaide
Department of Electrical and Electronic Engineering,
Adelaide, Australia, 5005
* E`mail: lying@eleceng.adelaide.edu.au
** E`mail: nesimi@eleceng.adelaide.edu.au

Abstract: In this paper, a mathematical model of the three-phase brushless permanent magnet AC motor
drives in abc reference frame is described. A computer simulation of the motor drive is provided which
utilised LabVIEW software. The simulation can be conveniently used to study the dynamic as well as the
steady-state performance of the three-phase permanent magnet AC motor drives, with either trapezoidal
or sinusoidal back emfs, under various operating conditions. Some of the simulation results have been
given in this paper.

1. INTRODUCTION
There is a requirement to increase the efficiency of
AC industrial drives, small or large scale, due to the
increased awareness about the energy conservation
world-wide. The recent advancements about the
permanent magnet materials, the switching power
devices and the microelectronic technology have
greatly contributed to the new energy efficient and
high performance electrical drives, such as Brushless
Permanent Magnet AC (PMAC) motor drives. These
motors have higher efficiency and higher power
factor, and their output power per mass and volume
are much greater than their counterparts. Furthermore,
they have superior dynamic performance, which make
them suitable for high performance motor drives [1].
Depending upon the stator winding arrangement and
the shape and the location of the permanent magnets
on the rotor, the motors can be broadly classified into
two groups.
The first group possesses trapezoidal back emfs and is
called Brushless Trapezoidal Permanent Magnet
(BTPM) Motor, which is also known as Brushless DC
Motor. The second type possesses sinusoidal back
emfs and is called Brushless Sinusoidal Permanent
Magnet (BSPM) motor. Since BSPM employs a
sinusoidal variable frequency PWM inverter as a
power supply, the motor is also called Brushless
Permanent Magnet Synchronous Motor.
The principal differences between the two types of
motors are the accuracy of the rotor position sensor
and their respective control requirements. In order to
produce constant electromagnetic torque, they require
sinusoidal or rectangular excitation currents
respectively. Any deviation from these ideal current
excitations produces torque pulsations.
Computer simulation is a very useful technique to
analyse the behaviour of the motor drive without
implementation of the hardware. In addition to this,
the drive simulations can easily be forced to operation
under the extreme conditions without the fair of
damaging the motor drive as in the practical systems.
Although there has been many simulation studies in
the literature to predict the behaviour of the Brushless
PMAC motors, not many studies are user friendly and
produce accurate results that can imitate the real
drives.
In this paper, the LabVIEW (Laboratory Virtual
Instrument Engineering Workbench) software is used
as a graphical programming language. In comparison
with the other software tools, the simulation with
LabVIEW provides easy debagging features and user
friendly environment.
The main objective of this work is to create a general
simulation tool for both types of the PMAC motors,
which can be utilised to study the dynamic as well as
the steady-state performance under the various
excitation modes and the load conditions.
2. THE MOTOR DRIVE MODEL
In order to obtain a general dynamic model for the
BSTM and BTPM motors, the three-phase abc
modelling approach is used in this paper. Since the
rotor of a PMAC motor has high resistivity, the effects
of the stator currents on the total flux distribution may
be ignored under the normal operating conditions.
Therefore, the three-phase star-connected PMAC
motor can be modelled by a network consisting of a
winding resistance, an equivalent winding inductance,
and a back emf source per phase, all connected in
series.
In this work, it is assumed that the stator resistances of
all the windings are equal and the self and the mutual
inductances are constant. Therefore, the voltage
equations in the matrix form of a three-phase PMAC
motor are expressed as:
 v 1  R 0 0i1  L 0 0 i1  e1  Note that Eq. 5 should be repeated to define the other
v = 0 R 0i  + 0 L 0 d     two phases of the motor simply by shifting 120o
 2   2    dt i 2 + e2  (1)
electrical. Moreover, please remember that the
v 3 
   
0 0 Ri3 
 0 0 L
  
i3  
e3 
 practical back emf waveforms are always deviate from
these ideal assumptions. In these cases, the back emf
Here v1, v2, and v3 are the phase voltages; R is the waveforms may be modelled by using a look-up table
winding resistance; i1, i2, and i3 are the line currents; L or multiple harmonic components of the real
is the equivalent winding inductance; and e1, e2, and e3 waveform [2].
are the back emfs of the phases.
In Eqns. 4 and 5, E m is the maximum value of the
In the PMAC motors, because of the large airgap
between stator and rotor, the saturation is neglected. back emfs that can be given by
Therefore, the flux linkages become a linear function
of the phase currents, and hence the electromagnetic E m = k eω r (6)
torque is given by:
Where k e the is back emf constant, and θe is
Te =
1
(e1i1 + e2 i2 + e3i3 ) (2) electrical rotor position that is given by:
ωr
Here ω r is the angular speed of the rotor. θ e = pθ r = p ∫ω r dt (7)

However, in order to study the transient behaviour of Here θ r is the mechanical rotor position and p is the
the PMAC motors, the mechanical state equation must number of pole pairs of the motor.
be also known, that is given as
Generally, three-phase brushless PM motors are
dω r powered from the three-phase inverters. Therefore, to
Te − Tl = J (3)
dt obtain a complete drive simulation, the inverter should
Here Tl is the load torque; J is the inertia of the motor also be modelled.
and the connected load. The inverter that is controlled by the switching signal
As mentioned earlier the two groups of motors provides the desired terminal voltage to the each phase
(BTPM and BSPM) posses different back emf winding of the motor. In practical motor drives, the
waveforms. motor is normally connected star, and the star point is
normally left floating (which means that the star point
If the stator windings of the three-phase motor are is not linked to anywhere in the power circuit).
symmetrically displaced, the ideal back emf equations
of the BSPM motor can be given by Therefore, due to the inverter switching the star point
voltage varies, and the effective winding voltage
depends on the star-point voltage, which should also
e1  Em sin(θe ) 
    be determined in the simulation. Due to the symmetry
e =
 2  m E sin(θ e − 2π 3)  (4) in the motor windings, only one of the phase (Phase 1)
 
 3  m
e E sin(θ e − 4π 3) 
 voltage is analysed below, which can be repeated for
the other two phases.
For the BTPM motors, however, the back emf
waveform of one phase is given in piecewise linear If va is the terminal voltage of the motor and vs is the
form as: floating star-point voltage of the motor, both relative
to the mid point of the DC link voltage of the inverter,
 Em π  the phase voltage v1 can be given as
 θe 0 < θe ≤ 
π 6 6  v1 = v a − v s (8)
 π 5π 
E m < θe ≤  The terminal voltage va is determined by the switching
 6 6  states of the phase, which can be either ±Vdc/2. For the
 E 5π 7π  star-connected PMAC motor, it is always true that the
e = − m (θe − π) < θe ≤  (5)
 π6 6 6  summation of the line currents equals to zero.
 7π 11π  Therefore, when all three phases conduct current, the
− E m < θe ≤  floating star-point voltage of the motor can be easily
 6 6 
derived from the three voltage equations that is given
 Em 11π 
 (θe − 2π) < θe ≤ 2π  in Eq.1.
π 6 6 
v a + vb + vc e + e2 + e3
vs = ( )− ( 1 ) (9)
3 3
Similarly, if only two of the phases (say Phase 1 and calculation interval are the other two inputs in the sub-
2) are conducting currents, the floating star-point VI. In order to set the initial values at the top-level VI,
voltage of the motor can be derived as the final step values of the currents and the rotor
position also provided as the inputs to the "motor"
v a + vb e + e
vs = ( )− ( 1 2 ) (10) sub-VI. An input signal named "mode" is defined to
2 2 select a BTPM or a BSPM motor to be simulated. The
The Table 1 summarises the estimated star point and outputs of the "motor" sub-VI include the line
phase voltage values of the inverter driven motors currents, the phase voltages, the torque and the rotor
[2,3], which are also used in the simulation model of position.
the motor drive. Similar to the practical motor drive system, the
"control" sub-VI is implemented as the current
Table 1 The summary of the star point controller of the motor. From the control point of
and the phase voltages view, both the BSPM and BTPM motors can
accommodate the identical current controller. The
TRAPEZOIDAL BACK EMFs SINUSOIDAL BACK EMFs only difference is that they need either sinusoidal or
e1 + e2 + e3 ≠ 0 e1 + e2 + e3 = 0 rectangular current reference signals respectively.
(except zero crossing instants) Therefore, the first function of the "control" sub-VI is
va, vb, vc = ± Vdc / 2 to generate the three-phase current reference signals.
vs = K [(va + vb + vc) - (e1 + e2 + e3)] The reference current waveforms can be either as ideal
sine waves or as piecewise rectangular waveforms,
If i1 = 0 K = 1/2 then v1 = e1 , v2 = vb - vs , v3 = vc - vs which can be expressed per phase as shown below.
If i2 = 0 K = 1/2 then v1 = va - vs , v2 = e2 , v3 = vc - vs
If i3 = 0 K = 1/2 then v1 = va - vs , v2 = vb - vs , v3 = e3 i1 = I m sin(θ e ) (11)
If i1≠ 0 , i2 ≠ 0 , i3 ≠ 0 K = 1/3 then
v1 = va - vs , v2 = vb - vs , v3 = vc - vs
 π 
0 0 < θe ≤ 
6
3. THE VIRTUAL INSTRUMENT  
I π 5π 
OF THE DRIVE < θe ≤
 m 6 6 
 
The programs in LabVIEW applications are called  5π 7π 
Virtual Instruments (VI). Similar to a subroutine used i1 = 0 < θe ≤  (12)
 6 6 
in the C language, any VI in LabVIEW can be used as
 7π 11π 
a sub-VI in the block diagram (where the − I m 6
< θe ≤
6 
programming is done) of a high-level VI. The sub-VIs  
0 11π
can also be called from the inside of another sub-VI,
 < θ e ≤2π 
and there is no limit to the number of sub-VI used in  6 
LabVIEW. This hierarchical nature provides a very
flexible and powerful programming environment. Here Im is the amplitude of the stator current
command.
A number of sub-VIs are implemented in this study.
The VI named "motor" is based on the motor's As stated previously, because of the three-phase
functions that are summarised in the previous sections. symmetry, only one phase of the current reference
The "motor" sub-VI consists of four computation sub- signal is given above. The other two phases can be
section. The first section calculates the electrical rotor obtained by shifting Eqns. 11 and 12 1200 electrical
position according to the Eq.7. The second section angle.
produces the back emfs by using the Eq.4 or 5. The Although there are various current control schemes
three phase voltages are estimated in the third section used in practice, which force the actual current to
based on the formulas given in Table 1. Finally, the follow the reference signal, only two of the commonly
section four solves the differential equations (Eq.1) used schemes are implemented here: the Hysteresis
and computes the electromagnetic torque of the motor Current Controller and the PWM Current Controller.
(Eq.2). The "control" sub-VI also contains the Hysteresis
The "motor" sub-VI is customised and six inputs and Current Controller and the PWM Current Controller,
four outputs are defined, which are used to link to the which to produce the switching signals required by the
other sub-VIs in the program. One of the inputs is the "motor" sub-VI.
array of the inputs for the switching signals, which are The "control" sub-VI has six inputs and one output.
used to link the control signals of the power devices in The rotor position and the amplitude value of the
the inverter. The parameters of the motor and the current command are the two inputs that are used to
generate the reference current signals. The line current All of the above mentioned three sub-VIs are linked to
inputs are used as the current feedback signals (which obtain a closed loop drive VI for the PMAC motors
simulate the current transducers in practice) for the (Fig. 2). This final VI can be used to study the various
controller. The parameters of the controller and the operating modes of the PMAC motor drives as
calculation interval are two other inputs. In order to mentioned previously.
distinguish the motor and controller types, two
4. SIMULATION RESULTS
"mode" input signals are also implemented.
The experimental results presented in this paper have
Fig. 1 shows the block diagram of the VI that can be
been specifically chosen to demonstrate some of the
used to simulate the steady-state operation of the
typical operations that occur in the operation of the
PMAC motor drives. There are four "waveform
motors. The results provide a confirmation of the
graphs" in this block diagram, which display the line
validity of the current, the voltage, and the torque
currents, the phase voltages, the electromagnetic
estimation of the drive in both the steady-state and the
torque and the rotor position. Per-phase or all three-
transient operation.
phase currents and voltages can be displayed in the
graph. The PMAC motor used in this paper has the
parameters given in Table 2. The below motor
speed p o s ition
parameters are used to simulate both the BTPM and
BSPM motor drives in this paper.
3..
mode
TABLE 2- The test motor nameplate data and
measured parameters.
current Torque constant 0.31 Nm/A
Back emf constant , ke 0.417 V/rad/s
torque
Moment of inertia , J 0.0008 kgm2
parameter 3.. Number of poles 8
command Winding resistance , R 0.8 Ω
i n t e r v a l

Equivalent winding inductance , L 3.12 mH

v o lta g e

Firstly, the "drive" VI was used to simulate the

Figure 1: The block diagram of the drive for the steady-state operation of the motor. To obtain the
steady-state operation. simulation results, the parameters of the drive must be
set on the Front Panel (the user interface) of the VI.
In order to simulate the transient performance of the The input parameters are classified into three groups.
motor drive while maintaining the simplicity, the VI The first group is the motor parameters, which include
shown in Fig.1 is integrated into the "drive" sub-VI the number of pole pairs, the winding resistance R, the
that has five inputs and three outputs. The inputs are equivalent winding inductance L, the back EMF
the current command, the angular speed, the rotor constant ke, and the DC link voltage of the inverter.
position, the currents and the calculation interval. The The second group contains the current controller's
outputs include the torque, the rotor position and the parameters, which are the modulation frequency of the
currents. PWM Current Controller, the hysteresis bandwidth of
The inputs and outputs of rotor position and currents the Hysteresis Current Controller and the phase
are used to set the initial values. Furthermore, two advance/delay angle of the current command. The
additional sub-VIs are built to solve the mechanical third group is called the "mode" parameters, which are
equation (Eq.3) and to simulate a PI speed regulator. used to select the type of the back emf (trapezoidal or
sinusoidal), the wave shape of the reference current
i n t e r v a l
(rectangular or sinusoidal), and finally the type of the
current controller (hysteresis or PWM). In addition to
torgue the above parameters, the reference speed, reference
speed
Reference current and integration time step are set on the Front
Panel of the VI.
a n g u la r s p e e d
60
Fig. 3 shows the simulated results under the steady-
position position state operating condition of the test motor. Phase 1
current current
voltage, Line 1 current and the total electromagnetic
torque of the BTPM motor are shown on the left-hand
side in Fig.3. The simulation results under the
identical conditions, but with the BSPM motor are
given on the right hand side in Fig. 3
Figure 2: The closed-loop system VI.
 Voltage(V)
40.0

Voltage(V)
40.0

20.0 20.0

0.0 0.0

-20.0 -20.0

-40.0 -40.0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32

Current(A)
6.0 6.0
Current(A)

4.0 4.0

2.0 2.0

0.0 0.0

-2.0 -2.0

-4.0 -4.0

-6.0 -6.0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32

Torque(Nm)
4.0 4.0

3.0 3.0
Torque(Nm)

2.0 2.0

1.0 1.0

0.0 0.0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32

time(ms) time(ms)
(a) (b)
Figure 3: The simulated waveforms under the steady-state operations of the BTPM (a) and the BSPM (b)
motors: 500 rpm, Vdc = 50 V, Hysteresis Current Controller with a bandwidth o f ±0.2A, Im(ref) = 5A, no phase
advance or delay angle.
voltage(V)
voltage(V)

40.0 40.0
30.0
20.0 20.0
10.0
0.0 0.0
-10.0
-20.0 -20.0
-30.0
-40.0 -40.0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32

8.0
current(A)

8.0
current(A)

6.0 6.0

4.0 4.0

2.0 2.0

0.0 0.0
-2.0 -2.0

-4.0 -4.0

-6.0 -6.0

-8.0 -8.0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
torque(Nm)

torque(Nm)

4.0 4.0

3.0 3.0

2.0 2.0

1.0 1.0

0.0 0.0

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32

time(ms) time(ms)

(a) BTPM (b) BSPM

Figure4: The simulated waveforms under steady-state operation for the similar settings as in Fig. 3 but without
the current control.
 12.0 12.0

current(A)

current(A)
5.0 5.0

0.0 0.0

-5.0 -5.0

-12.0 -12.0
0 50 100 150 200 250 300 0 50 100 150 200 250 300 350 400

torque(Nm)
torque(Nm)

10.0 10.0

8.0 8.0

6.0 6.0

4.0 4.0

2.0 2.0

0.0 0.0
0 50 100 150 200 250 300 0 50 100 150 200 250 300 350 400

600.0
speed(rpm)

600.0

speed(rpm)
400.0 400.0

200.0 200.0

0.0 0.0
0 50 100 150 200 250 300 0 50 100 150 200 250 300 350 400

position(rad)
8.0
position(rad)

8.0

6.0 6.0

4.0 4.0

2.0 2.0

0.0 0.0
0 50 100 150 200 250 300 0 50 100 150 200 250 300 350 400

time(ms) time(ms)

(a) BTPM (b) BSPM

Figure 5: The simulated waveforms under the transient operation of the motor drives, Vdc = 56 V.

Fig.4 also shows the simulated waveforms of the
phase voltage, the line current and the electromagnetic
torque of the BTPM and the BSPM motor drive under
the steady-state operation. In this case, however, since
the reference current amplitude was set to 10A, the
line current failed to reach the limits set previously, so
that no current control was achieved.
Fig.5 illustrates the results obtained under the
transient operation of the motor drives. In this test, the
motor accelerates from standstill up to the constant
speed of 500 rpm. In this mode of operation, since the
current demand was set to a maximum, and DC link
voltage was kept constant, after starting of the motor,
the current profile reduces as the speed increases.
5. CONCLUSIONS
The paper demonstrated that the complex drive
structure that exist in the BTPM and BSPM motor
drives can be simplified by using the graphical
programming language LabVIEW. The simulation
structure in LabVIEW provides easy debagging and a
very friendly user interface.
The simulation tool can work under the steady-state as
well as the dynamic operating conditions and can
provide an accurate analysis tool for the brushless
PMAC motor drives. In addition, the simulation tool
can be used to study further control and parameter
estimation concepts in the motor drives. To achieve
this the VIs constructed in this paper can be utilised.
References
[1] Jacek F. and Mitchell W., "Permanent Magnet
Motor Technology," Marcel Dekker, Inc., New
York, 1997.
[2] Ertugrul N., "Position Estimation and
Performance Prediction for Permanent-Magnet
Motor Drives, " Ph.D. Thesis, University of
Newcastle upon Tyne, UK, 1993.
[3] Ertugrul N., Eric Chong, "Modelling and
Simulation of an Axial Field Brushless Permanent
Magnet Motor Drive, " European Power
Electrical Conference, Trondheim, Norway, 1997.
[4] Azizur M. R. and Ping Z., "Analysis of Brushless
Permanent Magnet Synchronous Motor, " IEEE
Transactions on Industrial Electronics, Vol. 43,
No.2, April 1996.
[5] LabVIEW Tutorial Manual, National Instruments
1996.