PERFORMANCE OF A THREE-PHASE PERMANENT MAGNET MOTOR OPERATING AS A SYNCHRONOUS MOTOR AND A BRUSHLESS DC MOTOR

A Thesis

Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College In partial fulfillment of the Requirements for the degree of Master of Science in Electrical Engineering in The Department of Electrical and Computer Engineering

by Sophie Sekalala B.EE., Electrical Engineering, City University of New York, 2003 August, 2006

ACKNOWLEDGEMENTS I would like to thank Dr. Ernest Mendrela, my major professor, for his valuable guidance, direction, support and enduring patience! Without him, this project would not have been possible. Also, thanks to Dr. Leszek Czarnecki and Dr. Bingqing Wei for agreeing to be on my advisory committee. Your time, advice and effort are highly appreciated. Words cannot explain how grateful I am to have had my beloved husband beside me throughout this project. His invaluable advice and support were very helpful and are very much appreciated. To the rest of my family, I say, thank you for being so patient with me, even when the road was bumpy.

Author.

ii

TABLE OF CONTENTS ACKNOWLEDGEMENTS .....ii LIST OF TABLES.........v LIST OF FIGURES..vi SYMBOLS.ix ABSTRACT............x CHAPTER 1: INTRODUCTION.....1 1.1 Overview of Thesis Subject......1 1.2 Objectives of the Thesis............3 1.3 Outline of the Thesis.........4 CHAPTER 2: DESCRIPTION OF THE MOTOR UNDER ANALYSIS....6 2.1 Similarities and Differences in the Structure of Synchronous and BLDC Motors........6 2.2 Data of the Motor under Study..........8 CHAPTER 3: BRUSHLESS DC MOTOR OPERATION......11 3.1 Dynamics of the BLDC Motor.......11 3.1.1 Dynamic Model of the BLDC Motor...11 3.1.2 Simulation of Motor Dynamics........16 3.2 BLDC Motor Performance in Steady-State.........19 3.2.1 BLDC Motor and Brush PM DC Motor: Similarities and Differences...20 3.2.2 Steady-State Model of BLDC Motor.......24 3.2.3 Performance Characteristics in Steady-State.......27 CHAPTER 4: SYNCHRONOUS MOTOR OPERATION..30 4.1 Dynamics of the Synchronous Motor.....30 4.1.1 Dynamic Model of the Synchronous Motor.....30 4.1.2 Simulation of Motor Dynamics........32 4.2 Dq model of the Synchronous Motor..36 4.3 Steady-State Model of the Synchronous Motor...40 4.3.1 Equivalent Circuit Model......40 4.3.2 Power and Torque Characteristics....42 4.3.3 Performance Characteristics of Motor in Steady-State.....45 4.4 Experimental Determination of Synchronous Motors Performance..46 CHAPTER 5: CONCLUSIONS.....54 REFERENCES58

iii

APPENDIX A: M-FILES FOR BLDC MOTOR OPERATION59 APPENDIX B: M-FILES FOR SYNCHRONOUS MOTOR OPERATION ...62 VITA.........66

iv

LIST OF TABLES 2.1 Data of the motor under study..9 2.2 Electrical and mechanical parameters of the motor.....10 3.1 Electromechanical parameters of the BLDC motor in steady-state ....19 4.1Electromechanical parameters of synchronous motor performed at rated conditions..35 4.2 Parameters of the equivalent circuit.....46 4.3 Experimental and simulation results of synchronous motor....51

LIST OF FIGURES 1.1 Scheme of 3-phase PM, salient-pole motor with buried magnets in the rotor...1 1.2 Scheme of 3-phase PM motor with surface mounted magnets on the cylindrical rotor2 1.3 Supply circuit scheme of synchronous motor3 1.4 Supply circuit scheme of brushless DC PM motor3 2.1 Motor with 3-phase concentrated windings...6 2.2 Diagram showing 3-phase windings with: (a) overlapping spread coils, (b) concentrated coils.7 2.3 Circuit diagram of 3-phase inverters: (a) uni-polar inverter, (b) bi-polar inverter8 2.4 Three-phase PM motor manufactured by Motorsoft Company.....9 3.1 Equivalent circuit of 3-phase PM motor..11 3.2 Schematic representation of equation 3.1....12 3.3 Illustration of the rotor position angle e ....14 3.4 Scheme of the mechanical system...15 3.5 (a) Block diagram for BLDC motor operation, (b) Block diagram behind subsystem in Fig. 3.5.a ....17 3.6 Waveforms of electromechanical quantities obtained from: (a) the start-up process of the BLDC motor, (b) steady-state...18 3.7 Diagram of DC commutator motor, which visually explains its operation.20 3.8 Diagram of a DC motor with 3 coils (phases) in the armature: (a) coils commutated by mechanical commutator, (b) coils commutated by electronic commutator (inverter).....21 3.9 Scheme of DC motor with 3-phase star connected winding: (a) with mechanical commutator (winding placed on rotor), (b) with inverter (winding is on stator and magnetic poles rotate).....22

vi

3.10 Rotor positions at two subsequent instants: (a) at time t1, (b) at time t2 and (c) mutual positions of the armature flux with respect to the field flux position at the two time instants....23 3.11 The torque (T) and electromotive force (E) waveforms: (a) with more phases, (b) with a 3-phase motor....24 3.12 Brush DC motor armature circuit diagram24 3.13 Brush motors equivalent circuit model with armature inductance...26 3.14 Speed and current-torque characteristics determined in steady-state, from the dynamic simulation..28 3.15 Characteristics obtained from both the BLDC motor model after it reaches steadystate and the brush DC motor model (continuous lines)......29 4.1 Explanation of synchronous motor operation..31 4.2 (a) ABC equivalent circuit diagram, (b) motor side of Fig. 4.2.a....31 4.3 Block diagram of synchronous motor operation..33 4.4 Waveforms of electrical and mechanical quantities, (a) after motor was connected to supply voltage of 42 V at rated rotor speed, (b) in steady-state.....34 4.5 Phasor diagram showing angles, voltages and current35 4.6 Waveform of phasor quantities with a load torque of 0.86 N.m at steady-state.36 4.7 Scheme of ABC and dq system...37 4.8 (a) Block diagram for dq system of the synchronous motor (b) subsystem behind the block actual motor in Fig. 4.8.a (c) subsystem behind the block dqabc in Fig. 4.8.b39 4.9 Synchronous machine equivalent circuit: (a) armature reaction reactance Xar, armature leakage reactance Xal, (b) synchronous reactance Xs, (c) armature impedance Zs..............................................................................................................41 4.10 Phasor diagram drawn for Ra = 0...43 4.11 Torque-power angle characteristic.....45

vii

4.12 (a) Characteristics of steady-state and dynamic models for the synchronous motor (b) angle delta () and power factor cos (fi) obtained from the steady-state (continuous lines) and dynamic models.......47 4.13 Electric drives board..48 4.14 dSPACE control box.50 4.15 Experimental setup showing overall layout of all equipment used...51 4.16 Simulink model of the experiment.52 4.17 Layout of the experiment...52 4.18 Current waveforms from: (a) experiment, (b) simulation..53

viii

SYMBOLS E line-to-line electromotive force I average armature current K constant R armature resistance T torque on the shaft To static friction torque V source voltage Xs synchronous reactance m rotor angular speed angle, delta ohms flux flux linkage angle, fi angle, theta

ix

ABSTRACT The performance of a three-phase permanent magnet (PM) motor operating as a synchronous motor and brushless DC (BLDC) motor is discussed. The PM motor, when operating as a synchronous machine, is supplied with constant frequency, meaning constant rotor speed operation. When operating as a BLDC motor, the supply frequency changes according to the actual rotor speed. It means the speed signal is fed back to the controller, which generates the appropriate frequency to supply the stator winding. This means the BLDC motor cannot operate without a position sensor. The objective of the project was to analyze and compare the performance of the motor operating as a BLDC and synchronous motor. To do this, mathematical models for both operation modes (that allowed us to analyze the performance in dynamic conditions) were proposed. The analysis was carried using the results obtained from simulation done with the MATLAB/SIMULINK software package. To analyze the motors performance in steady-state conditions, simpler motor models were proposed and calculations carried out using MATLAB m-files. The results obtained from the two types of models were compared and a good match was observed. To verify the simulation mathematical model, an experiment was performed on the real object where the synchronous motor was tested. The motor was supplied from a 3phase inverter board (controlled by dSPACE controller) which operated in conjunction with MATLAB/SIMULINK. This allowed us to supply the motor with variable frequency at a constant voltage to frequency ratio. Despite the small discrepancies between the experimental and simulation results, the mathematical model used to analyze the synchronous motor was verified.

CHAPTER 1: INTRODUCTION 1.1 Overview of Thesis Subject AC permanent magnet (PM) motors can perform both as synchronous and brushless DC motors. In each case, the motor consists of a wound stator and a rotor. The stator may have a single-phase or multi-phase winding which is sometimes called the armature winding (Fig. 1.1). The rotor just has permanent magnets. There are two types of rotors. Salient-pole rotor (Fig. 1.1) mostly for low-speed machines. Cylindrical rotor (Fig. 1.2) usually for high-speed machines.

Fig. 1.1 Scheme of 3-phase PM, salient-pole motor with buried magnets in the rotor.

In both cases, the PM can be attached to the rotor surface (see Fig. 1.2) or it can be buried. (Fig. 1.1) Since there is no winding in the rotor (that would be supplied through the brushes), the AC PM machines are also called brushless PM machines. The rotor of the AC PM motor rotates synchronously with the magnetic field generated by the stator

winding. At the motor start-up, the rotor has zero speed. Due to its high inertia, it cannot instantaneously shoot to its synchronous speed. However, this can happen if we supply the stator winding with frequency, rising gradually from 0 to its rated value. To do this, the motor has to be supplied from a variable frequency inverter.

Fig. 1.2 Scheme of 3-phase PM motor with the surface mounted magnets on the cylindrical rotor. The frequency can be controlled by imposing the desired reference frequency (see Fig. 1.3), or the motor itself can set an appropriate frequency value required for its actual speed. In the latter case, the motor must be equipped with a speed or position sensor (Fig. 1.4). The AC PM motor with the frequency controlled as described in the former mode above operates as a synchronous motor. Conversely, if operating with self controlled frequency as explained in the latter mode above, then this operation is that of a brushless DC motor. The AC PM motor when operating in each of the above modes, i.e. synchronous or brushless DC motor, performs differently. Thus, its electromechanical characteristics will

differ significantly. The purpose of this project is to determine and study these differences in performance for each particular mode of operation.

Fig. 1.3 Supply circuit scheme of synchronous PM motor.

Fig. 1.4 Supply circuit scheme of brushless DC PM motor.

1.2 Objectives of the Thesis The objectives of this project are; To determine the performance of the permanent magnet motor while operating as a brushless DC motor and synchronous machine.

To study the difference in operation between the two mentioned above.

The tasks to be accomplished are as follows; A Literature study on: - Permanent magnet AC motors supplied from DC/AC converters. - Control of the motor using the DSP controller board. Modeling of the PM 3-phase synchronous and PM 3-phase brushless DC motors in dynamic and steady-state conditions. Determination of the synchronous motors performance in steady-state and dynamic conditions with the application of the DSP controller board. Comparison of simulation results with the ones obtained experimentally and conclusions. 1.3 Outline of the Thesis Chapter 2 describes the object of this research which is the motor. It shows a schematic drawing of the afore mentioned and its construction as well as the motors specifications/data. Chapter 3 discusses the brushless DC motors operation. This is based on the results obtained from the simulation carried out on the dynamic and steady-state model of the motor. Chapter 4 presents the synchronous motors operation. Both the dynamic and steady-state models are shown. Also included is the experimental determination of the synchronous motors performance. The theoretical and experimental results are then compared.

Finally, Chapter 5 offers a conclusion, including a general description of what was done in the project and a comparison of the motors performance in the dynamic and steady-state models.

CHAPTER 2: DESCRIPTION OF THE MOTOR UNDER ANALYSIS 2.1 Similarities and Differences in the Structure of Synchronous and BLDC Motors As mentioned in Chapter 1, any PM single-phase or multi-phase motor can operate as a synchronous or BLDC motor. The particular mode of operation depends on the supply and control circuit. This is generally true if the stator has conventional windings placed in slots and distributed symmetrically around the stator periphery (Fig. 1.1).

Fig. 2.1 Motor with 3-phase concentrated windings.

The BLDC motor, due to its different type of supply, may also have other types of windings with concentrated coils. Such a motor is shown in Fig. 2.1. The BLDC motor with concentrated windings evolved from the stepper motor structure. In fact, the stepper motor operating at high speed does not differ much from the BLDC motor operation. The winding diagrams with spread and concentrated coils are shown in Fig. 2.2. The 3-phase windings with distributed coils are supplied from a bi-polar inverter (Fig. 2.3b) and those with concentrated

coils are usually supplied from a uni-polar converter (Fig. 2.3a). As mentioned in Chapter 1, the BLDC motor has to be equipped with a position sensor which informs the controller what the position of the rotor magnetic pole is, with respect to the particular stator phase winding. This is done in order to switch the motor ON and OFF. The position sensors that are applied are usually optical and Halls [1]. The encoder will act as a position sensor, in the experiment carried out in this project. The BLDC motor can also operate by using sensor-less control. In this case, the position of the rotor is known from the value of the back EMF with respect to the particular phase [2]. (a)

(b)

Fig. 2.2 Diagram showing 3-phase windings with: (a) overlapping spread coils, (b) concentrated coils.

(a)

+
T 1 T 2 T 3

(b)

+
T 1 T 2 T 3

T 4

T 5

T 6

Fig. 2.3 Circuit diagrams of 3-phase inverters: (a) uni-polar inverter, (b) bipolar inverter. 2.2 Data of the Motor under Study The object used throughout this research, i.e. whose data is being used and whose performance is being analyzed, is the 3-phase PM brushless DC motor, manufactured by Motorsoft Company. In Table 2.1 is the data shown on the name plate. As seen in Fig. 2.1, the motor has three connectors for the three phases, A, B and C. During the experiment, these are connected to the drives board to supply the motor. Also, the speed encoder seen next to the connectors is connected to the dSPACE controller box (more details will be revealed in Chapter 4.3). In addition to the name plate data shown in Table 2.1, more data provided by the manufacturer of this motor, is listed in Table 2.2.

Table 2.1 Data of the motor under study. Model #: 2925

Serial Number: BR1057 Motor Type: 10 pole 3-phase BLDC Rated Voltage: 42 V DC Rated Speed: Rated Power: 3600 RPM 210 Watts

Rated Current: 7A (rms)/Ph

The motor is shown in Fig. 2.4.

Fig. 2.4 Three-phase PM motor manufactured by Motorsoft Company.

Table 2.2 Electrical and mechanical parameters of the motor. per phase resistance, Rph supply voltage, Vs per phase reactance, Xph per phase inductance, Lph moment of inertia, Jeq constant, KE rated load torque, TL frequency 0.077 ohms 42 V DC 0.396 ohms 0.210 mH 7.895e-7 Kg.m2 0.073 V.s/rad 0.56 N.m 60 Hz

The motor has a cylindrical rotor with surface mounted PMs (see Fig. 1.2).

10

CHAPTER 3: BRUSHLESS DC MOTOR OPERATION 3.1 Dynamics of the BLDC Motor 3.1.1 Dynamic Model of the BLDC Motor It is assumed that the BLDC motor is connected to the output of the inverter, while the inverter input terminals are connected to a constant supply voltage, as was shown in Fig. 2.3. The equivalent circuit model that refers to this circuit diagram is shown in Fig. 3.1. Another assumption is that there are no power losses in the inverter and the 3-phase motor winding is connected in star.

isk

iA DC vA
iB

RA

LA

eA

RB

LB

eB

Vs AC -

vB iC vC

RC

LC

eC

Fig. 3.1 Equivalent circuit of 3-phase PM BLDC motor.

The equivalent circuit shown in Fig.3.1 can be represented by the circuit diagram in Fig. 3.2. The equations that govern this model are as follows. v A = v N + v sA v B = v N + v sB vC = v N + v sC where: (3.1)

11

vsA, vsB, vsC are the inverter output voltages that supply the 3 phase winding. vA, vB, vC are the voltages across the motor armature winding. vN voltage at the neutral point.

v SA

iA

vA

v SB

iB

vB

v SC

iC

vC

vN

Fig. 3.2 Schematic representation of equation 3.1. [5]

For a symmetrical winding and balanced system, the voltage equation across the motor winding is as follows:
v A R A v = 0 B v C 0 0 RB 0 LA 0 i A d 0 i B + dt LBA RC iC LCA L AB LB LCB L AC i A e A LBC i B + e B LC iC eC

(3.2)

or in the shortened version representing the vector:

Va = R a I a + d La I a + Ea dt

(3.3)

Since RA = RB = RC =Ra, the resistance takes the following vector form:

Ra 0 0 Ra = 0 R a 0 0 0 R a

(3.4)

12

As for the inductances, since the self and mutual inductances are constant for surface mounted permanent magnets on the cylindrical rotor (see fig.1.2), and the winding is symmetrical: L A = LB = LC = L; and L AB = LBC = LCA = LBA = L AC = LCB = M Hence the inductance takes the form: L M M La = M L M M M L For a Y-connected stator winding, i A + i B + iC = 0 Therefore, the voltage takes the following form:
v A R A v = 0 B v C 0 0 RB 0 Ls 0 i A d 0 i B + 0 dt RC iC 0 0 Ls 0 0 i A e A 0 i B + e B Ls iC eC

(3.5)

(3.6)

(3.7)

(3.8)

where the synchronous inductance, Ls = L M The angle between a particular phase and the rotor, at any given time, is called e . Fig. 3.3 illustrates the position of this angle, with respect to phase A for example. Since phase A is chosen as the reference (see Fig. 3.3), the electromotive forces written in the form of a matrix, Ea take the form:
sin e 2 d e KE sin( e ) Ea = 3 dt p sin( e 4 ) 3

(3.9)

13

Fig. 3.3 Illustration of the rotor position angle e .

To link the input voltages and currents of the inverter with those of the output, the power equality equation, Pin = Pout is assumed at both sides. From this, the inverter input current is:
isk = 1 (iAvsA + iB vsB + iC vsC ) vs

(3.10)

where vsA, vsB and vsC are the phase voltages that supply the motor. The mechanical system shown schematically in Fig. 3.4 is defined by the following equations:

Tem = J eq

d m + B m + TL dt

(3.11)

where J eq = J M + J L is the equivalent moment of inertia, and JM, JL are the moments of inertia of the motor and load respectively, B-friction coefficient and TL-load torque.

14

Fig. 3.4 Scheme of the mechanical system.

The electromagnetic torque for this 3-phase motor is dependent on the current (i), speed ( m ) and electromotive force (e). The equation is:

Tem =
where,

e Ai A

eB iB

eC i C

= K E ( f a ( e ) i A + f b ( e ) i B + f c ( e ) iC )

(3.12)

f a ( e ) = sin( e ) 2 ) 3 4 f c ( e ) = sin( e ) 3 f b ( e ) = sin( e

(3.13)

Combining all the above equations, the system in state-space form is;
x = Ax + Bu

(3.14)

x = [i A iB
Rs L s 0 A= 0 K E ( f a ( e )) J 0

iC
0 Rs Ls 0

r e ]t
0 0 Rs Ls K E ( f c ( e )) J 0 K E ( f a ( e )) Ls K E ( f b ( e )) Ls K E ( f c ( e )) Ls D J P 2 0 0 0 0 0

(3.15)

(3.16)

K E ( f b ( e )) J 0

15

1 L s 0 B= 0 0 0

0 1 Ls 0 0 0

0 0 1 Ls 0 0

0 0 0 1 J 0

(3.17)

u = [v A

vB

vC

TL ]

(3.18)

3.1.2 Simulation of Motor Dynamics

The simulation of the BLDC motor was done using the software package MATLAB/SIMULINK. For this purpose, the motors block diagram was constructed, as shown in Fig.3.5. After running the simulation, the speed, torque, current, input and output power waveforms were recorded and analyzed. Fig. 3.6 shows the start-up process of the motor. The inverter was supplied with a voltage of 42V DC and the motor loaded with a rated torque of 0.56 Nm. Phase A voltage is a square wave of 21V DC. However, notice how the phase current is distorted from the square wave shape to something between a sine-wave and a square wave. This is a result of the inductance effect. After the initial start-up, the speed oscillates around 377 rad/s. As expected, the load torque is a constant straight line of 0.56 Nm. The electromagnetic torque depends on phase currents, constant K, speed and electromotive force (recall equation 3.12). Therefore, all these quantities affect the appearance of its waveform. After the start-up process, the motor reaches steady-state at around 0.016 seconds. Table 3.1 lists its parameters in this state. The motor efficiency was calculated as:

16

Fig. 3.5.a Block Diagram for BLDC motor operation.

Fig. 3.5.b Block diagram behind subsystem shown in Fig.3.5.a.

17

(a)

(b)
500

600 speed [rad/s] 400 200 0 0 0.005 0.01 0.015 time [s] 0.02 0.025 0.03
200 0.026 0.028 time [s] 0.03 400

300

phase A voltage [V] 20

20

-20 0 0.005 0.01

phase A current [A] 0.015 time [s] 0.02 0.025 0.03

-20 0.026 0.028 time [s] 0.03

2 1.5 1 0.5 0 -0.5 0 0.005 0.01 load torque [N.m] 0.015 time [s] 0.02 0.025 0.03 electromagnetic torque [N.m]

1 0.8 0.6 0.4 0.2 0.026 0.028 time [s] 0.03

800 source current [A] 600 400 200 0 0 0.005 0.01 0.015 time [s] 0.02 0.025 0.03

450 400 350 300 250 0.026 0.028 time [s] 0.03

Fig. 3.6 Waveforms of electromechanical quantities obtained from: (a) the start-up process of the BLDC motor, (b) steady-state.

18

Eff (%) =

PL 100 Pin

(3.19)

where the load power on the motor shaft, PL = TL m and the input power of the inverter equal to the input power of the motor, Pin = Vs I s (3.21) (3.20)

The above quantities were calculated as average values while the current as an rms value. The waveforms were plotted using m-file bldcfilesloader shown in Appendix A.

Table 3.1 Electromechanical parameters of BLDC motor in steady-state.

Supply voltage Output power Input power Load Torque Efficiency Electromagnetic torque Speed Current 42 V DC 176.9 W 207.85 W 0.56 N.m 85.1% 0.592 N.m 315.9 rad/sec 4.427 A

3.2 BLDC Motor Performance in Steady-State

The operation of the BLDC motor in steady-state may be regarded as that of a DC motor with commutators; hence the brush DC motor model may be used for analysis of

19

the BLDC motors performance in this state. Section 3.2.1 explains the similarities and differences between these two types of motors.

3.2.1 BLDC Motor and Brush PM DC Motor: Similarities and Differences

The brush DC motor is excited by either field windings or permanent magnets. In both cases, they are placed on the stator. The armature winding, which is placed on the rotor, consists of a number of coils (Fig. 3.7). When the rotor turns, the current of the subsequent coils that are approaching stationary brushes, are commutated. Due to the commutator, the resultant magnetic flux a produced by the coils is always perpendicular to the field flux despite the current changes in the rotor coils [2]. To digress just a bit, the subject of brushes is the first and obvious difference between the brush and BLDC motors. As the name suggests, the brushless DC motor has no brushes and is electronically commutated with permanent magnets placed on the rotor.

Fig. 3.7 Diagram of DC commutator motor, which visually explains its operation.

20

The commutator can be regarded as a mechanical inverter. This is seen clearly when the DC motor with three delta-connected coils or phases is considered as shown in Fig. 3.8a. At a particular time instant t1, coils A, B and C are supplied, generating the resultant flux, a perpendicular to the field flux (see Fig. 3.8). The same position of a can be achieved if the coils are supplied from a DC source through a 3-phase inverter as shown in Fig. 3.8b. This operation would be tantamount to that seen in the brushless DC motor.

(a)
N T, n
A

(b)
+ 1 2
i

4 _

A
S

Fig. 3.8 Diagram of a DC motor with 3 coils or phases in the armature: (a) coils commutated by mechanical commutator, (b) coils commutated by electronic commutator (inverter).
In the case of star (wye) connected windings shown in Fig. 3.9, two coils are energized at any one time. Again, this can also be achieved by using a 3-phase inverter as shown in Fig. 3.9b. The afore mentioned figure shows the position of the motor and coils energized by commutator and inverter at three different time instants. No changes are observed in the resulting flux position with respect to the stationary field. The only

21

minute changes observed in the position of the flux a, are between two subsequent winding commutations (see Fig. 3.9) [2].

(a)
N T, n
C

(b)
+ 1
A

2
i

i
aC aA

4 _

A
S

(i)
N T, n

+ 1
A

2
i

3 6

4 _

A
S

(ii)
+
N T, n
A

1
B

2
i

_
C

4 _

A
S

Fig. 3.9 Scheme of DC motor with 3-phase star connected winding: (a) with mechanical commutator (winding placed on rotor), (b) with inverter (winding is on stator and magnetic poles rotate).
22

Due to the change in position, i.e. = 60 (see Fig. 3.10c), the resultant interaction of stationary flux f and flux a, (also known as electromagnetic torque Tem) changes with time, producing some ripples in the torque (see Fig. 3.11). The more phases we have, the smoother the torque waveform will be [3]. Another essential difference to be noted between the brush and brushless DC motors is as follows. In a DC commutator motor, the armature winding, which is mechanically commutated, is placed on the rotor, while the field windings or permanent magnets are placed on the stator. In the BLDC motor, the reverse is true. This is because the armature winding in the former is a selfcommutated winding caused by the rotating commutator. On the other hand, the winding of the latter can be commutated by a stationary electronic inverter. The moment of commutation in the conventional DC motor is determined by the position of the coil with respect to the stationary brushes. In BLDC motors, it is determined by the position of the sensor signal. This, in turn, means that BLDC motors cannot operate without position sensors.

(a) t = t1
N
n

(b) t = t2
N
n

aB

C
aC

B
aB

C
aC

Fig. 3.10 Rotor positions at two subsequent instants: (a) at time t1, (b) at time t2 and (c) mutual positions of the armature flux with respect to the field flux position at the two time instants. (fig. contd.)

23

(c)
60 0

T (E)

t1

t2

Fig. 3.11 The torque (T) and electromotive force (E) waveforms: (a) with more phases, (b) with a 3-phase motor. 3.2.2 Steady-State Model of BLDC Motor
When the BLDC motor is in steady-state, it is simply operating as an ordinary brush DC motor. The model for this brush model, without any inductance, is shown in Fig. 3.12.

Fig. 3.12 Brush DC motor armature circuit diagram.

24

The equations that define this model are as follows. T = K I To E =Km V = E + R I where: T torque on the shaft, E line-to-line electromotive force, m rotor angular speed, To static friction torque, K constant, R armature resistance, I average armature current V source voltage. From equations 3.23 and 3.24 we obtain, (3.22) (3.23) (3.24)

m =

V RI K

(3.25)

or m =

V T + To R 2 K K

(3.26)

In the brush DC motor, at any time only a small part of the winding is commutated during the motor operation (see Fig. 3.9) and the inductance across this part does not affect the motor performance much. However, one negative effect exhibited by this commutator is that some sparks occur between the brushes and commutator. In the BLDC motor with a 3-phase winding, at any time one third of the winding is commutated (Fig. 3.10) and the current in the commutated part influences the motor
25

performance. The higher the motor speed, the shorter the commutation period and the faster the current changes in the commutated part of the winding. This change contributes to the relatively high voltage drop across this part, hence the inductance Lcom of this part (according to the equation vcom = Lcom

dia ) may significantly influence the motor dt

performance, in particular the speed-torque characteristics [3]. This means the winding inductance in the equivalent motor circuit should be considered. If the motor inductance has to be included, then the circuit diagram will appear as follows.

Fig. 3.13 Brush motors equivalent circuit model with armature inductance.

The voltage equation then changes to include the inductance. It is,

V = E + R I + K L m I n

(3.27)

where KL is a constant and the current is to the power n, usually between 1.5-2. From equation 3.23 and 3.27, the speed is,

m =

V RI K E + K L .I n

(3.28)

or expressed in terms of electromagnetic torque (recall torque Tem=KI),

26

Tem K m = T K E + K L ( em ) n K V R All other equations remain the same as previously.

3.2.3 Performance Characteristics in Steady-State

(3.29)

To find out how the analyzed BLDC motor performs in steady-state and to determine which brush motor model may be used for analysis, the motors electromechanical characteristics are first determined from the dynamic model. The characteristic that reflects the influence of the inductance is speed-torque curve. This along with the currenttorque curve is drawn in Fig. 3.14. According to equation 3.26, the motor model with no inductance (Fig. 3.12) gives the speed-torque characteristic as a straight line. For the motor model with inductance (Fig. 3.13), the speed-torque characteristic according to equation 3.27 is non-linear (Fig. 3.14). On the basis of these characteristics, the particular steady-state model was selected and its parameters determined. The brush model with no inductance was chosen because, as Fig. 3.14 shows, the speed-torque characteristics resemble those of that model. Also, the straight line characteristic shows that inductance played no part in affecting its shape; otherwise it would have appeared as a bent line. The constant KE = 0.23 was calculated using equation 3.22 or simply looking at Fig. 3.14, calculating the slope of the current curve and taking its inverse. The static friction torque constant To was determined to be 0.068 N.m. With the use of equations 3.22-3.28, the performance characteristics were calculated applying the MATLAB software (see m-file brush1 in Appendix A). The characteristics were plotted in Fig. 3.15.
27

3500 3000 2500 2000 1500 1000 500 0 speed/5 [rad/s]

current/200 [A]

0.5

1 1.5 torque [Nm]

2.5

Fig. 3.14 Speed and current-torque characteristics determined in steady-state, from the dynamic simulation.

Fig. 3.15 shows the characteristics of both models, i.e. the selected brush DC motor and the BLDC motor determined after it reaches steady state. By plotting both characteristics on the same graph, we are able to see how the BLDC motor (after it reached steady-state) performed compared to the brush motor. The brush motors characteristics are those represented by solid lines. From these characteristics, we see that the brush motor model describes the BLDC motor in steady-state relatively well. Regarding the particular characteristic, as the torque increases, the motor slows down and hence the speed decreases. The mechanical power was determined from the equation: Pm = mTL The motor efficiency was calculated as follows: Eff (%)= (3.30)

Pm 100 Pin
28

(3.31)

where Pin = VI Looking at Fig. 3.15, it is observed that at high torques, the speed-torque curve from the dynamic model becomes more flat. This means that the inductance does have some influence (albeit relatively small) on the motor performance. Since within the range of the rated torque (0.56 Nm) both characteristics match, the model without inductance can be applied to the analysis of this particular BLDC motor.
3500 3000 2500 2000 1500 speed/5 [rad/s] 1000 500 mechanical power [W] 0 0 0.5 1 1.5 torque [Nm] 2 2.5

efficiency/25 [%] current/200 [A]

Fig. 3.15 Characteristics obtained from both the BLDC motor model after it reaches steady-state and the brush DC motor model (continuous lines).

29

CHAPTER 4: SYNCHRONOUS MOTOR OPERATION 4.1 Dynamics of the Synchronous Motor 4.1.1 Dynamic Model of the Synchronous Motor

The 3-phase armature winding of the synchronous motor is connected to a 3-phase ac supply. The stator currents produce a rotating magnetic flux, as in 3-phase induction machines. The magnetic flux of the PMs is steady with respect to the rotor. To produce torque, these two magnetic fluxes cannot move with respect to one another. This means that the rotor should rotate with the same speed as the rotating flux produced by the stator (Fig. 4.1). The rotor follows the stator rotating field by an angle (Fig. 4.1). The motor being supplied with rated frequency does not develop any torque at zero speed. For the motor to operate, the rotor should first reach a synchronous speed. This can be done in three ways: by driving the rotor to the synchronous speed (with an external machine), by starting the rotor (which has to be equipped with a starting cage) as in induction motors, or by supplying the stator winding with variable frequency, beginning from zero to the rated frequency. The latter method can be used if the winding is supplied from a frequency converter. In the experiment (to be discussed in detail later on in this chapter), a frequency converter was used (recall Fig. 1.3 for the schematic arrangement) to try and bring the motor from standstill to its desired speed. The equivalent circuit for the dynamic model of the synchronous motor can be represented by two systems; namely the ABC system, which involves three phases and the dq system which involves two phases (see details in section 4.2). The ABC systems

30

model is very similar to that of the BLDC motor. Fig. 4.2 shows this ABC equivalent circuit diagram.

a s f

f-magnetic flux due to PM excitation . a-flux due to stator current.

s -synchronous speed.

Fig. 4.1 Explanation of synchronous motor operation.

isk

iA

RA

LA

eA

=
Vs

vA
iB

RB

LB

eB

~
-

vB iC vC

RC

LC

eC

Fig. 4.2.a ABC equivalent circuit diagram.

v SA iA vA

v SB

iB

vB

v SC

iC

vC

vN

Fig. 4.2.b Motor side of Fig. 4.2.a.

31

The motion and voltage equations that define the model are the same as those of the BLDC motor (see equations 3.1-3.18). The only difference between the two models is the type of supply voltage.
4.1.2 Simulation of the Motor Dynamics

The simulation of the synchronous motor was done using the software package MATLAB/SIMULINK. For this purpose, the motors block diagram was constructed, as shown in Fig.4.3. Also keep in mind that section 4.2 offers an alternative block diagram (dq system) to simulate the synchronous motors performance. After running the simulation, the speed, torque, current, input and output power waveforms were recorded and analyzed using m-file synchro (see appendix B). Fig. 4.4.a shows the waveforms of the electrical and mechanical quantities after the stator was supplied with a sinusoidal voltage of 24V (rms), 300Hz frequency and an initial speed of 377 rad/s (the motors rated speed). The motor was loaded with a rated torque of 0.56 N.m. Fig. 4.4.b shows the same waveforms after steady-state is reached. Looking at the waveforms, the power angle is negative, since this is motor operation. The current and voltage have a sinusoidal shape because, as Fig. 4.3 shows, the input was also sinusoidal. The motor efficiency is calculated in the same way as that of the BLDC motor (see equation 3.19). After the motor reached steady-state, the electromechanical parameters were recorded in Table 4.1. As mentioned earlier, the angle refers to the mutual position of the supply voltage and electromotive force, for the same phase. If the stator resistance Ra is negligible, this angle is equal to the angle between the d axis of the rotor and the stator flux.

32

During this motoring operation, this stator flux a pulls the rotor. The phasor diagram with the angle (and other quantities) is shown in Fig. 4.5. In Fig. 4.6, are the waveforms of the quantities used to construct the phasor diagram (Ef, iA, VA).

Fig. 4.3 Block diagram of synchronous motor operation.

33

(a)
30 20 voltage [V]

(b)
40 20
10 0 -10 -20 current[A] -30 0 0.01 0.02 time [s] 0.03 0.04

0 -20 -40 0.03

0.035 time [s]

0.04

10 pow er angle [rad]

-6.6 -6.7 -6.8 -6.9 0.03

-10

-20

0.005

0.01

0.015

0.02 0.025 time [s]

0.03

0.035

0.04

0.035 time [s]

0.04

600 speed [rad/sec]

377.5

400

377
200

0.005

0.01

0.015

0.02 0.025 time [s]

0.03

0.035

0.04

376.5 0.03

0.035 time [s]

0.04

1.5 1 0.5 electromagnetic torque[Nm]

1 0.8 0.6 0.4

0 -0.5

load torque[Nm]

0.2 0 0.03 0.035 time [s] 0.04

0.005

0.01

0.015

0.02 0.025 time [s]

0.03

0.035

0.04

Fig 4.4 Waveforms of electrical and mechanical quantities: (a) after the motor was connected to a supply voltage of 42V at rated rotor speed, (b) in steady state. (fig. contd.)
34

500 400 300 200 100 0 -100 0 0.005 0.01 0.015 0.02 0.025 time [s] 0.03 0.035 0.04 Pout [W] Pin [W]

500 400 300 200 100 0 -100 0.03 0.035 time [s] 0.04

Table 4.1 Electromechanical parameters of synchronous motor performed at rated conditions.

Supply voltage Output power Input power Load Torque Efficiency Electromagnetic torque Speed Current

24 V AC 211 W 244 W 0.56 N.m 86.5 % 0.598 N.m 377 rad/sec 9.19 A

Fig. 4.5 Phasor diagram showing angles, voltages and current.

35

30 iA [A] 20 10 0 -10 -20 -30 0.036 0.0365 0.037 0.0375 0.038 0.0385 0.039 0.0395 time [s] delta fi vA [V] ef [V]

0.04

Fig. 4.6 Waveform of the phasor diagram quantities with a load torque of 0.86 N.m at steady-state. 4.2 Dq Model of the Synchronous Motor

As mentioned earlier, the synchronous motor performance can be analyzed in two ways. Section 4.2 focuses on the second system known as dq, which normally stands for direct and quadrature axes as opposed to the conventional x, y and z axes. This means the system will now be 2-phase instead of 3. Fig. 4.7 shows how the two systems relate to each other. The equations that relate the stator currents of the two systems are as follows [4].
i sd (t ) i (t ) = sq 2 4 i (t ) cos( da ) cos( da ) cos( da ) a 2 3 3 ib (t ) 2 4 3 sin( da ) sin( da ) sin( da ) ic (t ) 34444444 34 44444444 14 4 2 44 3
[Ts ]abc dq

(4.1)

36

Fig. 4.7 Scheme of ABC and dq system.

cos( da ) sin( da ) ia (t ) i (t ) = 2 cos( + 4 ) sin( + 4 i sd da da b 3 i sq 3 3 ic (t ) 2 2 cos( da + ) sin( da + 3 3 4 1444444 4 2444444 3

[Ts ]dq abc

(4.2)

where [Ts]abcdq is the transformation matrix to transform stator a-b-c phase winding currents to the corresponding dq winding currents and [Ts]dqabc is the transformation matrix in the reverse direction. The same transformation matrix relates the flux linkages and voltages. The stator winding voltages are:
v sd = Rs i sd +
v sq = Rs i sq +

d sd m sq dt
d sq + m sd dt

(4.3)

(4.4)

37

where the speed of the equivalent dq windings is d = m (in electrical rad/s) in order to keep the d-axis always aligned with the rotor magnetic axis. The stator d and q- winding flux linkages can be written as follows:

sd = Ls i sd + fd
and,

(4.5)

sq = Ls i sq

(4.6)

where Ls= Lls + Lm and fd is the flux linkage of the stator d-winding due to flux produced by the rotor magnets (recognizing that the d-axis is always aligned with the rotor magnetic axis). The electromagnetic torque is,
Tem = p ( sd i sq sq i sd ) 2

(4.7)

Substituting for flux linkages in the above equation for the non-salient 10-pole motor,
Tem = p p [( Ls i sd + fd )i sq Ls i sq i sd ] = fd i sq 2 2

(4.8)

The acceleration is determined by the difference of the electromagnetic torque and the load torque (including friction torque Tfr = mechB) acting on Jeq , the combined inertia of the load and the motor T TL mech B d mech = em dt J eq where mech is the speed in rad/s. Fig. 4.8.a shows the block diagram for the dq system of the synchronous motor. The subsytem actual motor in Fig. 4.8.a is shown in Fig. 4.8.b. Further still, the subsystem dqabc in Fig. 4.8.b is detailed in Fig. 4.8.c.
38

(4.9)

Fig. 4.8.a Block diagram for dq system of the synchronous motor.

Fig. 4.8.b Subsystem behind the block actual motor in Fig. 4.8.a.

39

Fig. 4.8.c Subsystem behind the block dqabc in Fig. 4.8.b.

4.3 Steady-State Model of the Synchronous Motor

Just like the BLDC motor, the synchronous motor can be analyzed in steady-state without using a dynamic model. Discussed in this section, is its performance in this state; the motor is analyzed on the basis of the steady-state equivalent circuit and then its results compared to those obtained from the dynamic model.
4.3.1 Equivalent Circuit Model

The equivalent circuit model will be derived on a per-phase basis. The PMs produce the flux f . The current Ia in the stator produces flux a . Part of this, known as the leakage flux al , does not link the rotor. A major part known as the armature reaction flux ar is linked with the rotor. The resultant air gap flux is therefore the sum of two component fluxes f and a .Each flux component (and the resultant flux) induces a voltage component in the stator winding:

40

 f E f , ar E ar , al E al , E.
The excitation voltage Ef can be determined from the open circuit test. However, the armature reaction voltage, Ear and the leakage flux voltage Eal depend on the armature current. Therefore they can be represented as voltage drops across the reactances: Xar reactance of armature reaction and Xal leakage reactance [5]. This is shown in the equivalent circuit in Fig. 4.9. The relationship between these voltages is:
E = E f E ar

(4.10) (4.11)

or E = E f jX ar I a The voltage equation for the whole circuit (Fig. 4.9.a) is:
V = E f Ra I a jX al I a jX ar I a

(4.12)

(a)

Ia

X ar Ear

X al Eal E

Ra RaIa V ZL

Ef

(b)

Ia

Xs

Ra

Ef

ZL

Fig. 4.9 Synchronous machine equivalent circuit: (a) armature reaction reactance Xar, armature leakage reactance Xal, (b) synchronous reactance Xs, (c) armature impedance Zs. (fig. contd.)

41

(c)

Ia

Zs

Ef

ZL

or (Fig. 4.9.b)
V = E f Ra I a jX s I a

(4.13)

or (Fig. 4.9.c)
V = E f Zs I a

(4.14)

where: X s = X ar + X al - synchronous reactance

Zs = Ra + jX s - synchronous impedance
4.3.2 Power and Torque Characteristics

The complex power at the stator winding terminals is:

S = mVI a

(4.15)

The stator current from the equivalent circuit (Fig.4.9.c):

E f V E f V I = = Zs Zs Zs a

= =

Ef Zs s Ef Zs

V 0 Zs s V s Zs

(4.16)

( s )

42

2 V Ef V ( s ) s S = m Zs Zs

(4.17)

The real power P and reactive power Q are:

2 V Ef V P = m cos( s ) cos s Zs Zs 2 V Ef V Q = m sin( s ) sin s Zs Zs

(4.18)

(4.19)

In large synchronous machines Ra << Xs, thus Zs = Xs and s = 90o. From the above equations:

P=m

V Ef Zs

sin( )

(4.20)

2 V Ef V Q = m cos( ) Zs Z s

(4.21)

Equation 4.11 can be directly derived from the phasor diagram drawn for Ra = 0 (see Fig. 4.10)

Fig. 4.10 Phasor diagram drawn for Ra = 0.

43

In general, the active power

P = mVI a cos
and reactive power

(4.22)

Q = mVI a sin
From the phasor diagram (Fig. 4.10),
X s I a cos = E f sin , and

(4.23)

(4.24)

I a cos =

Ef Xs

sin

(4.25)

Combining Equations 4.16 and 4.13, we get:

P=m VE f Xs sin

(4.26)

Because the stator losses are neglected in this analysis, the power developed at the terminals is also the air gap power. The electromagnetic torque developed by the machine is:

T=
Inserting Equation 4.18 into 4.19:
T=

(4.27)

m VE f sin s Xs

(4.28)

where m is the number of phases. The torque-power angle characteristic drawn for the motor analyzed in this project is shown in Fig. 4.11. The torque was calculated for the following data: m = 3, Vrms=16.9V,
Ef = KE 2

s = 19.46V, s =

2f = 377 rad/s, Xs = 2fLs = 0.396. The calculations p/2

44

were done in MATLAB using m-file torque_delta (see Appendix B). The maximum torque, known also as the pull-out torque is at = 90o. The machine will lose synchronism if > 90o. The pull-out torque can be increased by increasing the excitation flux f.

8 6 4 Torque [N.m] 2 0 -pi -2 -4 -6 -8 -4 -3 -2 -1 0 delta [rad] 1 2 3 4 -pi/2 pi/2 pi

Fig. 4.11 Torque-power angle characteristic.

4.3.3 Performance Characteristics of the Motor in Steady-State

To determine the electromechanical characteristics of the synchronous motor, calculations were carried out by means of MATLAB (see m-file synchronousplot in Appendix B) using the equations in sections 4.3.1 and 4.3.2. The parameters of the equivalent circuit for a supply frequency of 300HZ are shown in Table 4.2.

45

Table 4.2 Parameters of the equivalent circuit.

supply voltage V emf Ef synchronous reactance Xs armature resistance Ra 24V AC 27V AC 0.396 0.077

The performance characteristics are presented in Fig. 4.12. The efficiency of the motor was calculated from the equation:

Eff =

Pout 100% Pin

(4.29)

where Pout= P-Pm, mechanical power loss Pm = B.s and Pin = Re(3VI*). To verify the dynamic and steady-state models to some extent, the simulation and calculation results obtained from both models respectively, are plotted on the same graph with the results from the dynamic model represented in form of points (Fig. 4.12). Looking at the results obtained, it is observed that both models are pretty consistent with each other.

4.4 Experimental Determination of the Synchronous Motors Performance

In addition to the dynamic and steady-state performance of the synchronous motor, it was tested to determine its performance experimentally. The experiment was performed using the program mentioned in Chapter 1, dSPACE controller desk in conjunction with MATLAB. A drives board (Fig. 4.13), on which the 42V DC supply was connected, supplied the motor. This board was designed to enable performance of a variety of experiments on AC/DC machines. The main features of the board are listed.

46

1000 efficiency/10 [%] 800 speed*5 [r.p.m.] 600

400 current/50 [A] 200 output power [W]

0.1

0.2

0.3 0.4 0.5 load torque [Nm]

0.6

0.7

0.8

Fig. 4.12.a Characteristics of steady-state and dynamic models for the synchronous motor.

1 0 -1 -2 -3 -4 -5 -6 -7 -8 0 0.1 0.2 0.3 0.4 0.5 0.6 load torque [Nm] 0.7 0.8 0.9 delta [degrees] cos(fi)

Fig. 4.12.b Angle delta () and power factor cos ( fi) obtained from the steady-state (continuous lines) and dynamic models.

47

Two completely independent 3-phase PWM inverters for complete simultaneous control of two machines.

42 V dc-bus voltage to reduce electrical hazards. Digital PWM input channels for real-time digital control. Complete digital/analog interface with dSPACE control box (Fig. 4.14).

42 V DC bus voltage

interface to dSPACE box

3-phase inverters

Fig. 4.13 Electric drives board.

This board was connected to the dSPACE control box (Fig. 4.14). The main features on this control box include: 8 analog to digital channels (ADCH) and vise versa. Digital output/input (I/O) interface.
48

Slave I/O pulse width modulation (PWM) interface with the drives board. Two speed encoder interfaces.

In order to record the speed of the motor, an encoder attached to the motor was connected to the control box (see Fig. 4.15). Fig. 4.15 shows the overall layout and connection arrangement of the computer, oscilloscope, drives board and control box. The experiment began with building a block diagram in MATLAB/SIMULINK as seen in Fig. 4.16. This was then built in real time. In so doing, controller desk (which has the ability to run models in real time) was loaded with the model and all its parameters. A layout as shown in Fig. 4.17 was then constructed, and the experiment ran. Changes in time were observed. Using the block diagram (Fig. 4.16), voltage was controlled using the v/f = constant method. The diagram was broken up into three different groups. Group 1 represented the actual motor. Group 2 controlled the frequency and duty-cycles shown in the layout (Fig. 4.17). At a constant frequency of, say, 1HZ in the block diagram, the frequency slider gain could still be varied in controllerdesk causing an effect on the duty cycles, current and speed. The duty-cycles were sinusoidally generated and phase-shifted by 120 .The current in phases a and b was also sinusoidal in steady-state. Group 3 controlled the speed m and current in the layout. As mentioned earlier, the parameters in the block diagram were assigned to their corresponding blocks in the layout. Also included in Fig. 4.17 was the block V-boost. This is the boost voltage which was given a very small value of 0.8V in order to compensate for resistance drops at low speeds and improve the start-up process of the motor. Other blocks in the layout that affected the results of the experiment were the acceleration and deceleration blocks. These represented the rate limiter block in group 2

49

of the block diagram which was inserted in the reference frequency path, to set the acceleration and deceleration times. Limitation of these two transients was necessary to limit the stator current during start-up and breaking of the motor. Efforts to get the motor to reach and operate at its rated frequency of 300Hz were futile; it was only able to get to 120Hz. In order for the motor to have been able to operate up to and upwards of its rated frequency, it should have started with the help of a control circuit which would have enabled it to keep the power angle = 90.However, building the control circuit for the synchronous motor was beyond the scope of this project. The experiment was, therefore, performed at a frequency of 120Hz.

speed encoder interface

ADC/DAC inputs

interface to drives board

Fig. 4.14 dSPACE control box.

50

Fig. 4.15 Experimental setup showing overall layout of all equipment used.
In order to see how consistent the experimental results were, the same voltage V= 13.7 V (rms) and frequency f = 120 Hz was supplied to the synchronous motor model in MATLAB/SIMULINK and the model simulated at the rated load torque of 0.56 N.m. The results are shown in table 4.3 below.

Table 4.3 Experimental and simulation results of the synchronous motor.

parameters output power, Pout [W] phase current, I [A] speed, m [rad/s] EMF, Ef [V] experimental results 84.44 13.5 150.79 11.2 simulation results 84.45 13.75 150.8 11

51

Fig. 4.16 Simulink model of experiment.

Fig. 4.17 Layout of the experiment.

52

(a)

(b)
20

10 current [A]

-10

-20

0.02

0.04 0.06 time [s]

0.08

0.1

Fig. 4.18 Current waveforms from: (a) experiment, (b) simulation.

53

CHAPTER 5: CONCLUSIONS
The performance of a 3-phase permanent magnet motor operating as a synchronous and BLDC motor was analyzed in this thesis. The motor performed quite differently under these two different schemes. To examine these differences, the software package MATLAB/SIMULINK was used to design the block diagrams and run the simulations. Models in both steady-state and dynamic conditions were taken into consideration. In addition, an experiment to see how the synchronous motor behaved in real-time was performed; the results were then compared with those gotten in MATLAB/SIMULINK for the same data. The BLDC motor, which operated with a self-controlled frequency, showed near consistencies when both of its models (i.e. in steady-state and dynamic conditions) were compared. The small discrepancies were attributed to some factors as explained in Chapter 3. As for the synchronous motor, both its steady-state and dynamic conditions performance, also matched each other. However, the experimental results somewhat differed from the theoretical ones. These discrepancies were also attributed to some factors as explained in Chapter 4. Differences between the BLDC and synchronous motor (in their operation), were noted as follows. The BLDC motor operates at a self-controlled frequency, meaning the motor itself imposes an appropriate value required for its actual speed. The motor should therefore be equipped with a speed or position sensor. Conversely, the

54

synchronous motor operates at set reference frequency and therefore has a constant speed. Looking at the characteristics, the efficiency of the BLDC motor first increases with a steep rise and then gradually decreases as the load torque is increased from no load. For the same condition however, the synchronous motors efficiency increases gradually. The equations that govern both motor models explain this difference. For instance, the BLDC motors efficiency depends on mechanical power and input power; at no load, the output power is zero (since it depends on load torque) leading to zero efficiency. As torque increases, both input and output power increase which leads to decreasing efficiency. As mentioned before, the synchronous motor speed stays constant regardless of the conditions. However, particularly looking at the speed-torque characteristics, the BLDC motor speed decreases as the load torque is increased. In building the SIMULINK models, the BLDC model had to be equipped with averaging blocks for the speed, power and torque; otherwise definite values could not have been read when the mentioned quantities reached steady-state (as the waveforms were unsettled). On the other hand, the synchronous motor needed no such block. When the power, torque and speed reached steady-state, the sinusoidal waveforms displayed definite values. Although the two models operate quite differently, a few similarities were also noted as follows; It is observed from the current-torque characteristics that both models have increasing current as the torque increases.

55

Similarly, observed from the power-torque characteristics, the mechanical power increases with load torque for both models.

Conclusions drawn from this project are as follows. The electromechanical characteristics of the BLDC and synchronous motor models show that regarding the average values, the motor behavior in steadystate is similar to that in dynamic conditions for each particular model. That means that the motors can be analyzed in this state, using the simpler steadystate model. Further still, from the efficiency-torque characteristics, it can be concluded that in order to get increasing efficiency out of the machine (as the torque increases), the synchronous motor should be chosen. This is because, even though the BLDC motor has increasing efficiency as the simulation begins, it gradually decreases with increasing the load torque (as mentioned earlier). Regarding the synchronous motor operation, analysis of both the dq and ABC models shows that either one would yield the desired results and ones choice of which model to use would just depend on the application. In particular, the dq model should be used when the motor works with a control circuit. When operating in an open system, the ABC system can be used for analysis. Based on the synchronous motors behavior of performing at a constant speed and frequency, this motor model should then be chosen if ones application requires maintaining a constant speed on their motor.

56

From the comparison of the theoretical and experimental results of the synchronous motor, the study done shows that for the motor to perform optimally in real time, it must operate at its rated frequency.

Future dSPACE experiments should focus on trying to get the synchronous motor to operate at its rated frequency within the closed loop control system. Further more, studies should be done on the experimental performance of a BLDC motor while its being supplied from the current inverter.

57

REFERENCES
[1] Dr. Ernest Mendrela, Introduction to Brushless DC Motors Lecture Notes, Louisiana State University, Baton Rouge, LA 2001. [2] http://en.wikipedia.org/wiki/Brushless_DC_motor. [3] Dr. Ernest Mendrela, Brushless DC Motors in Transients, Lecture Notes, Louisiana State University, Baton Rouge, LA 2001. [4] Ned Mohan, Advanced Electric Drives Minneapolis, MN 2001. [5] Dr. Ernest Mendrela, Synchronous Machines, Lecture Material, Louisiana State University, Baton Rouge, LA 2001. [6] J. F. Gieras, I. A. Gieras, Electrical Energy Utilisation Poland, Torun 1998. [7] M. A. El-Sharkawi, Fundamentals of Electric Drives Brooks/Cole 2000. [8] J.F. Gieras, M. Wing, Permanent Magnet Motor Technology Marcel Dekker 2002. [9] Deepti Chikkam, Performance of the Disc Brushless DC Motor Applied as a Gearless Drive for the Wheel Chair, Thesis Document, Louisiana State University, Baton Rouge, LA 2005. [10] P.C. Sen, Principles of Electric Machines and Power Electronics John Wiley 1989.

58

APPENDIX A: M-FILES FOR BLDC MOTOR OPERATION (a) M-file, bldcfilesloader1 for plotting waveforms in Fig. 3.6.
load current I; load speed w; load torque Tem; load LoadTorque TL; load voltage Vph; load SourceCurrent Is; I=I'; w=w'; Tem=Tem'; TL=TL'; Vph=Vph'; t=I(:,1);I=I(:,2);w=w(:,2); Tem=Tem(:,2);TL=TL(:,2);Vph=Vph(:,2);Is=Is(:,2); CLF figure (1), subplot (2, 1, 1), plot (t, w), xlabel ('time [s]'), ylabel ('speed [rad/s]'), subplot (2, 1, 2), plot (t, I, t, Vph), xlabel ('time [s]'), gtext ('phase A current [A]'), gtext (' phase A voltage [V]'), figure (2), subplot (2, 1, 1), plot (t, Tem, t, TL), xlabel ('time [s]'),gtext ('electromagnetic torque [N.m]'), gtext (' load torque [N.m]'), subplot(2,1,2), plot (t,w), xlabel ('time [s]'),ylabel ('source current [A]'),

(b) M-file, bldcplotter1 for plotting characteristics in Fig. 3.14.

TL = [0 0.16 0.26 0.36 0.46 0.56]; I = [2.59 2.658 2.870 3.255 3.810 4.427];

59

w = [342.1 335 330 325 320.3 315.9]; %Pou = [53.51 85.7 116.8 147.3 176.9] %Pin = [69.5 104 138.5 173.2 207.85] %eff = [77 82.4 84.3 85.0 85.1] figure(1), plot(TL,w,'ro',TL,I*25,'b*'), xlabel ('torque [Nm]'),gtext ('speed [rad/s]'),gtext ('current/25 [A]')

(c) M-file, brush1 for plotting characteristics in Fig. 3.15.

%simulation results-steady state model Is=0:1:12; Kms=0.23; Us=42;Rs=0.154; Ps=Us*Is; Ts=Is.*Kms-0.0001; omgas=(Us-Rs*Is)./Kms; ns=omgas*30/pi; %Pms=(Ts-ns*.00010).*omgas%0.0001 is the friction coeff Pms=Ts.*omgas; Effs=Pms./Ps*100; % simulation results from dynamic model Td=[0 0.285 .53 .775 1.025 1.78 2.025 2.5]; %Effd=[81 82 85 84 82.5 75 77 74]; Effd=[99.59 98.88 98.15 97.42 96.69 96.32 95.96 95.59];

60

Pmd=[0 10.23 93.46 215.99 296.14 375.05 414.04 452.72]; %Pmd=[0 31 72 114 155 195 236 276]; Id=Td/0.23; omgad=[182.6 181.27 179.93 179.26 177.92 177.25 176.58 175.28]; nd=omgad*30/pi; figure(1), plot(Ts,Effs*25,'r',Ts,100*Is,'r',Ts,ns,'r',Ts,Pms,'r',Td,Effd*25,'xr',Td,100*Id,'ro',Td,nd,'r x',Td,Pmd,'r*'),grid, xlabel('torque [Nm]'),gtext('efficiency/25 [%]'),gtext('speed [r.p.m.]'), gtext('mechanical power [W]'),gtext('current/100 [A]')

61

APPENDIX B: M-FILES FOR SYNCHRONOUS MOTOR OPERATION (a) M-file, synchro for plotting waveforms in Fig. 4.4.
% file name: synchro.m load angle fi;load torque T; load speed w; load loadsp TLw;load voltage VE;load Pout Pou;load input Pin;load Pelectrom Pem; fi=fi';T=T';w=w';TLw=TLw';VE=VE';Pou=Pou';Pin=Pin';Pem=Pem'; t=fi(:,1);fi=fi(:,2);w=w(:,2);Pem=Pem(:,2);Pin=Pin(:,2);Pou=Pou(:,2); figure(1), subplot(2,1,1), plot(t,fi,t,0*fi,'k'),xlabel('time [s]'), ylabel('rotational angle [rad]'),grid subplot(2,1,2),plot(t,w,t,0*w,'k'),xlabel('time [s]'), ylabel('speed [rad/sec]'),grid figure(2) subplot(2,1,1),plot(t,T,t,0*T,'k'),xlabel('time [s]'), ylabel('torque [Nm]'),grid subplot(2,1,2),plot(t,TLw,'k'),xlabel('time [s]'), ylabel('load torque, speed [Nm], [rad/sec]'),grid figure(3) plot(t,VE),xlabel('time [s]'), ylabel('voltage [V], Induced voltage Ef [V] '),grid figure(4) plot(t,Pin,t,Pou,t,Pem),xlabel('time [s]'),

62

ylabel('Pin [W], Pem [W], Pout [W]'),grid

(b) M-file, phasor for plotting waveform in Fig. 4.6.

load phasevoltage Ef; load sourcevoltage V1;load phasecurrent iA; V1=V1';Ef=Ef';iA=iA'; t=V1(:,1);V1=V1(:,2);Ef=Ef(:,2);iA=iA(:,2); figure(1), plot(t,V1,'r',t,Ef,'b',t,iA,'g'), xlabel('time [s]'), gtext('E_f [V]'),gtext('i_A [A]'),gtext('V_A [V]'),gtext('delta'),gtext('theta'),grid

(c) M-file, torque_delta for plotting characteristics in Fig. 4.11.

Xs=0.396; V=16.97; Ef=19.46;w=377; delta=-pi:pi/100:pi; T=(3*V*Ef*sin(delta))/(w*Xs); figure(1), plot(delta,T),grid xlabel('delta [rad]'), ylabel('Torque[N.m]'),

(d) M-file, synchronousplot for plotting characteristics in Fig. 4.12.a.

%simulation results-steady state model Xs=0.396; V=16.97; Ef=19.46; B=0.0001;Ra=0.077; w=377;

63

Ts=0.0001:0.76/1000:0.76; Pms=Ts*w+w*w*B; Pout=Ts*w; delta=asin(Pms*sqrt(Xs*Xs+Ra*Ra)/(3*V*Ef)); delta1=(180/pi)*delta; Pin=Pms; Is=(V-(Ef*cos(delta)+j*Ef*sin(delta)))/Xs; ns=w*30/pi; Effs=(Pout./Pin)*100;

% simulation results from dynamic model Td=[0 0.08 0.16 0.26 0.36 0.46 0.56 0.66 0.76]; Effd=[0 68 80 88 90 92 94 95 96 ]; Pmd=[0 30.2 60.4 98 136 174 211 248 288]; Id=[6.33 6.5 6.56 6.67 6.83 7 7.16 8 8.3]; omgad=377; nd=omgad*30/pi;

figure(1), plot(Ts,Effs*10,'r',Ts,50*abs(Is),'r',Ts,ns/5,'r',Ts,Pout,'r',Td,Effd*10,'xr',Td,50*Id,'ro',Td, nd/5,'rx',Td,Pmd,'r*'),grid, xlabel('load torque [Nm]'),gtext('efficiency/10 [%]'),gtext('speed*5 [r.p.m.]'), gtext('output power [W]'),gtext('current/50 [A]')

64

(e) M-file, fi_plot for plotting characteristics in Fig. 4.12.b.

Xs=0.396; V=16.97; Ef=19.46; B=0.0001;Ra=0.077; w=377; Ts=0.0001:0.96/10:0.96; Pms=Ts*w+w*w*B; delta=asin(Pms*sqrt(Xs*Xs+Ra*Ra)/(3*V*Ef)); deltas=-(180/pi)*delta; Is=(V-(Ef*cos(delta)+j*Ef*sin(delta)))/Xs; fi1s=-cos(angle(Is)); deltad=[-0.3 -1.8 -2.6 -3.6 -4.4 -5.4 -6.2 -7 -7.9 ]; fi=-[-9.27 -15 -18.97 -22.5 -26.02 -30 -33.53 -37.5 -41.03]; fi1=[0.99 0.97 0.95 0.92 0.90 0.87 0.83 0.79 0.75]; TL=[0 0.16 0.26 0.36 0.46 0.56 0.66 0.76 0.86]; figure(1), plot(TL,deltad,'xr',TL,fi1,'ro',Ts,deltas,'r',Ts,fi1s,'r'),grid xlabel('load torque [Nm]'),gtext('delta [degrees]'),gtext('cos(fi) [degrees]'),

65

VITA
Sophie Sekalala was born in Kampala, Uganda. She first came to the U.S in August 1999, via New York, where she started her undergraduate studies. She received her Bachelor of Electrical Engineering degree in May 2003 from the City University of New York. After graduation, she interned with Brookhaven National Laboratory in Long Island, New York, for six months, before starting on her masters degree at Louisiana State University in January 2004. She is currently a candidate for the degree of Master of Science in Electrical Engineering at Louisiana State University, which will be conferred in August 2006.

66