CHAPTER-3 MEASUREMENT OF COMMON MODE VOLTAGE IN 2- LEVEL INVERTER FED INDUCTION MOTOR DRIVE

Transcription

1 46 CHAPTER-3 MEASUREMENT OF COMMON MODE VOLTAGE IN 2- LEVEL INVERTER FED INDUCTION MOTOR DRIVE 3.1. INTRODUCTION Induction Motor (IM) is considered as a constant speed motor with certain limitations. Earlier days, speed control of IM was not as easy as it demands simultaneous variation in voltage and frequency but by the introduction of microprocessor this has become easier. CM voltage is a natural result of PWM techniques used in adjustable speed control of IM [7, 33]. It was found that the CM voltage in the drive system was responsible for disturbance to the measuring circuits and communication circuits connected to the same power line. Fig 3.1 shows the block schematic diagram of adjustable speed drive (ASD) [7, 33]. Fig CM voltage as EMI source CM voltage can be defined as follows [7, 8]. Vag = Van+Vng (3.1) Vbg = Vbn + Vng (3.2) Vcg = Vcn+ Vng (3.3) Sum of all phase to neutral voltage is zero

2 47 Vng = [Vag +Vbg+ Vcg] (3.4) CM current Ilg = C Where C is the total capacitance in the system [7, 8]. High frequency bearing currents are due to the flow of current in the CM circuit from the shaft of the IM to ground. This chapter proposes the experimental work on the 2-level inverter and measurement of the CM voltage at the star point of stator winding of a modified squirrel cage IM to the general ground. 2-level inverter is fabricated by using power MOSFETs (2SK962) and the Undeland snubbers are used along with fast recovery diodes. SVM scheme (as discussed in chapter-2) is used for the generation of gating signals and the inverter is designed to operate at 40Hz frequency. The simulation is done by using the MATLAB/SIMULINK software and the results are presented. By conducting the experiment the DSO recorded Voltages/current and the processed FFT results are presented LITERATURE SURVEY The most common method of speed control of IM utilizes the PWM [7, 22 & 60]. Though the speed control of IM gained popularity by the advent of modern power electronics associated with high power devices like power MOSFET /IGBT there is a lot of concern about the CM voltage and the EMI is due to the high switching transients of the power devices, which are harmful to the motor bearings, measuring and controlling systems/instruments connected to the same mains. There are many

3 48 undesirable effects of the use of solid-state power converters/inverters, due to its non-linear nature. Based on the above reasoning it is very much essential to minimize the CM voltage/current, there by the EMI interference, should be reduced to the acceptable limits by the use of advanced electronics instrumentation and control. The available techniques of CM voltage reduction are 1) An active circuit for cancellation of CM voltage generated by PWM, 2) A dual bridge inverter approach to eliminate CM voltages, 3) The mitigation techniques of CM voltages in PWM inverter, 4) The modulation technique used is characterized by a low switching frequency which also helps to reduce the CM noises. In this work SVM modulation scheme is used for MLIs to study and measure the CM voltage Generation, control and regulation of CM voltage and EMI from IM drives Adjustable speed IM drive (ASD) manufacturers now switched over to IGBTs from BJT due to its merit [32, 33]. The merits are, the rise/fall time of the switching capability is 5-10 times faster in IGBT, thereby lowering of the device switching losses and also requires only low current drive circuits. But the disadvantages are higher dv/dt transitions due to the high switching frequencies and faster output. This increases the EMI problems. There should be regulation standards on allowable conducted emissions for the successful drive system installations. The CM noise is an electrical noise induced on signals with respect to general ground. CM noise problems imply a source of noise, a means of

4 49 coupling noise by conduction or radiation from circuits/equipments, susceptible to the magnitude, frequency and repetition rate of the noise impressed. Fig Potential CM noise problems Fig.3.2. shows [7, 32&33] potential CM noise problems increase with susceptible equipment present, system input voltage, system drive quantity and length of motor cables. There are other factors such as ground system and cabinet layout. Higher drive carrier frequency (fc) increases the number of switch transitions and sum of the CM noise current. Generally if the length of the motor cable from the inverter is less than 20 feet [7, 33], which will exhibit low line to ground capacitance and hence low CM noise risk from capacitive dv/dt ground currents. As the cable length is long, the high frequency oscillations of reflected wave voltage transients appear on the motor terminals, which will weaken the insulation of the stator winding and cable capacitance [33]. Hence the EMI reduction must involve the

5 50 safety, equipment grounding, signal grounding and the effect of grounding system type depending on CM noise IM drive as an EMI noise generator The PWM inverter output voltage has abrupt voltage transitions to and from the DC bus controlled by power MOSFET/IGBT [7, 33]. Switching time and power are the main sources of conducted and radiated noise. Generally IGBT rise times are of the order of 0.05 µs to 0.2 µs, while BJTs are of the order of 1 µs to 2 µs, corresponding to fn of MHz [7, 32] and KHz, respectively, output dv/dt is now 20 to 40 times higher. Most drive related EMI [9] is due to conducted noise currents hence it is obvious that the CM noise current increases with low trise and higher bus voltages. Due to high carrier frequency fc increases the EMI, since the CM power repetition rate is faster. The CM current flows in the ground wire and hence it produces the CM voltages. The issue of conducted CM current inducing CM voltage in system noise coupling paths and some solutions to control EMI are well discussed in literature [33, 63]. There are four basic important steps for the mitigation of EMI [8, 9] and they are as follows. (1) Proper grounding (2) attenuate the noise source (3) shield noise away from sensitive equipment/instrument and (4) capture and return noise to the source (drive). All the above four important steps are well explained in the reference [21, 32] and are self-explanatory.

6 EMI in switching power converters The growth of power electronic equipments/systems in the recent years is due to the advancement of power converters [7, 9]. In these the switched mode power supplies, adjustable speed AC drives, utility interface and the other power electronic equipments in the industries are becoming increasingly common. These converter s switching frequencies are high due to advancement of power semiconductor devices. Due to fast switching frequency of the power devices there exists the CM voltage at the output of the converter [9]. This leads to the EMI and EMC problem Nature of conducted emissions The conducted emissions of a power electronic circuit measured across the line impedance stabilization network (LISN) can be broadly split up into CM and differential mode (DM) emissions [22, 64]. This division is distinctly due to different causes and noise current paths Causes of common mode emissions The normal operation of switching power converters requires the use of switching states that tend to generate the potential to the load [7, 21]. Very often the rate at which this oscillations caused is equal to the switching frequency of the converter. It is well known that PWM operation at high frequencies produces voltages with spectral contents extending to MHz. In the presence of parasitic coupling (capacitance) to ground, this spectral energy causes CM currents to flow between the

7 52 power lines and the ground. This, in a fundamental sense, is the mechanism by which CM EMI is generated [7, 21]. The magnitude of this current depends upon the value of the parasitic coupling capacitance of the motor to ground. Further, the neutral potential with respect to ground provides the CM excitation. This potential is by definition as the CM potential (CM voltage) applied to the load by the inverter. In the conventional PWM operation this potential oscillates with amplitude of one sixth of the total dc bus voltage of the inverter on either side of ground, and is never zero. As mentioned before it has substantial high frequency spectral energy and therefore the CM current resulting from this spectral energy is far from negligible. It is to be noted that if the motor is excited by symmetrical 3-phase sinusoidal source, the CM voltage applied to the load would be ideally zero, and no emissions would result. It follows from this reasoning that if the converter could be operated such that the neutral potential would be ideally zero at all times, then the CM emissions could be seriously attenuated. This method is popular for the active noise cancellation [7]. Although the discussion of CM emissions presented has focused on the 3-phase drive systems, it is applicable to almost all power electronic converters. In the practical implementations, the parasitic capacitance to ground can never be eliminated. In addition to the parasitic coupling of the load to the ground, the parasitic coupling between any part of the system that has a high dv/dt and the ground causes CM emissions. The

8 53 mechanism of noise current generation is again very much same as that discussed above. Fig.3.1 shows the schematic block diagram of a modern power electronic drive which consists of secondary of the 3-phase transformer, LISN, converter, PWM inverter and 3-phase IM. Vng is the voltage between neutral to the general ground which is also called as star point of stator winding of an IM to ground and also the direction of Ilg is as shown in Fig.3.1[7] SPACE VECTOR MODULATION SCHEME Reference [54] Compared SPWM & SVPWM and concluded that the later method yields 15% higher inverter output voltage. SVPWM is based on space vector theory which has its origin in the generalized theory of electrical machines. For a given minimum line side voltage SVPWM control strategy requires less DC bus voltage, consequently the voltage stress on the power devices is less, results in less EMI effects, further SVPWM results in reduced harmonic distortion in the line currents. The proposed scheme is said to have the merit over other PWM schemes in terms of efficiency and can be easily implemented. In order to obtain optimum performance and the minimum switching frequency for each of the power devices, the switching state sequence is arranged such that the transition from one state to the next is performed by switching only one inverter leg. This condition is met if the sequence begins with one zero state and the inverter poles are toggled until the other null state is reached. To complete the cycle, the sequence is reversed ending with the

9 54 first zero state. The pulse pattern arrangement of (right aligned sequence) is shown in Fig.3.3. [12]. SVM scheme, procedure and principles are already discussed in chapter-2. Fig.3.3. Signal switching sequence pattern 3.4. SIMULATION MODEL OF3-PHASE 2-LEVEL INVERTER Simulation is carried out using MATLAB/ SIMULINK software. Fig.3.4 Shows SIMULINK model for 2-level inverter fed IM drive using SPWM technique. PWM signals are generated using a high frequency triangular wave, called the carrier wave, is compared to a sinusoidal signal representing the desired output, called the reference wave. Whenever the carrier wave is less than the reference, a comparator produces a high output signal, which turns the upper switch in one leg of the inverter ON the lower switch OFF. In the other case the comparator sets the firing signal low, which turns the lower switch ON and upper switch OFF.

10 55 SIMULINK model also includes CM equivalent circuit with bearing model for measurement of shaft voltage and bearing current. Method used in modeling of IM and CM equivalent circuit parameters is as in the reference [30] TWO LEVEL SPWM INVERTER Line voltage Measurement Va A Vb powergui B Vc Van Phase Voltage Discrete, Ts = 2e-006 s. C Vbn Vcn 3 -PHASE INDUCTION MOTOR A a Source Impedence Conn1 Common mode voltage B b Conn2 Bearing current Conn3 PWM Geneation for the C c AC - DC - AC A AC - DC -AC i + - B Csr C N Rb Shaft voltage Csf Cg + - v + - v Cb Zb Common mode equivalent circuit with bearing model Fig.3.4.Simulink model of 2-Level inverter fed IM using SPWM. Fig.3.5 shows the SIMULINK model for 2 level inverter fed IM drive using SVPWM technique. Gating signals are generated using co-ordinate transformation a-b-c to d-q. Switching time duration for each sector and switching time of each switch is determined and applied to the switching devices.

11 S S S g D g D g D S S S g D g D g D 56 2-Level Inverter Fed Induction motor (Space vector PWM ) Discrete, Ts = 2e-006 s. powergui ramp generator L1a [s1] ramp ramp L1b [s2] 3phases sin generator trig Trig L2a [s3] 50 Freq_com L2b [s4] Vcom dir Vabc Vabc Angle Angle Sector ab vector sector Sector Gate timing gate_timing L3a [s5] L3b 400 Vbus FVbus Vbus ab_vbus a_b_vbus gates logic [s6] Scope low pass bus filter ab transform switching time calculator Line voltage1 Measurement Scope1 Va A Vb B Vc Van Neutral Voltages Bearing current [s1] [s3] [s5] C Vbn From From2 From4 Vcn Mosfet4 Mosfet2 Mosfet1 3 -PHASE INDUCTION MOTOR LOAD i + - Source Impedence Conn1 A B1 Csr B Conn2 Rb C B2 Subsystem [s2] From1 [s4] From3 [s6] Conn3 Bearing voltage Mosfet5 Mosfet3 From5 Mosfet Csf Csf Cg + - v Cb Zb Common mode voltage1 + - v Fig.3.5. Simulink model of 2-Level inverter fed IM using SVPWM HARDWARE CIRCUIT DESCRIPTION Phase-half bridge converter: Fig. 3.6 shows the 3-phase half controlled bridge converter. Three numbers of power diode and three numbers of SCRs are used for forming the bridge of the converter. Fig.3.6.The 3-phase half controlled bridge converter

12 57 3-numbers of line inductance (each of 0.8mH. not shown in the circuit) which are connected in series with the supply line to the converter so as to protect the converter bridge from di/dt effect and the snubbers used are the resistor and capacitor (100 Ω, 0.1µF not shown in the circuit.) across each SCR so as to protect it from dv/dt effect. The 3- phase AC supply is connected to the system by using an isolation transformer [7]. The ON-OFF switch (not shown in the circuit) is interlocked with the DC power supplies used for the converter and inverter. A filter capacitor of 80µF (not shown in Fig. 3.6) is used after the converter. Before connecting the converter output to the inverter circuit, a series resistance of 10 Ω, 100W is connected in series so as to protect the converter from dead short circuit in case of failure of the inverter devices Inverter circuit power module: In Fig.3.7 the insulated gate bipolar transistor (IGBT) or a power MOSFET can be used as a switching device for the inverter. The MOS devices are very sensitive to overvoltages; hence some safety measures have to be taken during handling the power MOSFET. (1) Protection against the static charging during handling of the components. (2) Protection against the over voltage in a circuit which can be remedied by connecting a zener diode as closely as possible to MOSFET module control circuit. (3) Fast recovery diode can be connected across each of the power MOSFET with opposite polarity to bypass the slow internal diodes for fast switching. (Power MOSFET will

13 58 have recovery diode built-in the structure). For the inverter circuit the power MOSFET (2SK962) is used. The gate drive circuit block diagram was discussed in the previous chapter (Fig.2.10) and the µ-controller program is given in Appendix-4. Fig.3.7. Power circuit diagram of 2- level inverter fed IM with undeland snubber circuit.(t1-t6 represents the power MOSFET) Snubber circuit: The snubber circuit in Fig. 3.8 [35] is provided across each leg of the inverter have the following advantages. It reduces the stresses on the switching device during the switching transients. As this reduces the power loss in the switching device the switch ratings can be better utilized by using the snubber. A number of configurations described by Undeland have been used to build the inverter as this has the following advantages over the other snubbers. It has fewer components and the snubber diodes do not cause difficulties

14 59 due to their reverse recovery and all the losses are dissipated in one resistor. Fig.3.8. The Undeland snubber circuit Design of LS, CS, C0 and RS: The inductor LS is used to limit the rate of rise of current, when, either of the switches is turned ON. The value of LS is decided by limiting the current through the switch at the end of the rise time (tr) to 10% of the full load current. This value is sufficient to consider the switching process of one device, as the switching process of other device is almost similar. Voltage across the device during rise time is V = {1- t/tr} X Vdc (3.5) LS = {Vdc X t} / tr = 10 X 0.1 (Vdc +t) dt. (3.6) LS = {1 / (10 X0.1)} X (600X2X10-6 )/2 = 600 X 10-6 H. (3.7) This value of the inductance also decides the current build up in the device after a short circuit and before the short circuit protection of all the devices. An inductor of 500 µh has been used for this purpose.

15 60 Design of CS and C0: The value of CS is decided such that the capacitor will be exactly recharged to Vdc at the end of current fall time during the switching OFF of the top and bottom device of the same leg. Cs = (IL/2Vdc) X tr = (10/2X600) X 2X10-6 = µf (3.8) CS is chosen as 0.05 µf as higher value will only further reduces the stress in the switch. The value of C0 is recommended to be about 10 times the value of CS and it is taken as 0.5 µf. Design of RS: The over voltage across the switching devices caused by the snubber energy dissipation is analyzed by equivalent circuit shown in Fig The value of C at turn ON of top device and turn OFF bottom device of the same leg is (C0 + CS) and it is C0 for the other switching. The initial inductor current I0 is Imax at turn ON and IL at turn OFF. The circuit may be over damped or under damped depending on the resistance value. The value of resistance for critical damping is Rcrit = {0.5 (Ls) ½ } / C ~ 15.8 Ω. (3.9) The value of RS is chosen as 2.5 Ω to keep the circuit over damped. Power circuit: The 3-phase inverter shown in Fig.3.7 is used in the power circuit driving the IM in adjustable speed operation. Power MOSFET switches are used with fast recovery diodes connected across each device so as to handle the inductive currents GATE DRIVE CIRCUIT IMPLEMENTATION This section presents the hardware implementation of the gate drive circuit. There are several steps involved in implementing the control

16 61 circuit discussion and the block diagram are shown in Chapter-2 (Fig.2.10) RESULTS Simulation results of 3-phase 2-level inverter (SPWM) 500 Line voltage Voltage(V) Time(s) Voltage(V) (a) Phase voltage Time(s) 300 (b) Common mode volatge Voltage(V) Time(s) (c) Fig.3.9.(a) Line voltage,(b) Phase voltage, (c) CM voltage

17 Simulation results of 3-phase 2-level inverter (SVPWM) Voltage(V) Line volatge Time(s) (a) 400 Phase voltage Voltage (V) Time(s) (b) Common mode voltage Voltage(V) Time(s) (c) Fig.3.10.(a) Line voltage,(b) Phase voltage, (c)cm voltage

18 Experimental results of 3-phase 2-level inverter (SVM) Fig Input to Motor (20ms/Div). Ch 1: 200 : 1 Phase voltage to IM. Ch 2: 200 : 1 Line voltage to IM. Ch 3: 20 : 1 wave form of vector sum of current in terms of voltage. Ch 4:1: 1 One phase current in terms of voltage. Fig.3.12 Input to Motor (20ms/Div). Ch 1: 200 : 1 CM voltage. Ch 2: 200 : 1 Line voltage to IM. Ch 3: 20 : 1 wave form of vector sum of Ph current in terms of voltage. Ch 4: 1 : 1 Phase current in terms of voltage

19 64 Fig.3.13.DSO recorded waveforms.ch.1.200:1 CM voltage. Ch :1Line voltage.ch.3.1:1.sum of phase current. Ch.4. 1:1.one phase current. The simulation is done using MATLAB/SIMULINK software for 2-level inverter fed IM and the simulated results are as shown in Figs.3.9 & The experimental CM voltage, line voltage, vector sum of phase currents in terms of voltage using current probe and the phase current in terms of voltage are recorded using 4-channel DSO is in Fig.3.11 to Fig 3.14 shows the experimental set-up of 2-level inerter. Table.3.1. shows the line voltage, phase voltage and CM voltage values of the 2-level inverter fed IM by simulation using MATLAB/SIMULINK software. Table.3.2. shows the experimental results of line voltage, phase voltage and CM voltage of the 2-level inverter fed IM drive with a DC link voltage of 450V.

20 65 Fig Experimental setup of 2-level inverter

21 66 Table-3.1: Simulated Results Parameters SPWM technique SVM technique Phase voltage Refer Fig 3.9(b) and 3.10(b) 330Vpeak 320Vpeak Line voltage Refer Fig 3.9(a) and 3.10(a) 500Vpeak 515Vpeak Common mode voltage Refer Fig 3.9(c) and 3.10(c) 250Vpeak 186Vpeak Table-3.2. Experimental Results Parameters Phase voltage Refer Fig Line voltage Refer Fig Common mode voltage Refer Fig DSO recording SVM 324Vpeak 538Vpeak 185Vpeak

22 Amplitude in volts Frequency (Hz) Fig FFT of Phase voltage to IM Note: (In the above FFT plot ve scale of frequency is shown deliberately for reading the zero frequency magnitude. Strictly there is no ve frequency will exist and the same Note is applicable for all FFT plots) Explanation: The Fig.3.15, the FFT of phase voltage waveform is as shown and its voltage magnitude is found to be 205Vpeak at zero frequency, however observing the phase voltage DSO recorded waveform (Fig.3.11,ch.1) there exists the steady state values (stair case) hence the phase voltage magnitude is shown at zero frequency but it should be considered as fundamental. It is seen from the above plot that at 40Hz the voltage magnitude is around 7Vpeak. In real sense the fundamental frequency voltage magnitude is around 212Vpeak (205+7). The even harmonic components should not be considered since while doing the FFT, the ve values are changed as +ve as per Origin signal analysis

23 68 software (in the symmetry of waveform the even harmonic voltages will get cancelled), and the triplen harmonic components will circulates hence should not be considered. The 5 th and the 7 th harmonic component is found to be 1V and 0.5V respectively. All the other higher frequency magnitudes of voltages are negligible. 600 Amplitude in dbµ V Frequency (Hz) Fig.3.16.FFT of Phase voltage to IM in dbµv Fig shows the FFT of phase voltage in dbµv. (Note: All the frequency vs dbµv/dbµa plots in this thesis may be utilized for comparison with FCC and CISPR standards which is not the objective/scope of the thesis work)

24 Amplitude in volts Frequency (Hz) Fig.3.17.FFT of Line voltage to IM (2-level Inverter) Explanation: Fig.3.17 shows the FFT of line voltage waveform. The voltage magnitude is found to be 320Vpeak at zero frequency, however observing the line voltage DSO recorded waveform (Fig.3.12,ch.2) there exists the steady state values (stair case) hence the line voltage magnitude is shown at zero frequency but it should be considered as fundamental. It is seen from the above plot that at 40Hz the voltage magnitude is around 30Vpeak. In real sense the fundamental frequency voltage magnitude is around 350Vpeak (320+30). The even harmonic components should not be considered since while doing the FFT, the ve values are changed as +ve as per Origin signal analysis software and due to symmetry of waveform the even harmonic voltages will get canceled and the triplen harmonic components will circulate hence should not be considered. The 5 th and the 7 th harmonic component are found to be

25 70 4.5V and 2V respectively. All the other higher frequency magnitudes of voltages are negligible. 300 Amplitude in db µv Frequency (Hz) Fig FFT of Line voltage to IM in db µv (2-level Inverter) Fig 3.18 shows the FFT of line voltage in db µv. The voltage magnitude is found to be 280 db µv. 60 Amplitude in volts Frequency (Hz) Fig.3.19.FFT of CM voltage

26 71 Explanation: Fig.3.19 shows the FFT of CM voltage waveform. The voltage magnitude is found to be 60Vpeak at zero frequency; however it is seen from the above plot that at 40Hz the voltage magnitude is around 8Vpeak. In real sense the fundamental frequency voltage magnitude may be around 68Vpeak (60+8). From the explanation of Fig.3.19, the FFT of CM voltage at the fundamental frequency is considered as 68 V. The even harmonic components should not be considered since while doing the FFT, the ve values are changed as +ve as per Origin signal analysis software and due to symmetry of waveform the even harmonic voltages will get canceled. The 5 th and the 7 th harmonic component are found to be 1 V and 0.5V respectively. All the other higher frequency magnitudes of voltages are negligible. 200 Amplitude in db µv Frequency (Hz) Fig FFT of CM voltage in dbµv Fig.3.20 shows the CM voltage in dbµv & its magnitude is found to be 190 dbµv.

27 CONCLUSION In this chapter, the measurement of CM voltage is done for the 2-level inverter fed IM. As indicated in the Table-3.1 the simulated CM voltage is 186Vpeak and as per Table-3.2 the CM voltage recorded by DSO (Fig.3.13, ch.1) is 185Vpeak. The simulated and the experimentally recorded waveforms of CM voltage values are in good agreement. The FFT plot of CM voltage at fundamental frequency (40Hz) is around 68V. However the FFT plots gives the voltages of various frequency components. Other high frequency FFT plots are given in appendix-1 for reference only. The frequency Vs dbµv graphs can be utilized to check for Federal Communication Commission (FCC) and International Special Committee on Radio Interference (CISPR) standards for the acceptable limits which are not under the preview of the thesis. Hence it is concluded that measurement of CM voltage has been done and the various frequency voltage components are plotted for 3-phase 2-level inverter fed IM drive.