EEEN30042 Power Electronics 2020 Current Control and Voltage Synthesis Lab – Part 2 Introduction This is the lab manual for the Power Electronics course. This lab has two parts: Part 1 – Simulation of inductor current control and voltage calculations, tasks 1-8 Part 2 – Virtual experiments on inductor current control and voltage synthesis (PWM), tasks 9-12. You should already have completed Part 1 (pre-lab exercises) and received feedback via blackboard. Complete Appendix B during this lab and submit as a pdf on blackboard by 1pm on the Thursday immediately following your Tues or Wed timetabled lab. You are not being marked on presentation. Screenshots from PLECS and photographs or scans of sketches and handwritten results are acceptable. If you have an automatic DASS extension, this can be applied, but I recommend you fill in Appendix B while you run the simulations. The types of control covered in this lab are the same as those used in many power electronic systems. In the PLECS model we control inductor current using a PWM (pulse width modulation) voltage and examine the results on a scope. This is a typical requirement in power electronic grid interface converters. The motor control is based on a Cypress 50 W, 4000 rpm motor development system. Bandwidths are representative for motor control and grid interface converters; modern DC/DC power converters would use much higher switching frequencies and bandwidths. Part 2: Voltage Synthesis and Inductor Current Control Objectives of this part of the lab: Take measurements on simulated inverter circuits using scopes. Approximate a desired voltage from a fixed DC supply, by using high frequency switching PWM. Investigate the influence the proportional gain () and the integral gain () exert on the inductor current control. Develop a basic understanding of how the practical control system behaves. Observe the effects of voltage limits and dead time on current and voltage outputs View the switching voltages and harmonic content of voltage and current waveforms for a three-phase motor load. The lab experiments use a PLECS simulation of a power electronic inverter. For Tasks 9-11, an inductor is connected to a single phase H-bridge. For Task 12, a 3-phase permanent magnet (PM) motor is connected to a 3-phase inverter. 1. Getting started You can open PLECS from the start menu (Start > All Programs > PLECS > PLEXS) to get screen shown in Fig. 1. Use File->Open from the menu toolbar and browse to the lab file ‘H- BridgeBothRLLab.plecs’. Alternatively, you can also double click on the file and it should open directly in PLECS. Fig. 1 PLECS opening menu The list in Fig. 1 gives you a set of PLECS components which you can drag into your model. Models have been provided to match the hardware in the lab, so you should not need to add extra components. The opened model should appear as shown in Fig. 2 Fig. 2 Menu bar controls Using the menu bar at the top, ‘File’ will allow you to open, edit and save files. Refer to ‘help’ for more instruction about all aspects of PLECS. The ‘Simulation->Simulation parameters’ menu, shown in Fig. 3 will allow you to change the stop time for the simulation. The units of time are seconds. A stop time of 2s should be Ok for the first test. The solver type and options have been set for you and should not need changing. Fig. 3 Simulation parameters 2. The single-phase H-bridge with RL load. Now take a closer look at the single phase H- bridge circuit model in Fig. 4. Starting from the top left, in the clockwise direction, you should be able to identify the following parts: 1. Reference selection 2. PWM generation 3. Power circuit 4. Displays 5. PI control Fig. 4 Single phase H-bridge PLECS model 2.1 Reference selection The signal selector allows you to switch between inputs in a multiplexed signal. Reference values should be given as desired voltages. They are then scaled by (1/VDC) before being passed to the PWM block, for comparison with a ±1V peak-peak triangle wave. To select the desired input, double-click on the reference selection block, Table i. The pop-up menu in Fig. 5. Should appear. Set the output indices to the channel number you want from Table i. Set to ‘1’ for the constant DC reference. Table i Input reference selection channel Source 1 Constant DC reference 2 Square wave (pulse) reference 3 Reference derived from PI controller. 1 2 3 5 4 Fig. 5 Selection box pop-up menu 2.2 PWM generation If you double-click on the PWM sub-block it will open the underlying model, Fig. 6. This block performs the reference-triangle comparison, and outputs the A-phase top switch gate signal. The inverted reference is used in a second reference-triangle comparison to output the B- phase top switch gate signal, but it is only use in unipolar PWM. Note that green lines are signals, and data is passed in the direction shown by the arrow. Connections for this sub- block are listed in Table ii. If your model doesn’t work at any point, check that input and output connections are within the expected range. Fig. 6 PWM sub-block Table ii PWM sub-block connections Connection Data type Range Name Sub name Function Input Float -1 to 1 VControl Reference voltage Output Boolean 0 or 1 swA+ leg A top switch gate Output Boolean 0 or 1 swB+ leg B top switch gate Output Float -1 to +1V RefTri VControl+ Reference voltage Output Float -1 to +1V Triangle Triangle Output Float -1 to +1V VControl- Inverted ref. (unipolar) In the main model, a second selector switch, Table iii, determines the signal source for the B-phase leg, top switch, for either bipolar or unipolar PWM mode. Set this to 1 for bipolar PWM. Table iii Speed reference selector channel Source 1 Ramp 2 Constant 2.3 Power circuit. Double click on the circuit sub-block to see the details of the power circuit in Fig. 7. Note that black lines are wires in a circuit and have values of both voltage and current associated with them. Hence, electric circuit components can only be used within a circuit block. Meters are used to measure these voltages and currents and convert them into signals for use in the control. The RL Load is password protected, as your task is to find the K and τ values for the load. Input and output connections are listed in Table iv. Fig. 7 Power circuit sub-block Table iv Power circuit connections Connection Data type Range Name Sub name Function Input Boolean 0 or 1 swA swA+ leg A top switch gate swA- leg A bottom switch gate Input Boolean 0 or 1 swB swB+ leg B top switch gate swB- leg B bottom switch gate Output Float -12 to +12V Out1 V_AO A phase leg voltage Output Float -12 to +12V V_BO B phase leg voltage Output Float -24 to +24V V_AB load voltage Output Float -15 to +15A Iload load current 2.4 Displays Simulation values are displayed on a series of scopes. Traces are listed in Table v. Note that the load voltage has been filtered with a 500Hz cut-off, 2nd order filter, so the average value can be viewed. Table v Display details Scope Plot Trace Colour Comment PWM signals Top V_Control+ Red Triangle Green V_Control- Blue Bottom swA+ Red swB+ Blue Unipolar only Load measurements Top V_AN Green Middle V_BN Green Bottom V_AB Green Load measurements 1 Top V_AB filtered Green 500Hz cut-off Bottom I_load Green Fig. 8 shows an example display. Use the commands at the top to change the scale, add cursors and save images. Useful control include: File-> scope parameters, allows you to add or remove channels and add titles, axis labels and adjust scale factors, File-> export, exports an image or data file, View->show cursors (or cursor icon), enables cursors, View-> analysis, allows you to select values based on the cursor positions such as difference, mean, rms, Edit-> zoom controls (or zoom icons) changes the axis scales. Further details on scope controls can be accessed through ‘Help’. Fig. 8 Scope display example. 2.5 PI controller The PI controller compares a square wave reference current with the measured current and generates a demanded voltage Vpi, as shown in Fig. 9. PI gains are set using the gain block labelled Kp and Ki respectively. Fig. 9 PI controller This is the start of the simulation tasks. Task 9 - Inductor characterisation In this task, you will: Measure the inductor current response to a square wave voltage input. Calculate experimental values for the inductor parameters and from the measured inductor current, as you did in Task 1 of the pre-lab exercises. Set-up: You should be using ‘H-BridgeBothRLLab.plecs’ for this part 1) Make sure that DC reference is selected for bipolar PWM and a 2s run time. 2) Run the model by selecting Simulation->start. 3) Adjust the value of the DC reference voltage to give approximately 1A of DC current (+/-10%), on the Load Measurements 1 scope. You may have to run the simulation several times to set this. 4) Change the reference selection to ‘2’ pulse, (see Table i). 5) Double click on the pulse generator and use the value of DC voltage from (3) to set the high-state and low-state outputs, Fig. 10, to give approx ±1A pulsed current. Fig. 10 Pulse generator controls. 6) From the data displayed in scope Load Measurements 1, use the cursors to read off the measurements you need to calculate the values of inductor parameters and. In appendix B, enter the values of inductor parameters and and units and hence write down the transfer function for the plant (i.e. inductive load). Task 10 - Inductor closed loop current control In this task you will Calculate values of and using the values of and for the same damping ratio (0.7) and bandwidth (100 Hz). Verify the closed-loop current response. Observe how the limit on voltage affects the response. Set up: 1) Use your new values K and τ to calculate the Kp and Ki gains for a 100Hz bandwidth and 0.7 damping factor, as you did in the pre-lab exercise. 2) Double click on Kp and Ki blocks and enter your gain values Fig. 11 Ki gain control 3) Change the reference selection to ‘3’ PI, (see Table i). 4) Check that the reference current, (Pulse Generator 1) is set for a high output of +1 and low output of -1 and a frequency of 2Hz. 5) Run the model by selecting Simulation->start. 6) View the response on the Load Measurements 1 scope. Note that the filtered voltage shows a small lag and transient due to the response of the filter. You should see that the voltage is limited to approximately VDC. 7) How well does your response match the expected response from the pre-lab? 8) Double click on the integrator block, and change the upper saturation limit to +24 and the lower saturation limit to -24, as shown in Fig. 12 Integration block. How does this change the response? Fig. 12 Integration block Comment: This effect is called integrator windup. If the output saturates and the reference does not match the measured current then the integration of the error accumulates, resulting in overshoot. Limiting the integrator is one way of preventing this. In appendix B, enter the Kp and Ki gains, including units. Insert screen shots of the closed- loop current response before and after adding the limit to the integrator block. Comment on how well the PLECS response matches the pre-lab in each case. You can now close H-BridgeBothRLLab.plecs’ 2.6 Addition of dead time Dead time is a small delay to the turn-on of each switch, to ensure that the other switch in a leg is switched off before the new device is switched on. This is implemented in a different model, H-bridgeDeadTimeRLLab.plecs, as shown in Fig. 13. Parts 2 and 4 are the same as in Fig. 4. Part 3 has been slightly modified so that the switch commands are passed separately, but the inverter circuit is unchanged. You will only be using this circuit with a constant reference for the reference generation so Parts 1 and 5 have been removed. However an additional Part 6 has been added, to introduce the dead time. Because this is very short (typically 1μs), the maximum simulation time-step has been reduced to 0.8μs. This makes it slow to run so the stop time has been reduced to 0.2s. Fig. 13 Single phase H-bridge PLECS model with dead time Task 11 – Effect of dead time For power converter control, it is often sufficient to assume the output voltage equals the reference value. However, in this task, you will be looking at the actual output voltages. In task you will: Verify the accuracy of the synthesised average voltage for Bipolar and unipolar PWM With and without dead time Set up: You should be using H-bridgeDeadTimeRLLab.plecs for this part. 1) Use the reference selector to set bipolar PWM (see Table iii). 2) Check the input reference (Constant) is set to 1 [V]. 3) Run the model using Simulation->start 4) Using the load measurements scope and the cursors function, measure the period of V_AB, and the time for which it is high, and record the high and low values of the switched voltage. 5) Using the load measurements 1 scope, estimate the average value of the filtered voltage in steady-state. There is some ripple on this value, so you will have to use the scope to find the average value. Fill in the first row in Appendix B, task 11 6) Change the reference selector to unipolar PWM. Repeat steps 2-4. Fill in the second row in Appendix B, task 11 2 6 4 3 7) Double click on each of the four Turn-on delay blocks. Set the delay to 1e-6 (1μs), Fig. 14. Make sure you do this for all 4 blocks. Fig. 14 Setting the dead time 8) Change the reference selector to bipolar PWM. Repeat steps 2-4. Fill in the third row in Appendix B, task 11 9) Change the reference selector to bipolar PWM. Repeat steps 2-4. Fill in the fourth row in Appendix B, task 11. Comment on how well your measured values match the calculations from the pre-lab, without and then with dead time. You can now close H-bridgeDeadTimeRLLab.plecs 3. The three-phase inverter with motor load. Fig. 15 3-phase inverter with PM motor load The 3-phase inverter is shown in Fig. 15. Like Fig. 4 it also contains the following parts: 2) PWM generation 3) Power circuit 4) Displays 5) Control (but motor control rather than inductor current control) 3.1 PWM generation In Fig. 15 the motor control block generates the reference voltages, which are then scaled by (2/VDC) for the sine-triangle comparison. The Osc block uses an oscillator circuit to generate the ±1V triangle wave and also a trigger signal for synchronising the digital control of the PM machine. The 3-phase sine triangle comparison generates logic gate commands for the top switches in the inverter. Bottom switch signals are derived as the inverse of the top switch signal, neglecting dead time. 3.2 Power circuit The power circuit is supplied from a 24V DC link. Three inverter phase legs are connected to a 3-phase PM machine, with external load torque. The circuit diagram is given in Fig. 16 and the block connections are listed in Table vi. The PM machine is a 3-phase, sinusoidal permanent magnet machine. If you are interested, you can double click on the block to view the machine parameters and read the help information to find out how it is been modelled in the PLECS library. 2 3 4 5 There is no access to the star point in the PM machine so a 3-phase 100 kΩ resistor network has been used to create a measurement star point. This is a common technique in power converters, which avoids the need for a neutral connection to the machine. Fig. 16 3-phase inverter power circuit Table vi 3-phase inverter power circuit connections Connection Data type Units Name Sub name Function Input Boolean Logic sw Sw[1..6] Switch control signals Input Float Nm Tload Load Torque Output Float A PM motor IA Phase A current Output Float A IA Phase B current Output Float A IC Phase C current Output Float Rad/s Speed Motor Speed Output Float Rad Angle Motor angle Output Float Nm Torque Motor Torque Output Float V Vmeas V_AN A phase-star voltage Output Float V V_BN B phase-star voltage Output Float V V_CN C phase-star voltage Output Float V V_NO Star-midpoint voltage 3.3 Display Signal displays on the scopes are listed in Table vii. To improve speed and reduce memory requirements, only the last 10ms of the high resolution switching signals have been saved. You will need to expand the traces to see the detail. Table vii PWM signals Scope Plot Trace Colour Units Comment PWM signals Top VA reference Red V Last 10ms only VB reference Yellow V VC reference Blue V Triangle Green V Bottom sw1 Red Logic sw3 Yellow Logic sw5 Blue Logic Load measurements Top I_A Red A I_B Yellow A I_C Blue A Middle V_AN Red V V_BN Yellow V V_CN Blue V Bottom V_NO Green V Motor measurements Top Speed Green Rad/s Middle Position Green Rad Bottom Torque Green Nm Reference and filtered voltages Top V_A Ref Red V V_B Ref Yellow V V_C Ref Blue V Middle V_AN Filtered Red V 1 kHz cut-off V_BN Filtered Yellow V V_CN Filtered Blue V Bottom Ref speed rpm Green Rad/s In the main model, a selector switch, Table iii determines the signal source for the Speed reference. Set this to 1 for the ramp input. Table viii Speed reference selector channel Source 1 Ramp input 2 Constant 3.4 Motor control. The motor dq control uses inner current control and outer speed control loops, with rotor field orientation. The implementation is beyond the scope of this exercise. Task 12 – 3-phase voltages In this task, for a 3-phase inverter, you will investigate the Fundamental voltage and current PWM harmonic voltage and current Switching voltage levels Set up: You should be using InverterMotor.plecs for this part. 1) Set the speed reference selector to 1, Ramp input. The speed should ramp up to 2000 rpm in 1s. Check that applied load torque, Tload, is zero. 2) Run the model using Simulation->Start 3) View the reference speed and voltage in the references and filtered voltages scope. In appendix B, comment on the relationship between motor speed and voltage magnitude. Comment on how the measured filtered voltages compare with reference values. 4) Set the speed reference selector to 2, constant. The reference speed is now constant at 2400 rpm. 5) Set the motor initial speed to match this value. Double click on the motor in the circuit model and set the initial rotor speed in rad/s as shown in Fig. 17. [Use 1 rpm=2π/60 rad/s]. Fig. 17 Setting the initial speed 6) Double-click on the constant Tload block and set the load torque to its rated value of 0.128Nm. 7) View the motor currents and voltages on the load measurements scope, and the filtered voltages on the ‘References and filtered voltages’ scope. In appendix B, record the magnitude of the current and filtered voltage in steady-state. 8) Use the cursors to estimate the frequency of the current. 9) Electrical frequency fe relates to mechanical frequency fm by the pole pairs. = This machine has 4 pole pairs. Check your electrical frequency value. You will need to convert angular speed in rpm to a frequency in Hz. 10) Select the FFT, using either View->Fourier spectrum or the icon shown in Fig. 18. Fig. 18 FFT control Fig. 19 FFT Settings 11) You will need to set up the FFT range. Click on the ‘f’ value above the plot. You should see a pop-up box. Set the base frequency to the fundamental from (7-8). You have 1s of data so you can look over at least 20 periods. 12) Click on the ‘N’ value above the plot. This sets the harmonic range. Again you could set this to 20 times the fundamental. In appendix B record the magnitude of fundamental for the phase current IA, phase voltage V_AN and star-point voltage V_NO. 13) Repeat steps 10 and 11 to view the switching frequency. Set the ‘f’ to 1000, and ‘N’ to 25. This should show the first two switching frequencies. In appendix B record the magnitude of the switching frequency and double frequency components for the phase current IA, phase voltage V_AN and star-point voltage V_NO. 14) Close the FFT window and return to the time domain view of load measurements. Expand the plot to see approx one cycle of the fundamental current. You should see five different voltage levels on the phase voltages (e.g. V_AN) and 4 levels on the V_NO In appendix B record the voltage levels for V_AN and V_NO. Comment on how well these match the pre-lab calculations. This is the end of the practical experiment part of the lab. Appendix B ID: Student Name: Task 9 K= Unit: = Unit: Plant Transfer function (use numeric values for K and ) = Task 10 Identify the values of and that will result in the closed loop dynamics with a natural frequency of 100 Hz and a damping ratio of 0.7. = Unit: = Unit: Plot of current response, without integrator limiting Plot of current response, with integrator limiting Comparison between PLECS and Matlab response: Task 11 Dead time PWM 〈〉 V , V , V Period μs Duty* No Bipolar No Unipolar Yes Bipolar Yes Unipolar *Use D1 for Bipolar, DNZ for unipolar Comparison between PLECS and pre-lab calculations, with and without dead time. Task 12 Comment on the relationship between motor speed and voltage magnitude. How well do the measured, filtered voltages compare with reference values? Time domain magnitudes Peak rms Current Filtered voltage (phase) Harmonic components Signal Fundamental Switching harmonic 2 x Switching harmonic Phase current IA A Phase voltage VAN V Star-midpoint voltage VNO V Put a ring around your choice. By comparison with the time domain values, are the harmonic components: a) peak b) rms? Voltage levels across phase A, (VAN) in V: Star-midpoint voltage levels (VNO) in V: How well do the measured voltages levels match pre-lab calculations?
欢迎咨询51作业君