Three phase inverter

Discussion in 'The Projects Forum' started by idrees.engineer, Jan 7, 2012.

  1. idrees.engineer

    Thread Starter New Member

    Jan 7, 2012
    6
    0
    I am using three half bridge IGBT modules specified as above in Three phase inverter.

    My design is as under

    Driver IC = IR2235
    Vg(ON) =15V for high side and also for low side
    Vg(OFF) =0V for high side and also for low side
    Rg(ON) = 25 Ohm
    Rg(OFF) = 5.6 Ohm
    Bootstrap capacitor value is = 1uF/25V
    Switching frequency is = 4KHz
    Dead Time = 6uSec.
    DC BUS Voltage = 600V
    Microcontroller= PIC18F4431.
    Optoisolater = HCPL2631.
    snubber capacitor across IGBTs C1 and E2 of value 0.22uF 1200VDC/600VAC.


    i have tested the inverter at 150VDC. with 500Watt motor.

    it is working good but there is a problem.

    From each half bridge when upper or lower IGBT is off and after dead time when other is going to ON a spike of 1.2uSec appears at rising edge of IGBT which is going ON and also this pulse appears in off time of other IGBT whis is OFF.
    The load current which is being drawn by load is = 1.2Amp.

    Although it did not give any problem but as i increase DC bus voltage the amplitude of this spike also increasing so i limited the DC bus to 150V.

    I have to apply 600VDC but i am afraid to spike which appears on both pulses

    Please suggest me the solution to remove this spike.

    Thanks in advance
     
  2. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    Post a schematic and pictures/drawings of the spikes.
    Also, unless you forgot to mention a transformer on the output, then this is not an inverter but something along the lines of VFD or variable frequency driver.
     
  3. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    If you have headroom in terms of temperature rise for the IGBTs you can try to increase the gate resistor values.

    I used for a 2kW inverter up to 56 Ohms.
    What type is that capacitor you mentioned? It has to be a low ESR type for this kind of application. I used up to 1uF for DC bus voltages of 460V.

    Post what Kubeek suggested and also a picture of the IGBT/capacitor layout.
     
  4. idrees.engineer

    Thread Starter New Member

    Jan 7, 2012
    6
    0
    I have attached the schematic of my inverter control PCB also spike picture is attached as well

    Kind regards
    View attachment THREEPHASEINVERTER.pdf

    View attachment SCHEMATIC1 _ IGBT Drive.pdf

    View attachment Waveform.pdf View attachment THREEPHASEINVERTER.pdf

    View attachment SCHEMATIC1 _ IGBT Drive.pdf
     
  5. jimkeith

    Active Member

    Oct 26, 2011
    539
    99
    Are you talking about a voltage spike or a current spike?
    and what are the magnitudes?
     
  6. idrees.engineer

    Thread Starter New Member

    Jan 7, 2012
    6
    0
    This is a voltage spike. Amplitude increases as i increase DC Bus voltage.
    At 60VDC Bus voltage the spike's amplitude is 4V
     
  7. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    Have you tried increasing dead time?
     
  8. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    I take as an example R7 and R23 in your second schematic. You are using the diode to discharge the gate faster than it was charged I suppose? Then these two resistors shouldn't have the same values. R7 would need to be higher to get different charge discharge times.

    Since you measured the voltage peak I suppose you have an oscilloscope. Please post the waveform captured on the scope.

    A voltage peak is normal when switching off IGBTs. How much is it at 150V?

    What IGBT are you using?
     
  9. idrees.engineer

    Thread Starter New Member

    Jan 7, 2012
    6
    0
    yes i increased dead time step by step up to 25uSec but the spike still there
     
  10. idrees.engineer

    Thread Starter New Member

    Jan 7, 2012
    6
    0
    The spike magnitude is 4V at 150V DC Bus voltage.
    I am using semikron IGBT "SKM400GB123D"
    The captured waveform is attached herewith

    regards View attachment Waveforms.pdf
     
  11. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Now I get it. You are talking about the gate voltage and you are worried about it being turned on when the other IGBT starts conducting.

    Try to increase the gate resistor values a bit, e.g. 10 Ohm for the ones that charge the gate ( not the ones with the diodes in series)
     
  12. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,004
    1,525
    Is your gate driver mounted close(less than 1 inch/25mm) to the IGBT? If you have a length of wire between the drivers and gates you maybe getting an inductive kick on the wires.
     
  13. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Good point, I always forget to ask this question. Where exactly did you measure the pulse, at the gate or at the driver output? It may not seem to but this makes a huge difference.
     
  14. idrees.engineer

    Thread Starter New Member

    Jan 7, 2012
    6
    0
    I did also increase the charge resistor step by step up to 56 ohm
    but the problem is still there.

    Secondly i have mounter IGBT drive directly on the IGBT so there is not wire distance between IGBT and driver
     
  15. Hansh

    New Member

    Jun 16, 2009
    1
    0
    I see some designs that each MOSFET / IGBT a Capacity of + / - 4n7 .. 12n over the drain / source is connected. (= 6x is a 3 phase motor).
    Could this be necessary?
     
Loading...