Vce is not responding to Vge in IGBT.

Discussion in 'The Projects Forum' started by Shiranga, Feb 15, 2016.

  1. Shiranga

    Thread Starter New Member

    Dec 22, 2015
    7
    0
    Hi all,
    I am making an IGBT inverter based PMSM motor controller. Operating voltage of the inverter is 330V(DC). I have chosen 1200V/150A rated IXYS igbts as power switching devices. My gate drivers produce +15/-8 PWM signals to control them. My anti parallel diodes are rated to 1200V/ 150A as well. Operating PWM frequency is 5kHz.
    Basically I tested them with 150V(DC) power supply. It worked nicely and inveter took around 75A from the supply. Then I stepped up the inverter voltage to 330V and tested the system. But the motor did not run properly because IGBT's failed in lesser supply current values.

    Then I inspect waveforms using an oscilloscope and I saw that Vce is not responding to Vge in some intervals. I have attached the oscilloscope image below which I saw many times.
    Is this the IGBT latching problem? or should there be any other reasons to generate this kind of a waveform (Anti parallel diode failure, etc) in motor controller inverter?
    Please help me to solve this problem.

    Thank you.
     
  2. crutschow

    Expert

    Mar 14, 2008
    13,052
    3,244
    Without seeing the circuit it's not possible to solve your problem.
     
  3. Shiranga

    Thread Starter New Member

    Dec 22, 2015
    7
    0
    Hi crutschow,
    I have attached basic schematics of our inverter. It is same as other basic invertes. DC supply voltage is 330Vdc. Each IGBT has an individual IGBT gate driver (+15/-8). Out put of the inverter is connected to 3-phase permanent magnet synchronous motor.
    Please ask any details if you want to know.
     
  4. crutschow

    Expert

    Mar 14, 2008
    13,052
    3,244
    You schematic doesn't show what is driving the transistor gates.
    The gate voltage to the top IGBT's must be greater than the supply voltage for them to fully turn on, since they are acting as emitter followers.
     
  5. Shiranga

    Thread Starter New Member

    Dec 22, 2015
    7
    0
    crutschow,
    What did you mean by the top igbt's gate voltage must be greater than the supply voltage ? Can you please explain that further?
    BTW I am using a isolated power supply to power the igbt gate driver. And 10ohm Rg was used to control the Igpeak to around 3.5A.
    I have attached a basic block diagram of gate driver.
     
  6. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,809
    1,105
    Which model gate driver are you using? Is the bootstrap functioning correctly? What is the output voltage, relative to ground, that the upper gate actually gets? It would help us to help you if you post your complete schematic, not just bits of it or block diagrams.
     
  7. Shiranga

    Thread Starter New Member

    Dec 22, 2015
    7
    0
    Alec_t,
    Honestly there is no complete schematic with me. I will prepare one and post it here.
    Powerex BG2B universal gate diver prototype board is used to drive IGBTs (gate driver: VLA503-01, DC/DC converter: VLA106-15242).
    Six ixdn 75n120 igbts were used as power switching devices. Six GSXF120A120S1 diodes were used as anti-parallel diodes.

    Is there any method to measure the proper bootstrap functionality of gate driver power supply?

    What did you mean by the output voltage, relative to ground? is that the phase voltage? we cannot measure the phase voltage of the motor because the connected point terminal of star winding connection is not out from motor. Line to line voltage can vary depending on duty cycle.

    Thank you
     
  8. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,809
    1,105
    Check with a scope. If the drain voltage of the top IGBT is 330V then the gate voltage should be 345V to turn the IGBT fully on. See crutschow's comment.
    Sorry, that was less than clear. I meant the output voltage of the top IGBT driver, i.e. the gate voltage when the bootstrap is functioning correctly.
     
  9. crutschow

    Expert

    Mar 14, 2008
    13,052
    3,244
    The top bridge IGBT's are acting as emitter followers.
    To fully turn on, the gate voltage must be higher than the emitter voltage by the Vgs turn-on voltage.
    Since, when fully turned on, the collector voltage is near the emitter voltage, this means the gate voltage must be greater than the collector (supply) voltage.
     
  10. Shiranga

    Thread Starter New Member

    Dec 22, 2015
    7
    0
    crutschow, Alec_t,

    It is absolutely correct that you are pointing out about Vge. I think that it happen with my inverter also. Because I have used an isolated power supply to power the gate drivers and it's ground is connected to emitter as done by everyone. If you look the oscilloscope photograph which I have attached in first message(I have attached again), you can see the Vge (yellow) and Vce(blue) with reference to emitter. When the Vce(blue) become zero then emitter voltage (Ve) become nearly equal to supply voltage (Vc, 330Vdc). At that time Vge(Yellow) is 15V with reference to emitter. That mean 345V. Is that the thing you are pointing out?

    My problem is, with some intervals Vge( {Vg-Ve}gate voltage with reference to emitter) go to +15V (345V relative Ve) and -8V according to PWM signals but Vce ({Vc-Ve}) seems latched to zero. That means gate is fully on and emitter and collector voltages become close to supply voltage. This is the problem I need your help with.
     
  11. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,809
    1,105
    Can you reduce the PWM frequency to see if the problem is speed-related?
     
  12. Shiranga

    Thread Starter New Member

    Dec 22, 2015
    7
    0
    Yes Alec_t,
    I have tested the system with different PWM frequencies started from 15kHz to 3kHz. But there was no significant improvement regarding this issue. Seems it is not frequency dependent. But It may depend on the current. I cannot see this issue for low current value (current from supply). When the current increase, this issue can be observed frequently.
     
Loading...