0

A pair of IGBTs that I was using in a high frequency application failed with 3 terminal short. The devices are rated 650V that's well higher than the operated voltage of 200 V. Snubbers were working effective in the circuit. So, the failure could not have happened due to voltage of high dv/dt, I suppose.

I specifically want to know if IGBT Gate insulation can breakdown and form a short circuit between gate and channel during channel breakdown due to high temperature from switching losses.

Emmanuel
  • 19
  • 4
  • 2
    So, the failure could not have happened due to voltage of high dv/dt, I believe. That's a bold remark and I do not agree with it. Maybe the snubber network isn't "good enough" to suppress very high dV/dt pulses. Can a three terminal short ... any other reasons excluding high voltage or high dv/dt? If I said: "No, it cannot be damaged from that." would you believe me? An IGBT has maximum ratings that should not be exceeded. Damage occurs from exceeding these ratings. Yes, that's a "duh" answer but what answer did you expect? – Bimpelrekkie Feb 04 '21 at 09:19
  • It's a well designed snubber. There were neither voltage spikes nor high voltage. I've verified that. I specifically want to know if IGBT Gate insulation can breakdown and form a short circuit between gate and channel due to high temperature from switching losses. – Emmanuel Feb 04 '21 at 09:34
  • Keep in mind that the input to a typical IGBT is a MOSFET and the 'collector' of the IGBT (NPN) is actually the emitter of a PNP device. Anything that a MOSFET suffers from, an IGBT can suffer from. – Peter Smith Feb 04 '21 at 09:49
  • I specifically want to know if IGBT Gate insulation can breakdown and form a short circuit between gate and channel due to high temperature from switching losses Yes it can. If anyone disagrees then please prove me wrong. – Bimpelrekkie Feb 04 '21 at 10:25
  • @Bimpelrekkie You are not wrong. – winny Feb 04 '21 at 11:38
  • "Yes, that's a "duh" answer but what answer did you expect?" Most questions asked here on EE.SE, including this one, can be answered with a "duh" answer.. IMO this question should just have been closed as it is not really EE related. –  Feb 04 '21 at 23:48
  • we ned to see your circuit and know operating conditions. || ANY failure mode can lead to multi-pin short. | Try a reverse biased zener gate to source close to device. Vz slightly above Vdrive peak. Ensure Vdrive peak is well short of Vgs max. – Russell McMahon Feb 07 '21 at 11:28
  • Thanks Mr. Peter and Mr. Russell, for your valuable inputs. I've incorporated your suggestions and now the circuit is working well. Thank you. – Emmanuel Feb 11 '21 at 08:08

0 Answers0