look for help about igbt short protection

hello. i recently in the process of the igbt short protection. but i meet with a little problem.
my igbt driver CANNT protect the igbt when the igbt is in the short circuit condition.
i use the VLA500-01 driver which is from isahaya, and the igbt module is CM1400DU-24NF.
the attachment is one of a picture from the ocilloscope. the yellow qurve stands for the VGE voltage,
the purple qurve stands for the current of the emitter of the igbt, and the blue stands for the VCE voltage of the igbt.

can someone give me some help, its ergent.thans in advance.

The curve looks plausible so far. It looks like the IGBT is destroyed by the huge overvoltage during emergency switch-off. The only way to reduce the overvoltage are soft switch-off and dynamic clamp. But they don't necessarily work for any load circuit.

I wonder, if this is a realistic short circuit scenario or an artificial test setup?

FvM said:

The curve looks plausible so far. It looks like the IGBT is destroyed by the huge overvoltage during emergency switch-off. The only way to reduce the overvoltage are soft switch-off and dynamic clamp. But they don't necessarily work for any load circuit.

I wonder, if this is a realistic short circuit scenario or an artificial test setup?

Click to expand...

from the datasheet of the cm1400du, the maximum VCE is 1200V, it seems that the the VCE voltage is about 1250V(5*250) which exceeds the rated value a little.
and i think it is no problem.

and this is an artificial test, please see the attachment, it is the experiment setup.
by the way, have u ever used VLA500-01 before?

The inductive overvoltage is nevertheless very high related to bus voltage. The question is, if you are measuring the voltage correctly. You should also the consider, that the inner chip voltage will be higher than the external one. This is particularly true with a low inductance short, as you have it in your setup.

As another point, Vce should be limited by the operation of the high side diode in a bridge circuit, which can't be seen in your measurement. This can happen even with reasonable bus voltage bypassing if the short inductance is too low, or due unsuitable bypassing.

If it's no an overvoltage induced breakdown, it's a case of too large absorbed energy. But according to the waveforms, the driver is acting on IGBT desaturation with acceptable delay. So I can't see a driver problem in this regard.

I'm not using VLA500-01. My experience is limited to Eupec driver modules and custom designed IGBT drivers. I notice, that the VLA500 has a slow switch-off feature, but doesn't support dynamic clamping. What's your Cs setting?

I would check for sufficient bus voltage bypassing with low inductance "snubber" capacitors and repeat the test with a more realistic short circuit impedance, involving a longer wires.