GaN FET blew up

[cupoftea]

Hi,
I was testing a "bidirectional forward converter" 240VAC with high voltage capacitor reset.
I broke the path in the HV cap to solder in a sense transformer.
But then took it out, and powered up, forgetting that the HV reset cap was now open circuit.
Due to this, the transformers leakage and magnetising inductance energy then had nowhere to go....and so blew up the low side NV6128 GaNFET

NV6128

https://www.mouser.co.uk/datasheet/2/1451/NV6128_Datasheet_FINAL_10_11_22-3076917.pdf

The thing is, the high voltage got connected, via the blown FET, directly to the control circuitry, and blew a good few things up there aswell.
Would you say these kind of GaNFET modules are more likely to blow up and cause control circuit damage, when compared to just plain old TO220 Si FETs?

 

[wwfeldman]

But then took it out, and powered up, forgetting that the HV reset cap was now open circuit.
Due to this, the transformers leakage and magnetising inductance energy then had nowhere to go....and so blew up the low side NV6128 GaNFETssssss

sounds like operator error
what happens if you put it back together, properly?

 

[cupoftea]

Thanks Yes it works then, though the accident has happened, and the board is well blown up.

 

[dick_freebird]

GaN FET reliability is not so well understood, key manufacturers
went to a lot of effort to tamp down discussion and knowledge
about the gate breakdown reliability issues (for me, a tasking to
make double sure my gate driver could not exceed the 5V limit
of the day).

Si MOSFETs have the "benefit" of decades of design-practice
sandbagging. The GaN guys just want you to scamper ahead,
full speed. Which is fine for them because field returns are
your problem if anything exceeded their low-margin-in-pocket
abs max specs, and who's going to prove anything about a
die sweated face down to the board that could pin it on them?

 

[Easy peasy]

you have provided yet another good reason to interrogate the cap voltage to deduce/measure the current.

 

[mtwieg]

Would you say these kind of GaNFET modules are more likely to blow up and cause control circuit damage, when compared to just plain old TO220 Si FETs?

By all means, repeat the same experiment with a MOSFET of comparable ratings and tell us what happens.