Fun with 230v and mosfets, or not... :(

Hello, so after getting some mosfet drivers and 600v mosfets i set about finishing my prototype ready for use, so after blowing up various things and burning out a few mosfets and a driver i finally thought i'd made progress.
So yesterday i had a slight issue where the laptop didn't like sharing a ground with the mosfet psu and 230v dc supply and i managed to blow a couple mosfets, led's and an arduino :slight_smile: replace one of the mosfets with a new spare i had (Infineon IPP65R225C7) and a old (MTP3N60) that i found laying around.
Got everything back up and running again to which it's spent 3 hours running perfectly, turned off as it should and then i put another hour on it, roll on an hour later and the lights are still on... pop down to see the arduino has crashed and the infineon mosfet has burnt out, my thoughts are either the mosfet burnt out and spiked the arduino causing it to crash or the arduino crashed and just sent a low signal to the mosfet causing it to overheat.
I just can't work out how i'm burning the infineon mosfets out, they run cold and then all of a sudden burn out randomly, i'm probably pulling an amp through each and the main difference between the infineon and the MTP3N60 is the RDS(on) where the Infineon is about 0.2ohms @ 25c and the other is about 2ohms @ 25c, gates are being driven by a 9A TC4422A driver at around 15v. Should i be current limiting the gates?
Cheers

Infineon

MTP3N60

Should i be current limiting the gates?

I dunno, should you? You have the all the information, not us.

This is one of those really simple to answer questions. The problem is no doubt lurking somewhere in the code and the schematics you did not post :wink:

Once at these voltage levels it only takes a few tens of microseconds to commit a MOSFET die to becoming vapour.

This means the protection circuitry has to be solid and absolutely prevent any instance of:

overvoltage spike on drain (for instance caused by ringing due to stray inductance)
shoot-through (ie gross overcurrent spikes)
excessive dV/dt (can lead to false switching)
overvoltage spike on gate w.r.t. source (source leads stray inductance is an issue you have to understand for this)
overvoltage spikes on gate driver 12V supply.
undervoltage gate drive (not switching fully on).

Protection also has to function reliably as the voltage rails come up and drop down as the power is applied
and removed.

With high power and high voltage everything gets harder.

These days IGBTs have replaced MOSFETs at these kind of voltages as they are more robust
for most applications (despite often having higher losses and slower switching).

WattsThat:
I dunno, should you? You have the all the information, not us.

This is one of those really simple to answer questions. The problem is no doubt lurking somewhere in the code and the schematics you did not post :wink:

The code is just basically just fading the loads up and down via pwm, under normal operation the Infineon Mosfet stays cold whereas the other mosfet just gets warm although i guess i can put that down to the difference in resistances across the mosfets, they are both loaded with around 1A worth of led's

Will have a better look into protection but just though it was strange how i've managed to randomly kill a few of the Infineon mosfets whereas the random old one thats alot lower spec'd seems happy enough :slight_smile:

Cheers

Without a 'scope and high-voltage probe(s) there’s no easy way to know what the failure
mode is that’s taking out the MOSFETs. Poor layout and stray inductance, lack of decoupling,
anything is possible, we have no information about the schematic and layout of the circuit.