We are using two ADZM3223ARZ's in a 400V Full-Bridge running at 100kHz. They run fine everything looks good, then out of the blue during testing for a few days they just stop working. 3.3V input signals are there , and 15V supplies are there, but no high side or low side output. On the first occurrence it was just one IC and we replaced just it and everything worked again. The instance prompting this request for help was a second prototype in which both IC's stopped at the same time. Now I am worried.
The circuit is attached. It is bootstrapped. We probably do not need the buffer PNP pull down transistor and we could remove it if deemed necessary, but I fail to see any failure cause mechanism. The IC's are not hot before and after stopping, the MOSFET's are not defective. At the driver output pins in unpowered state I measure 3.3K due to the 3.3K load resistor on MOSFET gates, so nothing appears damaged. I am pretty sure if I just swap out these two IC's and nothing else it will work again - for a while at least.
What could possibly be causing the IC's to just permanently stop outputting signal - dV/dt or voltage induced damage? Like I say the waveforms look clean to me. Most of my experience with damaged drivers is they run hot - these just stop working. I would appreciate some hints on possible causes.