@Deodand mentioned this in another thread:
In that temp sensor fault example what I observed was that the fault codes were recorded by
metr. as well as still available in vesc_tool because the focbox never fully bounced.
in this incident there were early cutouts and no recorded faults. which started my investigations before the final blown fet.
It’s still possible that the lower fault stop time contributed to this blown fet. it would have to be a contributing factor to something that did throw faults though which I didn’t catch. so jury’s still out on what blew this thing.