17:42 <carl0s> Hi, sometime ago i read about a test done overnight on the glasgow. IIRC the test involved reading the temperature of an IC. I can't find the issue on the github tracker. By any chance do you recall that issue?

17:45 <whitequark> i think that wasn't on the issue tracker, only irc

17:45 <whitequark> cc marcan

17:46 <whitequark> (marcan has done the test)

17:47 <carl0s> thanks, maybe i read it somewhere else

17:47 <whitequark> marcan's twitter possibly

17:48 <marcan> carl0s: it's on my twitter. it was a torture test of the shifters, shorted out.

17:48 <marcan> tl;dr they seem to survive and work fine after being shorted out hard for hours on end

17:49 <marcan> https://twitter.com/marcan42/status/1117125972254269440

17:49 <Hellsenberg> carl0s: https://github.com/GlasgowEmbedded/Glasgow/issues/135

17:49 <marcan> that's a different one

17:49 <marcan> no overnight test there

17:50 <marcan> that was debugging a specific issue, not a torture test, though it also involved shorting things out

17:50 <Hellsenberg> oh, I see

17:51 <Hellsenberg> marcan: I thought something ended up shorting out in the end, though?

17:51 <marcan> my TPD3S014 died at some point, precise cause unknown, during the testing in that github issue

17:51 <marcan> (pass fet shorted hard, which is not an uncommon failure mode)

17:51 <marcan> (so basically it no longer protected anything, I noticed when the thing was pulling 3A from USB)

17:52 <marcan> I don't have a good repro case and it could've been a fluke, so I didn't investigate further; I can't think of a design issue that could cause that, the thing should survive.

17:53 <marcan> so basically if this becomes an endemic problem somehow, we curse hard at TI, but I sure hope it doesn't

17:53 <marcan> I guess it's worth trying to repro it on the revC1s

18:07 <carl0s> Thanks whitequark, marcan and Hellsenberg :), i was about to explain some coleagues about stress tests and that issue was the first that came to my mind