Fabio Manganiello on Nostr: nprofile1q…yg4nq thanks for the link, it's very useful! In my case, I'd exclude ...
nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpquh6gl0fjh50ud9wv79tcgvr50ty935wpp4vxy5g0p6eyagmnggcstyg4nq (nprofile…g4nq) thanks for the link, it's very useful!
In my case, I'd exclude anything related to sudden over-voltage over either the 3.3V or 5V rails (I'm usually very careful with these things, the RPis in my living room can run for up to a year before frying, and in case of macroscopic over-voltage I'd expect them to fry nearly instantly). And also fiddling with the GPIO or touching the RPi (as I said, sometimes it even happened when the RPi was doing nothing else but sitting on the cupboard). Similar diagnosis for anything that would cause the board to overheat (all the times it happened, the board wasn't hot when I unplugged it).
But drawing large currents from the GPIO could probably be a culprit, as my I2C/SPI breakouts are all attached over GPIO, and that was already my primary suspicion.
The first test I'd try may just be to throw a PiJuice to fix any potential issues with voltage stability in my home network (something that I already suspect because the batteries of 3 different electric skateboards also got fried over a couple of years, maybe running a small data center in your home has its risks...), but then it sounds that plugging a current meter to the GPIO to get an idea of the absorbed current (and maybe doing the same test with different combinations of the breakout sensors) could also provide some valuable insights.
In my case, I'd exclude anything related to sudden over-voltage over either the 3.3V or 5V rails (I'm usually very careful with these things, the RPis in my living room can run for up to a year before frying, and in case of macroscopic over-voltage I'd expect them to fry nearly instantly). And also fiddling with the GPIO or touching the RPi (as I said, sometimes it even happened when the RPi was doing nothing else but sitting on the cupboard). Similar diagnosis for anything that would cause the board to overheat (all the times it happened, the board wasn't hot when I unplugged it).
But drawing large currents from the GPIO could probably be a culprit, as my I2C/SPI breakouts are all attached over GPIO, and that was already my primary suspicion.
The first test I'd try may just be to throw a PiJuice to fix any potential issues with voltage stability in my home network (something that I already suspect because the batteries of 3 different electric skateboards also got fried over a couple of years, maybe running a small data center in your home has its risks...), but then it sounds that plugging a current meter to the GPIO to get an idea of the absorbed current (and maybe doing the same test with different combinations of the breakout sensors) could also provide some valuable insights.