My experience has been that RPis have always run hot and always benefit from either active cooling or heat sinks larger than the rinkydink ones you get from vendors. Sure, they’ll “run” without, but weird behavior is par for the course on every single RPi I’ve had.
Heat increases electrical resistance which draws more power to maintain the same signal which increases heat which increases resistance… eventually you can’t get a signal (because there’s not enough current to supply the signal, or because something melted). Another experiment you could run is to add cooling to the RPi and see if the issues stop. I’ve unfortunately not had quite as good an experience with water on my electronics as Ron has, and I don’t make my ice from distilled water, so I can’t recommend an ice bath for that. Might try a fan. —dho On Thu, Oct 8, 2026 at 22:33 <[email protected]> wrote: > There is now what seems to me a large amount of information here. I'm > tempted to say much more than I expected. > > I feel I should clarify something.. I'm not personally all that bothered > by what happened. Something went wrong, but I managed to get the system > working again with a fresh install. From that perspective, my post had two > objectives: > 1) Report the information I had out of curiosity, to see if anybody out > there would say: "yup! seen that before! this is what the problem is!", > 2) Report the problem and the fact that unplugging the pi as I did is > likely not what you want to do. > > Both those objectives have been satisfied in spades. > > With respect to rons comment: "folks, fyi, they demo these embedded boards > sitting in boiling water all the time.. etc.." I would like to clarify > that when I bought the pi, the purchase was not the result of a long > intensive research into these devices to see how they would hold up in > "extreme environments". It was the result of learning that I could install > plan9 on them. Previous to that, I did not have a computer that could run > plan9, and qemu was something I was distantly aware of, at best. So I got > all excited and bought one. Needless to say, I have learned what feels > like an enormous amount from this thread. Yes, I have heard that you can > poor water on a thinkpad and that that won't hurt it, but I hadn't realized > that the water in those experiments actually flows over the electronics. I > can only assume that the electronics on these pcb's are covered with an > electrically insulating layer of some kind. > > All that said, a message I am receiving is that there is a very clear > desire to figure out what happened to the pi. I am being asked to run > various experiments. Out of appreciation for the help I am getting in > general, I am willing to do some experimentation. However, I would like > the objectives of said experimentation to be clear, and some conviction > that they will yield results that are not already expected. I can run the > intense calculation and do the test that demonstrates that the pi got hot, > but seeing as we already know that will happen, it won't teach us > anything. Or am I wrong? > > It seems evident that if I force the pi to do some intense calculation, it > will heat up, and eventually hit some thermal limit. I will make the > statement that when it came time to actually perform said calculations, I > would have researched just "how hard" I could push the pi before risking > damaging it. And I would have invested in heat-sinks, and fans, the whole > kit. Or I would have switched to a different SoC with much more ram, or > whatever else I needed, and equipped -it- with all those devices. > > So if I'm going to do testing, I would like a clear understanding of what > useful information we are hoping to get out of it. To the best that I can > ascertain so far, no ice-cube is necessary. All I need to do is get it to > heat up enough, preferably in such a way that I can control the heating so > that it stays within the advertised safe limits of the pi, to check if that > heating, on it's own, causes the error messages that I recorded. If it > does, well, we'll have information to support the theory that heating of > the pi caused them. > > The hypothesis of Bakul Shah could be the motivation for other experiments. > > I do not currently have access to the pi. I will again in a few days. > Thank you all! > *9fans <https://9fans.topicbox.com/latest>* / 9fans / see discussions > <https://9fans.topicbox.com/groups/9fans> + participants > <https://9fans.topicbox.com/groups/9fans/members> + delivery options > <https://9fans.topicbox.com/groups/9fans/subscription> Permalink > <https://9fans.topicbox.com/groups/9fans/Tc322db746dabdc8f-M0b9296086ae31f99cbb386ac> > ------------------------------------------ 9fans: 9fans Permalink: https://9fans.topicbox.com/groups/9fans/Tc322db746dabdc8f-M008f3843aaed08ab9566b78a Delivery options: https://9fans.topicbox.com/groups/9fans/subscription
