Got a PC that’s a hodgepodge build of components from a few different upgrades over the last several years. It’s been pretty stable throughout that span, after I replaced a flaky power supply a couple of years ago. However, for the last couple of weeks I’ve had a recurring reboot issue.
Seemingly out of the blue, about once per day, while idling, the PC reboots itself and lands back on the login screen. There’s no instigating events I can find in Windows Event Viewer; just a “unexpected power off just occurred” complaint from after it’s done rebooting. In one instance, the PC died while it was processing some TRIM operations on hard drives late at night, but it hasn’t been doing that during the others as far as I can tell.
It only ever happens when the PC has been left idle for long enough for the monitors to go off – never during use, including during heavy use when I was streaming some games at 1440p awhile back. That usually translates to “after I go to bed, between 2 and 9 AM,” but it rebooted while I was taking a midday nap today, for instance, so, no real consistency there.
There haven’t been any major hardware changes leading up to these reboots, nor any major software changes that I’m aware of.
How the heck do I diagnose a thing that isn’t leave traces of its cause?
Oh damn, I didn’t even realize that was a thing! Nice. Okay. Feeling better prepared already.
The first few times this happened late at night, it was happening in the same nebulous “partner and I are both sleeping” timespan that our brand new Samsung microwave keeps turning on its over-stove light, and I was worried we were getting some kinda weird power surge that was killing the PC and fucking with the microwave but leaving clocks intact. Didn’t even occur to me after disproving that (when the PC rebooted not at the same time the microwave light turned itself on) that I should be able to see the damn bluescreens.
In my experience, IRQL_NOT_LESS_OR_EQUAL errors have always had bad RAM as the root issue, though of course there can be other causes. Personally I’d throw a memtest at it first before ripping apart my computer. Time consuming? Sure. But the physical effort is minimal.
^This right here - every time I’ve had random reboots it’s been one of two things: failing PSU or a bad stick of memory.
You can also pull a stick out and see if it happens again, and if so, put that back in and pull another and repeat until you’ve either ruled out the memory or found the culprit.
When I started getting random reboots a couple years ago it turned out to be a secondary spinny hard drive that was about to fail. Might want to run a check just to eliminate that from the list of possibilities.
I used CrystalDiskInfo to diagnose, but there are probably dozens or hundreds of similar tools.
Maybe the faulty drive messed up 12v power delivery or something when it progressively kicked the bucket. That could do it. I wouldn’t think 3.3v would cause reboots. Maybe 5v.
I mean I do have some spinning disc hard drives from the early 00s in here. I actually disabled one in windows cuz it was falling and causing long hangs when anything tried to access it.
One other piece of weird behavior. Netflix and twitch videos occasionally hang the video for 5 seconds or so while audio continues, then skip forward to catch up. I figured that maybe having 80 open tabs across 3 browsers was finally catching up with my 16gb of ram, but even knocking that down to 20 or less, it’s still happening, and has been going on a similar amount of time to the reboots.
Speaking of which, past 24 hours now without one after doing a few of the software fixes suggested in one of the links above. Man it would be super neat if that were it. I really, really, really don’t wanna replace a power supply, ugh.