Question System shuts down after 10ish minutes once a day - works fine afterwards. Until next day.

May 7, 2023
7
0
10
Hey. I've found many threads that discuss similar topics, but the strangeness of this case is puzzling me.

System specs:
System
  • Intel core i7 6900 3,2 ghz
  • msi x99A sli plus (ms-7885)
  • intel X99 wellsburg-x
  • bios v. 1.D0
  • Psu: rm750i
Memory
  • 4 x 16gb Dimm Pc4 21300 DDR4
  • SDRAM G.Skill F4-2666C15-16GVR
Graphics card
  • Asus Rog Strix RTX 2080 Super Adv.

Case:
Everyday, once a day, I boot the system up, start a game such as bf2042. About 7-10 minutes later the light on the power button will flash violently, screen will go black and the system will go to "sleep" - the reason I say sleep is not because it actually goes to sleep, because I don't know, but because the light on the power button will stay lit after the crash has happened.

Then I press and hold the power button to shut it all the way down, and then turn it on again normally.

After this one strange crash(per day), the system is fine and plays bf2042 for hours without hiccups.

This will happen 95% of all days. I've had few days where I for some reason could avoid the crash but I can't say exactly what I did to cheat it.

I've also had one day where it happened twice within an hour.

Logs and dumps:
This is the critical error (see jpg) I think is the culprit but I'm a bit puzzled with the additional events after and during the crash?
jpg of the events
https://www.dropbox.com/s/26akoxcszyqspzw/waht_events.JPG?dl=0

Evtx dump of the events leading up to and beyond the crash
https://www.dropbox.com/s/ayqydr7yx001iyo/event_dump.evtx?dl=0

Mta dump - "local meta data" (created when i made the dump of the events[maybe some language stuff, i dont know if it is necessary]) ... (ignore)
https://www.dropbox.com/s/q31dchv4gzo0a31/event_dump_1033.MTA?dl=0
  • I've checked the minidump but it hasn't made one.
  • I've run OCCT test for half an hour on full stress. All 8 cpus go to 100C - Nothing happens.
  • I've had OCCT monitor running as a crash occur. Nothing peculiar is registered in the log.
  • I've set GPU max performance to 99% because of superstition. Nothing changes, crash still happens.
I can reproduce the crash once a day. 😕😕

To be honest I'm not so bothered by the actual crash. It's a thing I can come to live with. But. I'm very much afraid that it will evolve to a system breaking event.

Hope someone has experienced something similar and can help me.

Kind regards Andreas
 
Last edited:
I don't know if this is your problem but 100C is way too hot. What is the ambient temperature in the room when this happens?
yes i thought so too, but its while stress testing with OCCT, and computer doesnt seem to bother at all. Also, as mentioned, after the daily crash, everything works as expected. It also does before the crash... if it wasnt for the crash. Nothing seems out of the ordinary.
 
Memory instability maybe? 64G is a lot of RAM.
i would try the windows memory diagnostics, but i've had extremely bad history with that ancient tool hanging - so wondering if there is any other tool that could be recommended.
but i'm not sure why it would be the ram.
to me it sounds like some power issue.
i'm out of ideas how to fix it, which is why i turn to forums
 
i would try the windows memory diagnostics, but i've had extremely bad history with that ancient tool hanging - so wondering if there is any other tool that could be recommended.
but i'm not sure why it would be the ram.
to me it sounds like some power issue.
i'm out of ideas how to fix it, which is why i turn to forums
Well It is pretty easy to test if it is the RAM, just remove 3 of the 4 sticks for testing purposes and see if the issue persists. Just make sure you discharge any static by grounding yourself first and check in the motherboard manual (available online) as to which slot one stick of RAM should be inserted.
 
I've done an OCCT stress test AND playing bf2042 for 10 minutes. This is how the CPUs look according to OCCT.
https://www.dropbox.com/s/0ab94jbfwhukgap/whilegamingANDhavingstresstest.JPG?dl=0

This is after the daily crash, and i can do this the entire evening if i wanted to.

I'm not sure about OCCTs cpu temperature monitoring. Is there anyone with any opinions on the plausibility of the temperature monitoring? Seems to me, from reading alot of forums, that 95C would force my system to power down for safety. But it literally does not have a single hiccup.
 
OCCT stress test is way more stressful than normal computer use so I think you can safely say it is not a temperature issue. It could be a PSU problem but the test for that is to swap the PSU for a known good (usually new unless you have a newish spare) one. May as well eliminate the easy stuff first eh?
 
OCCT stress test is way more stressful than normal computer use so I think you can safely say it is not a temperature issue. It could be a PSU problem but the test for that is to swap the PSU for a known good (usually new unless you have a newish spare) one. May as well eliminate the easy stuff first eh?
yeah ok. totally. I'll see if i can borrow one from work maybe. Thanks. I'll be back