Weird restarting problem - collective wisdom, please help?

champ222

champ222

champ222

Associate
Joined
19 Aug 2005
Posts
1,569
Location
Beds, UK
Hi All

I have a problem with my PC, i'll try and brief, but i have lots of detail:

Upgraded the PC around the start of Feb. All was running perfectly for two months, used most evenings (by far the most use a PC of mine has ever had) until 2nd April, when the PC randomly restarted while playing Warzone. I had been playing for approx 20 minutes. Then it did it again 10 minutes later.

Prime 95 seemed to all work fine. ran a GPU test "Unigine Superpostion" PC restarted as it started. I feared the worst - GPU issue. Updated the Nvidia drivers. All seemed well, superposition ran, warzone ran. Warzone ran even with prime95 running in the background (forgot to stop it).

All was well until 7th April.. after 1.5 hours play, PC restarted again. Ran unigine superposition again, it ran fine in various modes. Back into warzone, instant restart when the match started. Went to bed i think.

8th April:
Ran furmark - fine
Ran memtest - 2 passes, no issues
Ran prime 95 for a bit - no issues

I didnt run these for very long as i didnt have that much time as i wanted to get windows rebuilt and warzone downloading before the kids got back home. i ran them enough to get the temps up a to about where they seemed to top out.

Updated mobo BIOS, checked ASUS GPU BIOS (tool said no update required), Reinstalled Windows, installed chipset driver from MSI website and Nvidia driver, ran the windows updates. installed Warzone. All was fine gaming that evening.

i didnt play 9th or 10th.

11th April (last night) played, 1 hr 45 mins all fine, then it restarts again, i rebooted, restarts as the match starts.

it was pretty late already, but i tried a few things and it seemed to get worse and worse to the point that it'll load to the windows password screen, i type the password, hit enter - restart.

Sometimes i got no display at all. Although once i plugged the monitor in via HDMI (it doesnt show the splay screen/bios etc over displayport, which is awkward enough) it seemed to be booting reliably, but would restart when asked to do pretty much anything. By this point, the lid was off the case and everything was cold to the touch.

i ran with only one ram stick, then the other, no difference.

Specs and date purchased:
Intel i5 10600KF - 3rd Feb 2021
MSI MEG Z490I Unify mini ITX motherboard - 3rd Feb 2021
WD Blue SN550 1TB - 3rd Feb 2021
Crucial MX500 M.2 Sata 500GB - May 2018
Corsair 16GB (2x8) Corsair Vengeance LPX 3200 - April 2017
Asus Strix GTX 1070 - was a warranty replacement that was supplied May 2017
Seasonic X series 750watt PSU - bought used from here in may 2017, but it may have been made in 2013
Corsair 16GB (2x8) Corsair Vengeance LPX 3200 - April 2017
WD Blue SN550 1TB - 3rd Feb 2021
New monitor - Iiyama Gmaster 2470 - 15th Feb 2021


I can try swapping out the SSD. the nvme is on the back of the motherboard, as MSI (and others) make the 2nd m.2 nmve only, and the 1st m.2 which is under the heatsink, is the only one that will run the sata drive.

I have a 2.5" SSD i can use testing. I have another PSU, its a good one, but its even older (corsair HX620) i have others at work i could probably borrow. they are decent PSUs but only 400w or so.

Ive reached out to a friend who might have some bits i could swap about, but ive not heard back yet.

If its the GPU, i'm out of gaming then. i cant even see how much they are supposed to cost as there is no prices on the websites etc... you lot know the score.

Sorry for the information overload. Any thoughts or ideas?

Thanks
 
Thanks

The case is a "Fractal Design Core 500 M-ITX"

The cooler is a "Thermalright AMD/Intel Macho Rev.B Multi CPU Quiet Air Cooler"

in the case, the motherboard is horizontal, which means the CPU cooler is sitting straight up, and so not "hanging" off the board. The GPU sits upright also, there is a vent all down the side of the case, right next to the GPU. the GPU never goes above 65 degrees C while gaming
 
mickyflinn said:
M2 drive see if theres a firmware updates.
Click to expand...
Thats a good shout actually.

I'm a little suspicious of the M.2, i tested the speed with crystal disk mark and i was getting speeds just over SATA speeds approx 800MB/s read. when i tested it again the following day, i was getting the typical 2400 MB/s read speed. i'll check the WD website for updates
 
i did run HDDTune, which reported a high temp for the Sata SSD of 60 something degrees, there is nothing on this SSD at the moment, it couldnt get a temperature from the NVME, though i dont know how up to date the free version of HDD tune is
 
So i fired it up tonight and it was the same as yesterday. Couldn't really do anything. Built it outside the case, and its working better. No restarts yet. Managed to download the WD dashboard and updated the ssd firmware.

Currently running furmark 7 minutes in, so far no issues... will keep testing
 
Ran furmark for 15 mins. No restarts.

Running prime95 now. I did small ffts for 10 mins or so, now running blend, been 10 mins maybe. Still no restarts.

Ill try some more tests, but im wondering what difference taking it out the case made...

Ive unplugged the usb 3 header for the case, and the audio header and a case fan, but thats it... i could plug them back in and see if it starts acting up again
 
Well... after an evening and a late night of faffing, i'm not really sure what i have learnt.

Initially, it was resetting every few minutes, or seconds, i could barely do anything. then once i took it out the case, all seemed well. I did update the ssd firmware. and ran a bunch of tests. I only had one restart the whole evening. that occured when trying to stress the CPU and GPU at the same time. i was also moving a cable at the time.

So maybe its a PSU issue, since CPU test or a GPU test seem fine, but both caused an issue (which would for gaming) but that doesnt explain the restarts when i first switched it on, as there was no stress, no heat. It could be a dodgy cable i suppose, but i did move the cables about, and i couldnt reproduce it.

It still could be a GPU issue i guess, while gaming, its it fairly constant 100% use for hours. i didnt replicate this with the testing since the superposition doesnt have a loop option (i'd need to pay for that i guess). I could use furmark, but do people still use this? could that explain the restarting issue from cold though? no heat in the system at that point.

the CPU hit 80 degrees max during testing (prime95 small ffts) its about 75 degrees on blend i think. GPU seems to be below 65 degrees. Feeling around the board, nothing seems overly hot (though you cant easily get to some of the heatsinks in a mini-itx when using a massive cpu cooler) the heat-sink that goes over the SSD was quite warm, which is odd as the sata ssd was in there, with nothing on it. though the heatsink does appear to be in contact with the other heatsinks, so maybe its getting some heat from there.

Ive been running optimised defaults for a while now, so ram is running at 2666. Everything is on auto, i havent run any overclocks. Perhaps there are some settings i can manually set to improve stability?

In the end, i took the sata SSD out, moved the NVME to the slot on the top and rebuilt it back into the case. i left off the USB 3 header cable and the hd audio header, since i'm not using those anyway. I might see if have any more cables for the PSU --> GPU and the PSU -->CPU. i left the vent filters out to help air flow a bit.

i set the fans up in bios. Though all the fans were chosen, some builds ago, for silence, rather than performance and are rated at about 50 CFM ish. They never seem to run at full speed though, so i'm not sure if there is any benefit of getting some that create more air flow, since they are temperature controlled anyway?

I'll do some gaming again and see what happens. I was hoping to find a test that reliably causes it to fail. Running Prime95 (changed to low priority) with superposition runs the CPU at about 100% and the GPU ran about 30-40%, though there was a few occasions when it ran up at 80-90%, but i couldnt fathom why that could be, perhaps at that moment the tests run by prime95 were lighter? not sure if it goes through a suite of test one after another, or if it is constant.

Running prime95 and furmark does run both the CPU and GPU close to 100%. it was 2am at this point though and i needed to get to bed.

So i'm not sure what ive learnt really, if anything.
 
Yeah, i need to do more testing really. Ive been a light/occasional gamer forever really, but i like that experience to be as trouble free as possible. Its only recently that ive been playing online with my mates (i'm 42 now, so pretty late to the game), and its a good laugh. So it prompted the CPU/mobo upgrade and a new high refresh monitor.

I cant change the GPU at the moment, so i either need to find a fix, or find a way to deal with the restarting once it starts. It'll be fine for a while, but once it starts restarting, it seemingly gets worse until, this time at least, took it out the case, then its all seemingly fine again. i cant do that everytime, so i need to find what fixes it. Underclocking would be ok for me if it fixed the issue, as id happily drop a few fps to improve stability. If its a CPU, mobo or a ram issue though, those a still under warranty.

Its early days still so i'm hoping i'll figure it out. My gaming time is short though, with work and kids etc, so its frustrating when the restarts persist to the point where i cant play.
 
Moeks said:
Did you have a look in the even viewer to see if it recorded anything prior to the last reset ?
Click to expand...
no, i should have done. i did look in there before, but i just found the kernel 41 errors which as i understand it, are the result of the restart rather than the cause. If/when it does it again, i'll check in Windows Logs, Application and System sections.
 
Emjeadea said:
@champ222 did you ever get to the bottom of this issue? i have the same thing and it works okay most of the time when the pc is under a little load from gaming ?!
Click to expand...
Yes. It turned out to be the psu. Perhaps a dry joint or a broken cable or something. Looking back at it now, it makes sense. It would work ok until the temp of the psu/system was just right that it failed. Then it would constantly fail, and rebuilding out the case fixed it for a time... all good now the psu has been replaced. Got a good deal on a good psu too which is always nice.
 
You must log in or register to reply here.
Back
Top Bottom