(Resolved) Random/intermittent shut down

[mrryan]

Thanks for the help all. I am going to class this issue as resolved for now as I'm pretty sure it is related to heat. CPU temps have always been fine but my guess is that there is some part of the motherboard getting hot.

GPU was idling at 50c (not clocking down due to running 2 monitors @ 144hz, seems to be a common thing) and with having seemingly nice temps everywhere else I had case fans set to 20%, increasing to 40% under load. This was adequate for my CPU but clearly not for some other component.

Since getting the Kraken I have only used CAM for monitoring but have now moved back to HWMonitor to get more info than just CPU/GPU temps.

I have monitors set to 120hz, shaving 17c off GPU temps and fans set to 50% min rising to 100% if things are getting seriously toasty. Hopefully that will keep everything under control. Only issue now is noise when not wearing headphones but it's still quieter than my previous build and covers up the annoying coil whine from GPU so I can't really complain.

Been having some issues for a while now and hit a dead end on where to turn.

PC shuts down mid use, not limited to specific usage (gaming/browsing/idling).
When trying to start it up again the fans spin up then it shuts down again. This can take between <1 to 5 seconds.
Usually it repeats this 3 times before staying off.

PC built nearly 9 months ago. Worked completely fine for three months before these issues. It appears to be working now but I am not convinced so am looking for suggestions on where to go next if (when) the issues return.

Here is what has happened so far:

- PC built.
- Worked for ~3 months.
- Reseated RAM.
- Reseated PSU.
- Replugged in all cables.
- PC started as normal, worked fine for a while before shut down again.

- Removed individual components, testing machine after each one removed (they were not put back in until after all had been removed)
- RAM/tried in different slots.
- GPU.
- HDD/SSD.
- Everything else (fans, fan hub, usb hub etc.)
- Same issues.

- Put everything back in - PC started, worked fine for a while before shut down.
- Installed different PSU that is working on another machine.
- Everything worked fine for a few months before shut down.
- Installed original PSU, unplugged everything from MOBO and set things up from scratch, including fresh windows install, new thermal paste, cleaning/dusting everything. Everything seemed to be working fine.

That brings us to today. PC shut down earlier and wouldn't reboot.
I tried plugging the PSU directly into wall socket - same issue.
I removed/reseated the CPU, checked all cables plugged in correctly and now it is running but I am doubtful as to how long.

I am thinking this is likely a motherboard issue?

Could it be a dodgy power button on case? Perhaps it is sticking/shorting and acting as though it is being held down?

Myself and partner both vape. Could vapour be possibly settling somewhere in the machine and causing a short? I have looked inside and can't see anything.

Any ideas would be most welcome!
Thanks.

Specs
CPU: i7-8700K
Cooler: NZXT Kraken X52
Mobo: Asus ROG Maximus X Code
RAM: 16GB G.Skill Trident Z RGB 3600MHz, 2x8GB
GPU: Asus GeForce GTX 1080Ti Strix
PSU: Corsair AX860i
NVME: Samsung 960EVO 250GB
SSD: Samsung 850EVO 1TB
HDD: Seagate BarraCuda 2TB
Case: Corsair Crystal 570X
Case/Rad fans: Corsair ML120 Pro x6
Additional: NZXT Internal USB hub, Corsair Commander Pro.

 

[LuckyBenski]

Firstly I'd check the temperatures. If your AIO pump has failed, the CPU will eventually overheat. It may not be obviously related to load.

Secondly after one of these events I'd check the Windows event viewer log to see what happened before shutdown.

 

[mrryan]

Hi Benski, thanks for the response.

I initially thought it might be temp related as after fiddling with things (during which everything will cool down) it seems to start again. Been keeping an eye on temps since then. Usually around 40c. Don't think I've really noticed higher than 60 unless I'm pushing things. I suppose things could potentially be getting very hot very quick so I don't notice anything. I can't say I've ever been looking at temps right when it shuts down.

Just had a shut down now again. Nothing in event viewer for 10 or so minutes before the warning about "rebooting without cleanly shutting down". It followed the usual process of shut down, automatic restart that fails. I pressed the switch myself and now everything appears fine again. (usually doesn't start back again so soon)

 

[theforce]

A PSU issue would be my first guess as well. Since you have tried another PSU with the issue persisting then the motherboard is the other most probable culprit. You can try moving the PC to another case if you have one laying around to make sure the case is not the cause of the issues before splashing cash for a new motherboard.

 

[Journey]

Have you tested running bear minimum hardware, and just using the onboard GPU and no overclocks on the CPU or RAM (run the RAM at 2133MHz) and just plug in one OS drive nothing extra. Get rid of any USB devices and unplug the front ports and headers etc. take the front I/O HDD LED/Power LED/PWR Switch etc off, and turn it on using something else to short the contacts.

Once it's booted, plug in the keyboard and mouse, and then mess around for a bit, make sure you have no network cable in, or wireless enabled etc.

 

[mrryan]

Thanks for the replies guys. So for an update...

The issue returned adter resetting cmos and leaving only m/kb, RAM, single drive and AiO connected. Machine shut down and refused to reboot. I removed everything leaving just the ATX and CPU cables plugged in. Same issue. Seemed to get to 16 or 20 on the q code display which indicates "pre memory system agent is started" or "pre memory pch is started".

I switched out the CPU cable from the one in use (corsair brand, designed for use with psu) with the original that was supplied with psu and it started working again. I tried the other cpu cable again and still working.

The difficulty in removing and reinserting the CPU cable has lead to me wonder if this issue could be caused if that cable was slightly loose? Or is it time to RMA the motherboard?

 

[EsaT]

Loose connector could certainly cause issues.
At worst it would lead to connector melting from heat generated by bad contact.

 

[mrryan]

Thanks for the help all. I am going to class this issue as resolved for now as I'm pretty sure it is related to heat. CPU temps have always been fine but my guess is that there is some part of the motherboard getting hot.

GPU was idling at 50c (not clocking down due to running 2 monitors @ 144hz, seems to be a common thing) and with having seemingly nice temps everywhere else I had case fans set to 20%, increasing to 40% under load. This was adequate for my CPU but clearly not for some other component.

Since getting the Kraken I have only used CAM for monitoring but have now moved back to HWMonitor to get more info than just CPU/GPU temps.

I have monitors set to 120hz, shaving 17c off GPU temps and fans set to 50% min rising to 100% if things are getting seriously toasty. Hopefully that will keep everything under control. Only issue now is noise when not wearing headphones but it's still quieter than my previous build and covers up the annoying coil whine from GPU so I can't really complain.