Question PC rebooting on idle - a real challenge

[misthic]

Hi, I'm encountering pretty annoying problem that I can't seem to resolve. I'm literally hopeless, so I hope you guys can do a miracle. Reboots sometimes are 5 seconds after logging to Windows, sometimes after 20-30 minutes. Background: I recently upgraded my system: - swapping out Gigabyte RTX 3070 Eagle OC for a Gainward 4080 Super - swapping out SilentiumPC Supremo FM2 Gold 750W PSU for a Fortron Hydro Ti PRO 1000W Titanium Other components: - Gigabyte Z690 GAMING X DDR4 - Intel i9-12900K - Crucial 32GB (4x8GB) 3600MHz CL16 Ballistix Black RGB Problem description: After replacing the GPU and PSU, and ensuring all cables were properly connected, my PC started randomly rebooting when idle. Here's what I've tried so far: - Clean installed GPU drivers using DDU. - Conducted tests on each component: CPU stress test (OCCT)​ GPU stress test (OCCT)​ MEMTEST ​ TM5 ​ test power stress test (OCCT)​ Intel Processor Diagnostic Tool​ - Ran sfc, dism, and chkdsk, and even reinstalled Windows (repair, not clean install). - Investigated Event Viewer logs, noticed Event 56 ACPI 2 errors before each reboot. Managed to resolve this (they are no longer appearing in Event Viewer) and a BitLocker error, but the problem with reboots persisted. - Updated BIOS to the latest version (F29), enabled "Power Loading" option in BIOS, and disabled C-States ("C-States" option to Disabled and second version - Enabled and blocked everything manually (C1E, C6/C7 State Support, C8 State Support, C10 State Support)). - Disabled power savings options for disks and PCI-Express in Windows Power Options. - Even tried to manuallny set vcore offset / vdroop in BIOS , but system didn't even boot and the VGA red LED was on (?). - Took off 2 RAMs, then other 2 RAMs, used different slots. - Trued using both: psu 12vhpwr <-> gpu 12vhpwr and psu 2x8 pin pcie -> adapter -> gpu 12vhpwr used his GPU in my PC - no restart (~1 hour)​ used mine GPU in his PC - no restart (~2 hours)​ - Switched slot for GPU, used PCIe 4 - reboot - Switched slot settings for GPU, set Gen3 for CPU PCIe Link Speed - didn't test enough since I'm half asleep, will test more tomorrow. UPDATE: 2 reboots on gen3. Interestingly, my PC never crashes during rendering things or/and gaming sessions (2-4 hours renders or gaming at demanding titles). It only occurs on idle periods and web/desktop activities. Despite all my efforts, the issue persists, and I'm running out of ideas. Searching throught internet I've found most of these issues were cause by CPU getting too low voltage when idle, but it didn't help in my case (or did I miss something?). Has anyone more ideas or possible solution? It's 6 days since I've started diagnosing PC and I still have no clue what is the issue.

 

[Ralston18]

Look in Reliability History/Monitor.

Much more end user friendly than Event Viewer and the timeline format may reveal some pattern(s) of events.

Another place to look is Update History: any failed or problem updates?

Also take a look in Task Scheduler for some action being triggered that results in a reboot.

 

[misthic]

Look in Reliability History/Monitor.

Much more end user friendly than Event Viewer and the timeline format may reveal some pattern(s) of events.

Another place to look is Update History: any failed or problem updates?

Also take a look in Task Scheduler for some action being triggered that results in a reboot.

I have no problem with looking through Event Viewer - anyways there are no errors except Kernel 41 (shutdown), therefore I don’t see any pattern.
No problems with update, everything up to date, even optional updates.
There is no consistency in rebooting issue - sometimes It’s after login into system, sometimes after 20 minutes.

 

[Ralston18]

Look beyond errors: i..e, warnings and informational events.

It may also prove helpful to clear the Event Viewer logs so new events are readily apparent and viewable.

The inconsistency suggests, to me anyway, a loose connection somewhere.

However. the problem could also just be the result of two simultaneous events. Either one harmless on its' own but when concurrent or closely concurrent then the reboot.

Reading back: "It only occurs on idle periods and web/desktop activities". Do any fans slow or stop?

Are those times when the PC would cool down and a connection could loosen or something contract and touch causing short/reboot?

Or perhaps trigger some other process via Task Scheduler. And that process causes the reboot.

In any case, I would expect the logs to capture such things...

Leaning towards a physical culprit - short or disconnect somewhere.

= = = =

Power down, unplug, open the case.

Clean out dirt and debris.

Use a bright flashlight and inspect everything for signs of damage: loose connections, bare conductor showing, melted insulation, kinked or pinched wires, cracks, scratches, browned or blackened areas, swollen components, loose or missing screws, components that move and wiggle where that should not happen.

Pay close attention to case connectors and the area (inside and out) around the I/O panel.

Check incoming cables and connections - fully in place and secure?

Intermittent problems are difficult to diagnose and you have been thorough in the troubleshooting.

Take another look at it all and hopefully something will be found.

 

[misthic]

I checked every cable one time after switching PSU and cables, but maybe I'll try one more time.
But I'm not sure it's about cables or some loose connectors - I think problem would occur on my PC with my friends GPU, right?
"Reading back: "It only occurs on idle periods and web/desktop activities". Do any fans slow or stop?" - really hard to tell, they barely move on idle, since everything is on low temps. I've tried to observe HWMonitor / HWInfo, but didn't notice anything suspicious.

 

[misthic]

Different PCIe / Gen3 setting didn't help.
I've used my friends GPU (3080) in my PC - no restarts.
He used my GPU in his CPU (pretty similiar build) - no restarts.

No idea what's going on...