PC won't wake from sleep - random restarts without BSOD in some games

[More Sly]

Hey folks, I have an issue I've posted about previously that's ongoing and I'd like some advice if anyone's willing.

After installing a and Ryzen 7 5800x and B550 Aorus Elite V2 in August with a fresh Windows 11 install, my computer was unable to wake from sleep. On wake, the fans spin at what sounds like 100% but it never actually wake up. If I force a shutdown and start it back up manually it restarts as if waking. I've been in touch with Gigabyte's support and am slowly troubleshooting, but I just disabled sleep and figured I'd deal with it slowly. I've reset my BIOS with the B550's CLR_BIOS pins and tried a new installation of windows on a new drive - the issue persists. Next step is flashing to the new 15d firmware (I'm currently on 15c).

But another issue has come up and I'm not sure if it's related. In some games I'm getting random restarts. No BSOD, just a sudden restart and a kernel power error in the events log. This has consistently occurred in one game (Oxygen Not Included) but has only ever happened once in one other game (No Man's Sky).

My biggest suspicion is still that this is a faulty board with power delivery issues, but I'm exploring other options before I can RMA. The restart issue lessened when in-game actions in ONI reduced the memory load, so I wonder if this might be a RAM compatibility problem - I have 2x8 GB G.Skill DDR4-3000 DIMM CL16-18-18-38. BIOS reports both sticks running at 2133 MHz and enabling XMP gives me pretty much immediate BSODs.

I'm also wondering if this could be considered erratic CPU behavior? CPU stress tests don't cause the issue - I've run 3DMark and Cinebench with no problems - and I think this is probably just a little residual paranoia about installing an AM4 CPU since it was my first time plugging in a PGA chip. I've had a bit of high heat in very CPU intensive games where it's had odd peaks at 91 °C, but it was otherwise stable in safe ranges with an average temperature around the mid 60's to maybe 70 ° during long sessions.

I've attached a HWiNFO log leading up to one of the restarts in case that might suggest anything, the event log, DxDiag, and a log file from ONI (just in case). Any thoughts would be greatly appreciated!

 

[_JP_]

Howdy!

I never came to any conclusion, but I had the exact same behavior on my desktop, back until I last used it, ever since it upgraded Windows to 1803. I'm at awe that anybody managed to have the same issue.
Not that I have the same configuration you do, of course, it was a R7 1700 + X370 Gaming Plus + 2x8GB GSkill Trident Z 3200MHz (XMP Profile) (Hynix chips) and for discreet cards, there was a R9 280X and an Asus Xonar DX. PSU was a Seasonic-built 650W, 3 SSDs all in very working order (health between 95~100%).
Later on, already with the upgrade to 1909, 2004 and 20H2 the issues persisted, GPU was already a RX5500XT and the sound card was different.
I could not launch a damn game that wouldn't BSoD after a random amount of time, but just sitting idle would randomly BSoD-out too.
If the desktop went into S3, taking it out of it would take some 7 to 10 minutes, fans low but then sped-up until it rebooted all by itself like it was cold-booted.

This was very peculiar because BSoDs always pointed at Kernel and the AMD Drivers, but this spanned several driver versions and most boggling, worked 100% fine in versions 1703 and 1709 of Windows 10.
Eventually, I stopped having time to troubleshoot it, the computer was just so unstable I stopped using it.

 

[eidairaman1]

Breadboard the system and then get hwinfo64 screenshots of power supply voltages

Howdy!

I never came to any conclusion, but I had the exact same behavior on my desktop, back until I last used it, ever since it upgraded Windows to 1803. I'm at awe that anybody managed to have the same issue.
Not that I have the same configuration you do, of course, it was a R7 1700 + X370 Gaming Plus + 2x8GB GSkill Trident Z 3200MHz (XMP Profile) (Hynix chips) and for discreet cards, there was a R9 280X and an Asus Xonar DX. PSU was a Seasonic-built 650W, 3 SSDs all in very working order (health between 95~100%).
Later on, already with the upgrade to 1909, 2004 and 20H2 the issues persisted, GPU was already a RX5500XT and the sound card was different.
I could not launch a damn game that wouldn't BSoD after a random amount of time, but just sitting idle would randomly BSoD-out too.
If the desktop went into S3, taking it out of it would take some 7 to 10 minutes, fans low but then sped-up until it rebooted all by itself like it was cold-booted.

This was very peculiar because BSoDs always pointed at Kernel and the AMD Drivers, but this spanned several driver versions and most boggling, worked 100% fine in versions 1703 and 1709 of Windows 10.
Eventually, I stopped having time to troubleshoot it, the computer was just so unstable I stopped using it.

Sounds like MS broke the OS lol

 

[More Sly]

Breadboard the system and then get hwinfo64 screenshots of power supply voltages

So with my GPU and a monitor since I don't have integrated graphics, yeah? I'll do that this weekend after flashing the BIOS to 15d (was instructed to by Gigabyte but I don't think it'll do anything).

 

[More Sly]

I could not launch a damn game that wouldn't BSoD after a random amount of time, but just sitting idle would randomly BSoD-out too.
If the desktop went into S3, taking it out of it would take some 7 to 10 minutes, fans low but then sped-up until it rebooted all by itself like it was cold-booted.

This was very peculiar because BSoDs always pointed at Kernel and the AMD Drivers, but this spanned several driver versions and most boggling, worked 100% fine in versions 1703 and 1709 of Windows 10.
Eventually, I stopped having time to troubleshoot it, the computer was just so unstable I stopped using it.

Fortunately my issue isn't quite this bad! It's still usable, it's just the sleep issue and one game getting random restarts. Very frustrating, but not ready for the trash heap. Fortunately, the issue started with a mobo and CPU replacement so it should be fixable replacing one of those components, or just switching back to Intel. Which I'd rather not do - total waste of component money.

I'm convinced it's a motherboard issue, so if breadboarding doesn't point me to a fix and I end up with an RMA I might buy a new mobo to test in the meantime. Terrified of bending pins while messing with this damn PGA chip, though...

 

[eidairaman1]

Fortunately my issue isn't quite this bad! It's still usable, it's just the sleep issue and one game getting random restarts. Very frustrating, but not ready for the trash heap. Fortunately, the issue started with a mobo and CPU replacement so it should be fixable replacing one of those components, or just switching back to Intel. Which I'd rather not do - total waste of component money.

I'm convinced it's a motherboard issue, so if breadboarding doesn't point me to a fix and I end up with an RMA I might buy a new mobo to test in the meantime. Terrified of bending pins while messing with this damn PGA chip, though...

Warm the cpu up then gently twist the heatsink. Idk LGA pins are more difficult to fix than PGA.

 

[More Sly]

Warm the cpu up then gently twist the heatsink. Idk LGA pins are more difficult to fix than PGA.

Thanks for the reminder! I'd forgotten this and probably definitely would have hurt something. Don't have the time to breadboard this weekend so I'll have to try over the holidays - BIOS update didn't work.

 

[More Sly]

I've finally had time to take my PC apart and breadboard. Breadboarding has fixed my sleep issue, thankfully, but the the random restarts continue; though it appears to be less frequent. HWiNFO of the last restart to compare.

I'd only had this happen in a single fairly low-resource game (Oxygen Not Included) when I made this topic, but I've since had random restarts in other games on two occasions. It's rare enough in other titles that I probably wouldn't have thought anything of it were it not for the one game where restarts are more consistent.

Thoughts? Should I RMA the mobo?

 

[Vya Domus]

I have 2x8 GB G.Skill DDR4-3000 DIMM CL16-18-18-38. BIOS reports both sticks running at 2133 MHz and enabling XMP gives me pretty much immediate BSODs.

That's very odd, there shouldn't be any problem running at it's rated speeds, 5000 series should not be that finicky about memory, I can't say for sure but I am pretty confident there's a problem with the RAM sticks.

 

[More Sly]

I can't say for sure but I am pretty confident there's a problem with the RAM sticks.

The game that crashes most does have a bad reputation for memory leaking. I might have to cough up to test some new RAM - memory is the oldest component on this build at about 5 years olds anyways.

 

[Vya Domus]

The game that crashes most does have a bad reputation for memory leaking. I might have to cough up to test some new RAM - memory is the oldest component on this build at about 5 years olds anyways.

It shouldn't be related to memory leaks, even with that the system should not crash, it's just a matter of instability. After the RAM it might be the board and least likely the CPU.

 

[More Sly]

It shouldn't be related to memory leaks, even with that the system should not crash, it's just a matter of instability. After the RAM it might be the board and least likely the CPU.

Unfortunate. Guess I'll grab some new memory and hope I don't have to deal with RMAs. Thanks for the help!

Breadboard the system and then get hwinfo64 screenshots of power supply voltages

@eidairaman1 the log above should have those PSU voltages, right? I'm not totally clear on what you were suggesting I screenshot.

 

[Braegnok]

Hey folks, I have an issue I've posted about previously that's ongoing and I'd like some advice if anyone's willing.

After installing a and Ryzen 7 5800x and B550 Aorus Elite V2 in August with a fresh Windows 11 install, my computer was unable to wake from sleep. On wake, the fans spin at what sounds like 100% but it never actually wake up. If I force a shutdown and start it back up manually it restarts as if waking. I've been in touch with Gigabyte's support and am slowly troubleshooting, but I just disabled sleep and figured I'd deal with it slowly. I've reset my BIOS with the B550's CLR_BIOS pins and tried a new installation of windows on a new drive - the issue persists. Next step is flashing to the new 15d firmware (I'm currently on 15c).

But another issue has come up and I'm not sure if it's related. In some games I'm getting random restarts. No BSOD, just a sudden restart and a kernel power error in the events log. This has consistently occurred in one game (Oxygen Not Included) but has only ever happened once in one other game (No Man's Sky).

My biggest suspicion is still that this is a faulty board with power delivery issues, but I'm exploring other options before I can RMA. The restart issue lessened when in-game actions in ONI reduced the memory load, so I wonder if this might be a RAM compatibility problem - I have 2x8 GB G.Skill DDR4-3000 DIMM CL16-18-18-38. BIOS reports both sticks running at 2133 MHz and enabling XMP gives me pretty much immediate BSODs.

I'm also wondering if this could be considered erratic CPU behavior? CPU stress tests don't cause the issue - I've run 3DMark and Cinebench with no problems - and I think this is probably just a little residual paranoia about installing an AM4 CPU since it was my first time plugging in a PGA chip. I've had a bit of high heat in very CPU intensive games where it's had odd peaks at 91 °C, but it was otherwise stable in safe ranges with an average temperature around the mid 60's to maybe 70 ° during long sessions.

I've attached a HWiNFO log leading up to one of the restarts in case that might suggest anything, the event log, DxDiag, and a log file from ONI (just in case). Any thoughts would be greatly appreciated!

Issue of random restarts, no BSOD,.. like pushing restart button.

Try adjusting your PBO +6 all-cores,.. and see if that solves issue.

 

[freeagent]

It’s your memory, system does not like it.

 

[More Sly]

It’s your memory, system does not like it.

Thanks for the help. I'm ordering some new memory to test now but you sound confident - would you mind explaining why? Would just love to better understand.

 

[freeagent]

Thanks for the help. I'm ordering some new memory to test now but you sound confident - would you mind explaining why? Would just love to better understand.

enabling XMP gives me pretty much immediate BSODs.

This was the big tipoff. But when I was fresh to AM4 I was trying all kinds of memory configs that would act as you described. That's when I quit using that dram calculator

You could try feeling a little more vdimm, maybe 1.375v, but.. it should run. I have some crap Adatas that are 16-20-20-38 but they run ok. Maybe try stepping up to those timings to see if it helps?

Sorry I am still a memory noob, I only know my sticks which are b-die. My Adatas are not though, they are in my sons system.

 

[More Sly]

I didn't realize AM4 could be finicky with memory! I just ordered a TridentZ 16-19-19-39 kit. I don't know much at all about memory, so nothing to apologize for. I appreciate the help.

Reviews say TridentZ's recent kits are good for AMD boards, though I wouldn't know, so I figure I'll stick with them for now and send them back to try another if the problem persists.

 

[RJARRRPCGP]

just a sudden restart and a kernel power error in the events log.

More likely a PSU fault or motherboard power delivery fault. There is usually a crash report that points to memory, if the RAM is the issue.

 

[eidairaman1]

I didn't realize AM4 could be finicky with memory! I just ordered a TridentZ 16-19-19-39 kit. I don't know much at all about memory, so nothing to apologize for. I appreciate the help.

Reviews say TridentZ's recent kits are good for AMD boards, though I wouldn't know, so I figure I'll stick with them for now and send them back to try another if the problem persists.

Crucial Ballistix Gaming 3600 on an AsRock B550 Steel Legend with Ry7 5800 runs flawless with DOCP (XMP)

More likely a PSU fault or motherboard power delivery fault. There is usually a crash report that points to memory, if the RAM is the issue.

Yup BSODs start flying

 

[More Sly]

More likely a PSU fault or motherboard power delivery fault. There is usually a crash report that points to memory, if the RAM is the issue.

The motherboard has been my main suspicion. PSU is fairly new and worked well on the Intel rig it was previously on, but I'm going to test a spare PSU tomorrow and come back.

New memory seems worth a try since I was hovering around upgrading anyways. Would be nice to have XMP working. I think I'll try this new memory and then RMA the board if the problem persists.

I've attached the ErrorLog from the same crash I previously shared the Kernel error for.

 

[RJARRRPCGP]

I suspect that kernel power error is for Windows not being cleanly shut down.

That's what happens to me, when I get the Windows 11 infamous "blank screen with monitor backlight still lit" crashes. (Have to hold down power button)

 

[Vya Domus]

You can try memtest86 and see if that throws any errors.

 

[freeagent]

I suspect that kernel power error is for Windows not being cleanly shut down.

That's what happens to me, when I get the Windows 11 infamous "blank screen with monitor backlight still lit" crashes. (Have to hold down power button)

Sell that stuff and get some B-Die if you can find it. I got mine Nov. 2020 and have not had any problems like that. Just smooth sailing this whole time

 

[eidairaman1]

I suspect that kernel power error is for Windows not being cleanly shut down.

That's what happens to me, when I get the Windows 11 infamous "blank screen with monitor backlight still lit" crashes. (Have to hold down power button)

Hmm never had that issue, then again i disabled WUD altogether

 

[More Sly]

Issue of random restarts, no BSOD,.. like pushing restart button.

Try adjusting your PBO +6 all-cores,.. and see if that solves issue.

View attachment 279140

It's possible this worked. I saw a little barely perceptible screen flickering, but no restarts so far. Yesterday restarts were already pretty infrequent, though, so I'll have to test a lot longer.