Question Diagnosing Random Shutdowns

[cdriccio]

Hello all.
My PC started randomly shutting down recently and I'm trying to diagnose the issue.
I found a good deal on a prebuilt a little less than a year ago, and use it for day to day use and gaming.
Here's the PC:

https://www.newegg.com/abs-ali634/p/N82E16883360299?Item=N82E16883360299

High level details:

• Operating System: Windows 10 Home (have not upgrade to 11 yet... not sure if I should)
• CPU: Intel Core i9-11900KF Rocket Lake 3.5GHz (5.3GHz Turbo) 8-Core 16-Thread
• GPU: MSI NVIDIA GeForce RTX 3080 Ti 12GB
• Motherboard: Z590 Motherboard
• RAM: OLOy Blade 16GB DDR4 3200MHz (2 x 8GB) (upped to 32)
• SSD: 1TB M.2 NVMe SSD
• CPU Cooler: 240MM RGB AIO Cooler (closest I can find to it: https://www.newegg.com/gamdias-liquid-cooling-system/p/N82E16835099012?Item=9SIAJGNJ2B9758)
• PSU: 750W 80 Plus Gold
• Case: Rosewill SPECTRA D100

I'm using CoreTemp to watch the temperature of the CPU and it can get pretty high while gaming. I did see it hit 100 degrees.
Last night I was playing BG3 and it was running in the 90s.
So I emailed the company and told them I thought the cooler was not doing it's job. They are sending me a new one to replace. I do not know the brand of the cooler, so that is a concern. I linked the closest thing I could find, but the cooler has the ABS brand on it and I can't find it for sale anywhere. Newegg has 1 "open box" ABS cooler and says it's manufactured by Gamdias.

However... my other concern is that today, the PC shut down while NOT at high temps. CoreTemp logs showed this as the last entry for the temps:

Time	Core 0 Temp. (°)	Core 1 Temp. (°)	Core 2 Temp. (°)	Core 3 Temp. (°)	Core 4 Temp. (°)	Core 5 Temp. (°)	Core 6 Temp. (°)	Core 7 Temp. (°)
07:46:52 09/09/23	48	48	58	48	47	50	49	45

The PC powers down, waits a few seconds, and then starts back up normally.
I did download HWiNFO64 as well, so I'm looking for other possible issues that may cause the PC to shut down other than cpu temp.

Does anyone have any advise on where to go from here?

Thanks!

 

[cdriccio]

Some additional information...
the GPU seems to be running hot as well. Could it be an overall cooling problem?

With HWiNFO, it says that my GPU's thermal limit is 83 degrees.
However I see a max temps of 82.4, 92, and 98, for GPU Temp, Memory Junction Temp, and Hot spot Temp respectively.

 

[jkevery]

I recently upgraded a GTX 1070 (wattage ~ 150) to a RTX 3070 (wattage 240 @ full load per HWMonitor) and was experiencing random restarts during hard gaming. Powering it with an older Corsair HX-850 so I figured I was good. CPU and GPU temps were hitting upper limits, but I think that modern CPUs throttle themselves to limit thermal overload and I wasn't having screen freezing that I associate with GPU crashes. After I was out of ideas I replaced the old PSU with identical new unit and that solved my problem. Your 3080 Ti can pull 350 watts. Something to think about.

 

[Kona45primo]

Does it randomly restart in games only or just on desktop as well?

Did buy the system with 16GB of memory and add 16GB later or did it all come with 32GB from the factory?

Do the temps go down significantly if the side panel is removed?

Try using HWiNFO64 for your temp monitoring. It should give you a better overall picture of temps everywhere.

If the computer is stable at idle & desktop you may want to consider updating your BIOS, chipset and GPU drivers if they are old.

 

[cdriccio]

Thank you both so much for the replies.

I recently upgraded a GTX 1070 (wattage ~ 150) to a RTX 3070 (wattage 240 @ full load per HWMonitor) and was experiencing random restarts during hard gaming. Powering it with an older Corsair HX-850 so I figured I was good. CPU and GPU temps were hitting upper limits, but I think that modern CPUs throttle themselves to limit thermal overload and I wasn't having screen freezing that I associate with GPU crashes. After I was out of ideas I replaced the old PSU with identical new unit and that solved my problem. Your 3080 Ti can pull 350 watts. Something to think about.

Understandable and good point. This is a recent problem though and I've had the PC almost a year. I've played other high end games on it that have ramped up the power consumption. Though BG3 is the most intensive. That said, it powers down even when I'm not using the pc. Today it powered down before I even logged into it.

Does it randomly restart in games only or just on desktop as well?

Just on the desktop as well.

Did buy the system with 16GB of memory and add 16GB later or did it all come with 32GB from the factory?

I added the additional 16G immediately upon getting the rig. So it's been in there for 11+ months. One concern I've always had about it is that the tubes for the AIO infringe on the ram stick closest to it. If this connection is not stable, maybe that's doing it?

Do the temps go down significantly if the side panel is removed?

I have not tried with it removed, but I opened it slightly to get better ventilation and I don't believe I saw any chagnes.

Try using HWiNFO64 for your temp monitoring. It should give you a better overall picture of temps everywhere.

I am using HWiNFO64, that's what I used to post the GPU temps above.

If the computer is stable at idle & desktop you may want to consider updating your BIOS, chipset and GPU drivers if they are old.

Yeah as mentioned above, it is not. Shuts down at desktop as well. I've watched it just shut down while I was sitting on the couch not using it. That said... GPU drivers are updated. BIOS and chipset are likely not. I'll look into that.

Tonight I will see about testing out:
BIOS / chipset updates
Temps with side panel completely off
Removal of added RAM sticks

 

[MEMOFLEX]

What is the make, model and age of your PSU?

 

[Kona45primo]

You may want to check stability with A1 & B1 ram sticks removed to see if it's a memory related crash.

Also go into command prompt (As Admin) and type sfc /scannow to check for any errors

If it's randomly unstable a bios or chipset update may be a little dangerous, as you don't want to risk a restart in the middle.

 

[cdriccio]

What is the make, model and age of your PSU?

DQ750-M-V2L - DeepCool

The DQ750M-V2L fully modular 750W 80 PLUS Gold power supply features 100% Japanese capacitors and high quality switch MOSFETs. Built to deliver stable output capacity with high conversion efficiency, the DQ750-M-V2L offers better reliability for your PC system. - DeepCool was founded with the...

www.deepcool.com

PC is just less than a year old.

You may want to check stability with A1 & B1 ram sticks removed to see if it's a memory related crash.

Also go into command prompt (As Admin) and type sfc /scannow to check for any errors

If it's randomly unstable a bios or chipset update may be a little dangerous, as you don't want to risk a restart in the middle.

I ran it and it came back with:
"Windows Resource Protection found corrupt files and successfully repaired them."
I looked in the log and I see repair to a OneDrive.lnk file twice and some of these...
Warning: Overlap: Directory ... is owned twice or has it security set twice

 

[jkevery]

Has anything helped or are you still having the same issue?

 

[cdriccio]

Has anything helped or are you still having the same issue?

Over the past few days, I've seen less shutting down, but it did it again overnight last night it seems.
I moved it to a more open area (it was essentially between 2 desks and now it's on a table) and I've been running it with the side panel off and it's definitely running cooler.
I got the new AIO from the manufacturer.

 

[jkevery]

I hope the new AIO does it. The fact that it’s happening at idle and minimal load, though, suggests that it’s not temperature related. If this continues with the new parts and your other efforts, I think you’d have to suspect a bad PSU. Tom’s review of the DQ-M V2L 850W wasn’t particularly good - especially the Bottom Line. Not trying to be a troll and not saying don’t try everything else. But when you’re having random shutdowns and nothing else works it’s time to try another PSU. I keep a spare RM750 and a GTX 1070 in the attic as I’m running two gaming rigs and sometimes the only way to verify that a component is bad is to replace it with a known working unit.

 

[cdriccio]

I hope the new AIO does it. The fact that it’s happening at idle and minimal load, though, suggests that it’s not temperature related. If this continues with the new parts and your other efforts, I think you’d have to suspect a bad PSU. Tom’s review of the DQ-M V2L 850W wasn’t particularly good - especially the Bottom Line. Not trying to be a troll and not saying don’t try everything else. But when you’re having random shutdowns and nothing else works it’s time to try another PSU. I keep a spare RM750 and a GTX 1070 in the attic as I’m running two gaming rigs and sometimes the only way to verify that a component is bad is to replace it with a known working unit.

Yeah I know the AIO isn't the problem. I'm also beginning to suspect PSU. It's just odd how it ha been mostly random. I have a spare psu laying around too... but it's rather old and I think I'd just rather get a newer better one. I guess now I'm debating sending it back... seems like a huge hassle to pack it up, ship it back and be without it for who knows how long. Warranty is up in less than 2 weeks... so I need to make a decision soon.

Any suggestions on PSU for that rig?

 

[jkevery]

Seasonic Focus GM-750 is on sale at Newegg for $100 (4,200+ reviews), $20 more will buy a Corsair RM 750x. Yea, I hate returning things on warranty - especially for something that happens only intermittently. But, since it is under warranty you shouldn't have to spend $ to fix it.

 

[cdriccio]

Thoughts on this one:

https://www.newegg.com/p/pl?N=4841%2050001697%2050001459%20600112163%20100007657&d=850w+power+supply

Seems like a good price for an 850 Platinum.

 

[vertuallinsanity]

Don't see any details on the mobo..

Does it support overclocking? Have you modified anything in bios?

First step for those issues is to reset the BIOS to default and see if the problem continues. Same if you've made any GPU changes. Reset to default clocks.

 

[jkevery]

I don't think you can go wrong with a Corsair RM or HX series. My original HX lasted 10+ years and three rebuilds.

 

[Zerk2012]

From the reviews looks like a common problem.

https://www.newegg.com/abs-ali634/p/N82E16883360299?Item=N82E16883360299&nm_mc=AFC-RAN-COM&cm_mmc=afc-ran-com-_-Future+Publishing+Ltd&utm_medium=affiliate&utm_campaign=afc-ran-com-_-Future+Publishing+Ltd&utm_source=afc-Future+Publishing+Ltd&AFFID=2294204&AFFNAME=Future+Publishing+Ltd&ACRID=1&ASUBID=tomshardware-us-8240642821978208000&ASID=https%3A%2F%2Fforums.tomshardware.com%2Fthreads%2Fdiagnosing-random-shutdowns.3819519%2F&ranMID=44583&ranEAID=2294204&ranSiteID=kXQk6.ivFEQ-G2c2Dmd6BBcFbkDuG3wHwQ

 

[cdriccio]

I don't think you can go wrong with a Corsair RM or HX series. My original HX lasted 10+ years and three rebuilds.

So I did order that PSU. Based on a power calculator I found, my PC is pretty close to being too much for the PSU. I received it yesterday and plan to install it asap.
I'll report back on whether or not it fixes the issue.

PC shut down on me again yesterday while playing BG3

 

[jkevery]

The 3000 series RTX are power hungry cards. My son's 4070 pulls far fewer watts than my 3070. Fingers crossed that this cures it.

 

[cdriccio]

The 3000 series RTX are power hungry cards. My son's 4070 pulls far fewer watts than my 3070. Fingers crossed that this cures it.

I added a second SSD and an traditional harddrive to the box as well.... granted that while ago, so I'm still a little concerned that the PSU is not the problem... but I feel like it's worth the risk at this point.

 

[cdriccio]

New symptom I guess... I haven't changed out the psu yet.
System just powered down while I had a USB powered Key light plugged in and while the system powered down, the light never went off? Not sure if that's telling of anything.

 

[cdriccio]

I replaced my power supply and all the power cables over the weekend. The PC is back up and running and I'm waiting to see if the problem appears again. It's only been 20 or so hours and since this problem started, I've seen it go as long as 4 days, so I'm keeping an eye on it. I did a few hours of gaming on it last night as well.

I did find something unusual that I think could be the issue. While removing the old power supply, I noticed that I did not need to "pinch" the motherboard power cable to remove it from the PSU. Seems like the tab may have been broken and it was only loosely connected requiring very little force to remove it. I'm HOPING that it was an unstable connection to the PSU that was causing the shutdowns.

Thoughts?

 

[jkevery]

May not have been pushed in fully or enough for the tab to click into place. Should have thought to suggest re-seating the two motherboard connectors, but probably wouldn't have thought to recheck the PSU side connectors. Reminds me of the 4090 issues with the melting connectors which NVIDIA blamed on not being fully seated. Is it still chugging along OK?

 

[cdriccio]

May not have been pushed in fully or enough for the tab to click into place.

Oh I tried several times to get the cable to click into the PSU... it wouldn't... the tab was broken. So it was very easy to just gently pull the cable out of the PSU.

Is it still chugging along OK?

Well... it's been more than 2 days so far and it's been up the whole time. *crosses fingers*

 

[jkevery]

Still running without shutdowns?