Question Computer Turns Off While Playing Games

[Bgran31]

This has been happening for a couple of days now, But it happens when the GPU and CPU are under load, the computer loses power but my peripherals and RAM stay on, the power buttons have no effect until I cut power via the outlet and reboot from there, it normally happens anywhere from 30mins to a couple of hours into gaming, never when idle.
Specs :
AMD Ryzen 7 2700X with Wraith Prism
Gigabyte Radeon RX 580 Gaming 8GB
MSI X470 Gaming Plus Motherboard
Corsair Vengeance RGB Pro CMW16GX4M2C3000C15 16GB (2x8GB) DDR4
SilverStone Strider ST65F-GS Gold 650W Power Supply
Samsung 860 EVO 2.5in SATA SSD 500GB
Western Digital WD Black 1TB WD1003FZEX 3.5in Hard Drive

I built the PC 5 months ago, since having these issues i've tried clean installing my graphics drivers, reseating the CPU/RAM/GPU, checked all the cables and connections, I've stress tested the CPU via Cinebench and GPU via 3DMARK, no crashes there, tests look all normal. I've reset my ram to the default values from 3000MHz to 2133MHz. It's getting to the point where troubleshooting takes so long due to not being able to find out if what I did has changed anything.

I'm not too sure but i've noticed if my RAM is loaded in sleep mode the RGB lights stay on, in this instance the RGB lights stay on after the crash so i've assumed they are still loaded.

I've look at the temps while I play to monitor them, they dont seem to get over 70°C under load, idling at around 50°C

 

[Obakasama]

I'm a bit suspicious of the idle temperature you mention which seems high. What is your current ambient temperature?

Any overclocking?

Checked event viewer to see if there are any clues to the sudden shutdowns?

Checked BIOS or a hardware monitor (eg. HWiNFO) to see what voltages are being read by the motherboard?

 

[Bgran31]

I'm a bit suspicious of the idle temperature you mention which seems high. What is your current ambient temperature?

Any overclocking?

Checked event viewer to see if there are any clues to the sudden shutdowns?

Checked BIOS or a hardware monitor (eg. HWiNFO) to see what voltages are being read by the motherboard?

Well I live in NE Queensland, Australia. so the ambient is around 25°-35°, the idle temp is most likely reflecting that.

I've overclocked in the past, but not very far, and not for very long when I first got the PC.

The Event Viewer critical error is pasted here https://pastebin.com/kbXkpCzr

I'm not too sure what voltages you were looking for, so I grabbed pictures of two separate bios screens

https://imgur.com/9lBA7Z8

View: https://i.imgur.com/9lBA7Z8.png

https://imgur.com/hj6mGVy

View: https://i.imgur.com/hj6mGVy.png

 

[Obakasama]

I'd use HWiNFO for voltage readings as it might be easier to navigate (given BIOS of motherboards can vary); it would be under the motherboard section looking something like:

You'd be looking at those yellow lightning bolts to see whether the 3V, 5V and 12V readings are within 5%, 5% and 10% of tolerance. Basically it's to see if there's anything potentially wrong with the PSU.

Can't make out much from the event information you linked, perhaps others will be able to.

 

[Bgran31]

I'd use HWiNFO for voltage readings as it might be easier to navigate (given BIOS of motherboards can vary); it would be under the motherboard section looking something like:

You'd be looking at those yellow lightning bolts to see whether the 3V, 5V and 12V readings are within 5%, 5% and 10% of tolerance. Basically it's to see if there's anything potentially wrong with the PSU.

Can't make out much from the event information you linked, perhaps others will be able to.

https://imgur.com/w9e1oDu

View: https://i.imgur.com/w9e1oDu.png


Let me know if there's anything else you need to know

 

[Obakasama]

Hm... I would tentatively rule out a PSU issue then, as the motherboard readings suggest the voltages running through it are fine (from what I can tell anyway).

It seems you've tried most things I can think of. But... JayzTwoCents had a video where RGB was an issue with a PC booting.

View: https://www.youtube.com/watch?v=vnST5rA64Oc

Whether it's applicable to you I've no idea, but might be worth trying to eliminate it as a possible cause. (How to do so I'm not entirely sure.)

 

[Bgran31]

I have isolated it to be the GPU causing the crashes, I can recreate the crash if I throttle the GPU with uncapped framerates, not sure where to go from here

 

[Obakasama]

That would seem to suggest a cooling issue if it crashes when it throttles.

 

[Bgran31]

So far the temps haven't showed anything that indicates it to be a cooling issue, the crashes have persisted to the point of happening after boot, even when it's not throttled. I've swapped out the GPU and I cant replicate the crash, so i'm going to send my GPU in for warranty and see how that goes.

 

[Bgran31]

So far the temps haven't showed anything that indicates it to be a cooling issue, the crashes have persisted to the point of happening after boot, even when it's not throttled. I've swapped out the GPU and I cant replicate the crash, so i'm going to send my GPU in for warranty and see how that goes.

Edit:
Nope, crash persists even with a different GPU

 

[Obakasama]

Hm... sounds like something else going on if a different GPU also crashes.

 

[Bgran31]

I'm going to do a clean install, i'll update the thread when I do

 

[kaehligj]

My favorite : broken (dry) solder joints on the power connector block pins, because of thermal cycling.
ANY slight discolorationon the connector block would hint you.
During PC normal running try an ever so slight movement of the connector block.
If it fails your PC will either freeze, get a BSoD or reboot.

 

[Bgran31]

None of the power connectors cause a freeze, bsod or reboot when jostled, no discoloration that I can find.

I've reinstalled windows, so far 12 hours without a crash, i'm going to try and get most of the software/drivers I had before installed to see if I can recreate it.

 

[Bgran31]

The crash has persisted the fresh install, leaving me to believe that it's a hardware issue with the PSU, CPU or MB

Is there any method of troubleshooting the PSU more then I already have by simply checking the voltages?

 

[kaehligj]

Using a voltmeter only tells you the instantaneous voltages.
The mb will not boot with a missing or faulty voltage.
And it wont tell you if one of the voltages fluctuates over time.

Best and sure way to test is a replacement with a known good PSU - or swap it with a PSU from a working PC.
If the symptom moves, you have hit bullseye.

But IMHO you are on track, the PSU is the first unit to verify.

 

[Bgran31]

Using a voltmeter only tells you the instantaneous voltages.
The mb will not boot with a missing or faulty voltage.
And it wont tell you if one of the voltages fluctuates over time.

Best and sure way to test is a replacement with a known good PSU - or swap it with a PSU from a working PC.
If the symptom moves, you have hit bullseye.

But IMHO you are on track, the PSU is the first unit to verify.

Yeah I've been troubleshooting it for last, but I'm at the point where I just need to throw a known good PSU in there, my only concern is it's a 550w, will that be alright for my system?

 

[kaehligj]

No, I believe you should grab the chance to upgrade for a 750W PSU.
550W is defo too weak for your setup.

Your 650W may be working fine with all shiny stars, it just cannot deliver the needed umph for your specific setup. It seems to shut down either on thermal or overcurrent.

 

[Bgran31]

I dont have access to a known good 650w/750w PSU, I can either test my system with the 550, calculations here (https://outervision.com/b/9NCA7b) Or I can transplant my PSU into another very similar machine to see if I can recreate the crash

 

[kaehligj]

Use the transplant option then.

 

[kaehligj]

Use the transplant option then.

And for fun and giggles try the 550W PSU also, while you have your PC disassembled anyway.
If that somewhat weaker 550W PSU can run your setup, you are dead sure you have a stinky 650W PSU

 

[Bgran31]

Yeah I think I'm going to swap PSUs and see if either machines crash.

 

[Bgran31]

So the PC that had the PSU in question put into it is all in all going ok, but it has had 2 freeze crashes in a couple of days, i'm not sure if that's enough to rule out the PSU