monitor and peripherals turn off but tower still running

[Chocaille]

Hello everyone : )

I explain my case:

I treated myself and bought a gaming PC a little over a year ago now, at the beginning everything was going perfectly well then the problem that I'm going to try to describe arose, I would say 3-4 months after purchasing the computer, you never know if the detail matters.

The Problem

Basically, in the middle of a game, the PC screen will turn off, the sound will saturate on the last note for about 5 seconds, the keyboard, the mouse, and everything connected to the box (usb key, wifi key, and therefore the screen display cable) will turn off. The tower and everything inside stays on, then all the devices restart except the screen, I turn it off and on again, unplug it and plug it back in, it says "dp no signal" on it . So I can't do anything anymore.

To get everything working again I have to manually turn off the tower or turn off my power strip, and then everything turns back on and it's as usual.

I would point out, at the start, so the first months when it happened, that the crash happened after a few hours of play anyway, and not every day, sometimes once a week, and only on very big games like Cyberpunk, GTA, The forest 2, for honor, street fighter 6 etc...

But over time the case got worse, it became more frequent, but it didn't happen instantly yet, it happened after 4-5 hours of play (obviously when I restart everything and play again, it immediately crashes once the game is launched after 5 minutes)

Now, after a year and a few months, it's different, it's worse, the crash occurs on any game (skyrim, the forest 1, gmod, totally accurate battle simulator) instantly on launch, or after 5min no more (as soon as I start moving around in the game) and on some big games the crash even occurs at the title screen, always the same, the devices turn off, restart but not the screen, I reboot All.

Moreover, the PC is not particularly hot at the time of the crash, it makes no noise, none of the CM LEDs light up at any time.

Original PC Spec

First of all, here is my starting config (when purchasing the PC and when the crash appeared, I prefer to specify everything just in case, and I apologize because I am not very knowledgeable):

CG: VGP AMD RADEON RX 6900XT 16GB GIGABYTE GAMING OC 16GB

Proc: AMD AM4 RYZEN 7 5800X 8-core

Cooler: kuhler be quiet Pure Rock Slim 2 130W

Motherboard: MSI B550-A Pro ATX

RAM: DDR4-3200 16GB (x2)

POWER SUPPLY: MSI MPG A850GF 850W fully modular

HARD DRIVE: SSD 1000GB M.2 PCI-E NVMe WD Blue

My screen is a 165Hz from MSI

And in trying to resolve this problem I replaced several things, one after the other: The cooler, the power supply, the motherboard, the ram, (and also the keyboard and the mouse but that I don't think matters much )

Current PC Spec

So here is my current config:

CG: VGP AMD RADEON RX 6900XT 16GB GIGABYTE GAMING OC 16GB

Proc: AMD AM4 RYZEN 7 5800X 8-core

Cooler: Thermalright peerless assassin 120SE here: https://www.amazon.fr/gp/product/B09NSTR7JZ/ref=ppx_yo_dt_b_asin_image_o02_s00?ie=UTF8&psc=1

Motherboard: MSI B550 Tomahawk Motherboard: https://www.amazon.fr/gp/product/B08B4V583Q/ref=ppx_yo_dt_b_asin_image_o08_s00?ie=UTF8&psc=1

RAM: Corsair vengeance RGB Pro 32GB (X2): https://www.amazon.fr/gp/product/B082DGZJ9C/ref=ppx_yo_dt_b_asin_image_o09_s00?ie=UTF8&psc=1

POWER SUPPLY: Corsair RM1000x 80 PLUS Gold 1000 Watts: https://www.amazon.fr/gp/product/B08ZYTBVXJ/ref=ppx_yo_dt_b_asin_image_o01_s00?ie=UTF8&psc=1

HARD DRIVE: SSD 1000GB M.2 PCI-E NVMe WD Blue

I also replaced my power strip which was very old with this: https://www.amazon.fr/gp/product/B09MQ9ZG6W/ref=ppx_yo_dt_b_asin_image_o07_s00?ie=UTF8&psc=1

I specify that I am on Windows 11

Troubleshooting tried

What else I tried to fix the problem:

Reset

Reinstalling the motherboard and CG drivers (including adrenalin from AMD, updated the same)

Reinstalling windows

BIOS update (I specify that I have the XPM profile activated too)

Setting the power supply management, I also deactivated the option which turned off the devices beyond a certain threshold

I even plugged my PC into another outlet, just in case, nothing changed

The element observer tells me Kernel power 41 (63) every time this crash occurs, but I think it's because I just turn off the PC by pressing the tower to turn it back on Next

The SFC/Scannow command does not indicate any errors, commands with HEALTH do not change anything either

I launched several power supply tests with OCCT, all lasting 1 hour, as well as the tests of each tab, for each component, for a duration of 1 hour each.

the pc heated up but never crashed

I show what the data shows during a 30min power supply test with HWMonitor open:

https://forum.tomshardware.fr/attachments/capture-décran-2023-10-20-131722-jpg.15420/

https://forum.tomshardware.fr/attachments/capture-décran-2023-10-20-131733-jpg.15421/

https://forum.tomshardware.fr/attachments/capture-d_écran-2023-10-20-131749-jpeg.15422/

https://forum.tomshardware.fr/attachments/capture-d_écran-2023-10-20-131759-jpeg.15423/

In idle:

56 degrees for the CG and 38 for the CPU

My Theories

There you go, personally I feel like I've tried everything, I've searched everywhere on the internet, I've scoured the forums and I can't find anything that comes close to my problem, and therefore even less the solution, I leave it to you

However, I have two theories: either the power supply of my barrack is faulty and does not support the PC, therefore the PC does not have enough current to operate and sacrifices peripherals, which is possible because it is of an old house, and so I am lost.

Either the processor/CG has a problem.

I also notice something else, I don't know if it's related, there are like "short circuits" in my case, for example if I plug a USB key in the front, therefore close to my wifi key, the wifi key cuts out, in short, a nice mess

There you go, I've laid out everything a bit, I'm still available if there are any questions obviously, thank you to those who read, I've been in distress for quite some time. x)

New CPU

Note: I have since opted for a new processor, the Ryzen 7 5800X 3D, no choice because the motherboard is am4

No change though, still the same crashes

(Sorry for my not great English by the way I didn't find any solution in my native country)

[Mod edit: added headings for clarity.]

 

[Brian Boru]

Welcome to the forum

I added some headings for clarity, since your excellent explanation is a long read

Experts will be along shortly, but I notice you didn't mention drivers. If you haven't done so, update ALL hardware drivers from part maker sites.

 

[Chocaille]

Welcome to the forum

I added some headings for clarity, since your excellent explanation is a long read

Experts will be along shortly, but I notice you didn't mention drivers. If you haven't done so, update ALL hardware drivers from part maker sites.

i did, after a complete reset of my pc, i'm trying to delete adrenalin now, just for testing

thx a lot btw

update : other stuff i did ---> undervolt the gpu and reset amd adrenalin, no success

 

[Chocaille]

I did three more things : disable freesync premium, limit games to 60fps and reset everything in adrenalin to default
with this no crash, then i tried to disable the fps limit, and my computer turn off minutes later
maybe a problem with the cg or the monitor ? I don't know, or maybe windows...

 

[Brian Boru]

it's worse, the crash occurs on any game (skyrim, the forest 1, gmod, totally accurate battle simulator)

Have you tried any small games like solitaire etc? Just wondering if there's a cut-off point somewhere…

If they're ok, ramp up a notch to say a bigger old game, eg Civ 4 from 00s.

 

[Chocaille]

Have you tried any small games like solitaire etc? Just wondering if there's a cut-off point somewhere…

If they're ok, ramp up a notch to say a bigger old game, eg Civ 4 from 00s.

I can play civ V, i already tried
depends on the game, but the computer holds the old games globaly
instant crash on skyrim, lethal company (yeah, really), I can play armored core for 20 minutes average, if I set a limit of 30 fps on cyberpunk I can play it without any crashes
here I use TABS for my tests, the game crashes if I remove the 60fps limit

 

[Brian Boru]

Someone knowledgeable should be along soon—I'm not au fait with hardware, especially anything recent—my last build was 9 years ago.

Apart from a little Skyrim, I haven't played any of your problem games. I suppose their obvious common trait is they put significant load on a PC. Can you think of any more particular trait they might share, eg load a lot of textures or cause power draw spikes?

Do you have 100GB or more free on your drive?

 

[Kaamos_Llama]

That sounds like a really frustrating one, you tried almost everything already.

Since you replaced everything except the GPU and the case, thats what I would be looking at. Your GPU has massive power spikes of over 600W at 20ms, but I'd assume the PSU you have should handle that with your system. I would have rather replaced with a good 1000W unit for that reason, to eliminate any doubt there.

You mentioned USB shorting. Have you tried taking everything out of the case and setting up the motherboard and rest of PC on a non conductive surface like cardboard? The motherboard shorting out would be a big problem.

 

[Chocaille]

That sounds like a really frustrating one, you tried almost everything already.

Since you replaced everything except the GPU and the case, thats what I would be looking at. Your GPU has massive power spikes of over 600W at 20ms, but I'd assume the PSU you have should handle that with your system. I would have rather replaced with a good 1000W unit for that reason, to eliminate any doubt there.

You mentioned USB shorting. Have you tried taking everything out of the case and setting up the motherboard and rest of PC on a non conductive surface like cardboard? The motherboard shorting out would be a big problem.

Basically the tower was on a polystyrene plate, I don't really have the strength to pay even more, I'm afraid it won't be of any use
I will try in the future, I especially hoped that there was a Windows error, or something, the graphics card was new like the rest, it seems strange to me that it was malfunctioning, but that might be the case , yet I tested it on software like occt for long hours and on big tests, and no crash

I just don't urderstant x)

 

[Kaamos_Llama]

Basically the tower was on a polystyrene plate, I don't really have the strength to pay even more, I'm afraid it won't be of any use
I will try in the future, I especially hoped that there was a Windows error, or something, the graphics card was new like the rest, it seems strange to me that it was malfunctioning, but that might be the case , yet I tested it on software like occt for long hours and on big tests, and no crash

I just don't urderstant x)

I'm not sure if you understood. A short could be happening via the case through the back of the motherboard or through a stand off. By removing the mobo and all components entirely from the case and running the system in the open air you would be able to tell if that was the case.

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

 

[Chocaille]

I may have found some semblance of an answer, the power supply fan does not turn in idle as in game they do not turn and I just saw H24 whether in game or on the desk, thegraphic card fans are immobile
Do you have an explanation? the fans only spin at startup then stop

 

[Kaamos_Llama]

Sorry I missed that you have an RMx1000W earlier.

It has a 'Zero Fan' feature so that the fan only spins up when they are warming up under heavy load. Thats normal.

 

[mjs warlord]

I just read what you said about fans only spin a start up ..... thats supposed to happen , after that they dont normally run again till the card reaches around 70 degrees , to eliminate the possibility of an over heating card crashing your system i would recommend you try the following .... load up a game that you know will eventually crash your pc but dont actually use it , use ctrl , alt , del to bring up task manager on top of the game , now monitor the gpu temp and as i say at around 70 degrees you should see your cards fans kick in

 

[Inspireless Llama]

Even though I can't give an idea on what the issue might be, I would strongly recommend like what @Kaamos_Llama said, rebuilt the PC outside of your case and jump start.

Simular to my PC, I replaced literally every part of my case, never fixed the issue, until I rebuilt the PC outside of my case and it booted and worked without issues. My case was the issue (more specific a single cable from the case to the motherboard).

While I can't say the same is true for you, it can help excluding the case as a potential for the issues you're having.

If it turns out your case is the issue and you dont want to instantly replace it: I think I had my "PC" running of my motherboard box for about 2 weeks before I rebuilt it

Even if it's your case you can try to single out what exactly it is. I'm still using my case but I never replugged my RESET SWITCH cables to my motherboard since I've never used them and can do without them (they were my issue).

Long time RAM issues, could it be my case?

Hey guys, I have a question regarding my PC, which has been having issues for years now. Every issue came back to the RAM not being detected or signs that appeared to be that the RAM was faulty. The issues started after I moved. I didn't use my pc for a few months because i didn't have a desk...

forums.pcgamer.com