Black Screen Issue - No BSOD. No warnings. No error log. No dump files

Hey all

Hoping you can help with this issue i'm facing. I've recently swapped out some of the hardware on my PC (and sold the old parts so can't use these for testing) but have been facing some issues. Randomly, either when using the computer of if it's idle, the PC will black screen and the monitor will show a 'No Signal' error when I try to manually power it back on. The monitor is fine as a work laptop is also hooked up to it and works with no issue.

I have also faced it where when it black screens it will force a reboot (Gigabyte logo appears before forcing me to log-in again, complete fresh reboot). I've also had it in some games i.e. Watch Dogs:Legion where the screen goes black during gameplay and reports a 'Driver Timeout' problem.

The weirdest part to me is the fact that despite all of this, no BSOD appears or DUMP file is logged. When I check the event viewer, the oldest 'System' log is always: "The System log file was cleared." which I'm not doing so I don't know why this is. I've done a malware scan and there's no viruses on the machine. The date of this event log that clears it is always when I boot up the PC. It does not have a filter on the log to only show certain results, it is genuinely just clearing the log upon boot-up.

Things i've tried:

Turned off fast startup
Used DDU to wipe AMD drivers and restore to a previous version - Tried 21.10.2 and 21.10.3 - Not sure what other drivers could be used with how new the GPU is
Changed power settings to 'High Performance'
Checked event viewer/Dump files - nothing available. It is enabled as I have checked my advanced system settings.
Tried a Windows Update
Disconnecting peripherals - A bluetooth USB was causing the system to freeze-up and not POST previously but this has been disconnected and binned.
Used sfc /scannow
Checked that I have connected the correct amount of power cables to GPU - 6+2pin cable needed which is connected to it.
Checking the temperatures - All temperatures are well within the guidance from the manufacturer. The only thing I've picked up is that the SSD is operating at 44 degrees non-idle (but does not go past 62 when going through intense r/w tasks)
Tried a small stress-test with Prime95 and FurMark (didn't try for too long - heard this can burn out your components on your PC). Tried a benchmark with 3DMark - didn't crash and ran at optimal performance.

One notable thing of relevance is when I accessed BIOS this morning it reported that it had reset due to the CMOS battery being removed. I have done no such thing at all. The CPU/GPU/Mobo are about less than a month old, NVME is about 3 months old and the rest is less than 1 year old.

I would also like to add that when it black screens, I have obviously tried the combination of ctrl+windows key+shift+b to reboot the driver manually. I have also tried ctrl+alt+delete and tried to press the power button. Nothing happens. There is a LED Indicator on the back of my monitor that shows when signal is being received. This is not indicated when it black screens.

What I haven't tried:

Updating BIOS
Running Windows in Safe Mode to replicate problems
Trying a different OS
Reinstalling Windows

Is there something else I should look at first before trying these fixes? Or is there any point even as it's not logging the events at all so should I look to diagnose this first?

I should add that the computer is still running (unless it does the reboot) as the Wi-Fi Card on the back is flashing and the ARGB fans on the front continue to operate.

Specs:

AMD Ryzen 5 5600x
AMD Radeon RX 6600 xt
2x 8gb A-DATA 2666mhz RAM
B550M-DS3H Motherboard
NVME Western Digital 1TB SSD - Gaming drive
ADATA SU630 SSD - Boot drive (Only reason this is not on the NVME is that I couldn't be bothered to reinstall)
Coolermaster 500w PSU - Upon checking PSU calculators it advised that this amount of wattage would remain sufficient to keep the components powered
Coolermaster Hyper 212 Black Edition CPU cooler
Windows 10 64 bit

Hope you can help.

Thank you!

Welcome to the Forum ImpossibleTop45

What parts did you changed?

Open Device Manager. Is there any yellow alerts of missing drivers?
Update your BIOS to latest Non-Beta

When that was last happening to me: PC just went dead, never a crash dump, BSOD or whatever, just a black screen - power off-on was the only fix - it was a failing motherboard - but it was 7 years old, and you say yours is under a month - still, it could be bad. I don't know how to test that bar changing it. I don't know if the ARGB fans or Wifi car light signify anything, as they will still be getting power.

Sounds like either the video card is failing or the motherboard is failing.

I had an old PC (Windows 7) that would black-screen when waking up after hours of being asleep. It continued doing that for a while until the day finally came when it wouldn't boot at all. Diagnosis: bad motherboard

Try replacing the video card with one from a PC that you know works OK. Maybe you'll get lucky.

If that fails, take your pc to a professional repair shop.

@megaherts - CPU, GPU and Motherboard (convenient I know...)
@mngerhold/Eksy61 - I have a spare mobo, would you suggest trying this first? I would like to add that I previously had a 1050 ti and although I didn't have the same issues the PC wouldn't turn its screen back on and i'd had to reboot anyway - Since that was with my old mobo, gpu and cpu would you think this leans more towards the PSU? Never had any driver timeout problems with this.

Should I look at a full reinstall before writing anything off? It's just a bit odd that the event viewer seems to wipe everything in my opinion.

Make it simple as possible.
If you have overclocked, return it to defauts.
Detach all other drives leaving on only the OS drive.
If your CPU has an internal GPU, remove the graphics card.
Remove any add in card (like Wi-Fi)
What about your PS? How powerful it is (W)?

@Megahertz

So I've got basic USBs plugged in (Keyboard wireless dongle and a mouse separate)
I can try my 1050 ti in the PC tomorrow and will let you know the results
I'll remove the PCIE Wi-Fi card as well during the test.
Using onboard graphics is not an option as the 5600x doesn't have an integrated graphics chip unfortunately.

The PSU's my only concern. It's less than a year old but from what I've read the PSU isn't the best. It's a Coolermaster MWE White 500w PSU - do you think I should look at getting rid of this first or RMA the Motherboard/CPU/GPU?

Is anything overclocked?

Difficult to guess what is the best thing to do. I will start by replacing the PS.
I think 500W is under the minimum for a system like yours.
Open a new thread under https://www.tenforums.com/pc-custom-...-overclocking/ or https://www.tenforums.com/gaming/ and ask their PS suggestion for a hardware like yours.

On BIOS you have a hardware monitor. Are the temperatures below 80ºC?
OpenHardwareMonitor Is very good.
Extract under C:\Program Files (x86)\ and run OpenHardwareMonitor.exe

Check through Reliability History for flagged entries.
If there are some,the ---view technical details--- will give usable information in the first three lines.
Also,run ---dism /online /cleanup-image /restorehealth---then run ---sfc /scannow----to recheck.
PSU calculators are useful as kindergarten toys,Add up each piece of hardware's claimed power consumption,then add a minimum of 20% of that total and see how close that is to your PSU's nominated output.There should be a difference of 100w to be usable.

@Megahertz

No overclock at all - no need for it. I've ordered a new PSU as this will likely need replacing anyway. Temperatures have been on the medium-side but not hit 80 degrees (60 degrees when surfing multiple pages simultaneously).

@joeandmarg0

I've checked reliability history. Got this for yesterday (Doesn't seem to have logged for everything that's crashed):
Description
A problem with your hardware caused Windows to stop working correctly.

Problem signature
Problem Event Name: LiveKernelEvent
Code: 1a8
Parameter 1: 1
Parameter 2: 0
Parameter 3: 0
Parameter 4: 0
OS version: 10_0_19044
Service Pack: 0_0
Product: 256_1
OS Version: 10.0.19044.2.0.0.256.48
Locale ID: 2057

It doesn't show the crashes I've had today.

With the PSU i've worked out the max power draw reported from 3DMark for the GPU/CPU to see if I can meet any 'Spikes' in wattage needed. I've added on fans, NVME, RAM, SSD, etc and 20% and that's left me with about 30w to spare so I'm thinking the PSU may be struggling; the manufacturer recommends 650w for the GPU (XFX). My only concern is though, why would this mainly happen at idle rather than at full load? Surely if the PSU wasn't good enough it'd be more struggling when I'm gaming? I'm not trying to question you as I'm sure you know a heck of a lot more than me about this stuff but it just seems a little bit odd that it's happening moreso under low load rather than high.

Example: With replying to your posts now I've been on the PC for an hour. During this time, no crashes, no driver timeouts, no problems whatsoever the PC's worked as if nothing's ever wrong. But as soon as I leave the PC to its own devices it'll crash and black screen and i'll be left in this situation. I did say in the original post it happens in Watch Dogs:Legion but this is less frequent than when it's left idling. I went out earlier and left it on to do a few things and I came back to the login screen showing it had obviously shut down while I was out. Put password in, came back to it 30 mins later, bang, needs me to log-in again

Does this sound more like the PSU failing or the fact that the GPU's not waking up properly? I do have another PCIE power cable and another PCIE port, would it be worth moving the GPU (swapping it with my Wi-Fi card) and putting it in its place? My Wi-Fi card's 4 pin in a 16 pin but I read that this shouldn't cause any harm if it's appropriately supported (supported with a case screw) so it'd be straightforward to do a swap.

EDIT: I did the commands, dism restored successfully, sfc reported no issues