[SOLVED] PC randomly shuts down while gaming after upgrading CPU

Apr 14, 2020
3
0
10
Hello.
After installing a Ryzen 5 3600 on my old MSI Tomahawk b450 motherboard for 1st gen Ryzen, my PC has been regularly shutting down after some time gaming (maybe 30 minutes, maybe several hours).
I've updated all of my drivers. I flashed the bios with the newest version to be able to install a 3rd gen Ryzen, and this shouldn't have been the problem because on the MSI website for this motherboard, the new CPU was on the supported list. From the b450 support site I also downloaded and installed the newest chipset.
Thermals aren't the issue. The CPU never goes above 80º while gaming, and even stress testing with Prime95 and other benchmarks caused the PC to shut down.
Whenever the computer crashes, I get the BSOD with the following error: CLOCK_WATCHDOG_TIMEOUT. Looking at the event logger, there are always the same two errors: EventLog id6008 and id1101 ("event processing") and Kernel-power id41. No other fatal issues are registered.
When I did the new CPU installation, I also installed an N.VME SSD. I have never gotten these errors with the old CPU, a Ryzen 5 1400, or while not playing videogames.

My specs are:
Ryzen 5 3600
8GB DDR4
PNY N.VME SSD
Nox Hummer GD750w Gold+ PSU
NVidia GTX 1050 Ti from Gigabyte

I've been troubleshooting for a while and haven't come up with any solution yet. I worry for how these shutdowns may affect the hardware. Any help would be appreciated.

unknown.png
unknown.png
IMG-20200413-WA0008.jpg
 
Solution
....
How can I test the PSU if I don't have any others?
It's very hard to adequately test a PSU and that's just part of the reasons people rag on off-brands that have never been closely scrutinized by the few reviewers who test them properly. About the best you can do is watch how much the +12V input varies using a decent monitoring utility, e.g. HWInfo64. It shouldn't vary more than +/- .6V, from 19.4-12.6V, as the processing load changes. But even if a little bit under that, although apparently 'in spec', is terrible performance for modern PSU's.

But i'm somewhat with you: 750W is way overpowered for your system, so even though it may be a poor unit as long as it's relatively new it should work OK. But it's a big unknown...
Have you tried using another power supply?
I don't have any other PSUs lying around.
I suspected it may have been a problem with the PSU but it is gold and I haven't found any other people having problems with the same model. Maybe it's because it isn't from a brand as well-known as Corsair or EVGA?
It has plenty enough wattage for that matter. How can I test the PSU if I don't have any others?
 
....
How can I test the PSU if I don't have any others?
It's very hard to adequately test a PSU and that's just part of the reasons people rag on off-brands that have never been closely scrutinized by the few reviewers who test them properly. About the best you can do is watch how much the +12V input varies using a decent monitoring utility, e.g. HWInfo64. It shouldn't vary more than +/- .6V, from 19.4-12.6V, as the processing load changes. But even if a little bit under that, although apparently 'in spec', is terrible performance for modern PSU's.

But i'm somewhat with you: 750W is way overpowered for your system, so even though it may be a poor unit as long as it's relatively new it should work OK. But it's a big unknown, and it may not last long.

At any rate, I'm not sure what you've done to set up your system. If you haven't, do a CMOS reset and don't fiddle with anything for the time being, let it run with default settings.

It would probably be good to also uninstall and reinstall your video card drivers and chipset drivers. In fact, uninstall the AMD chipset drivers, uninstall the video card drivers and run DDU (display driver uninstaller) from a safe mode bootup. Run it for both AMD video card as well as Nvidia. Then re-install the chipset drivers and video card drivers.

Also be sure Windows is up-to-date, so run Windows Update. And run 'SFC /scannow' from a command prompt with admin rights.
 
Solution
It's very hard to adequately test a PSU and that's just part of the reasons people rag on off-brands that have never been closely scrutinized by the few reviewers who test them properly. About the best you can do is watch how much the +12V input varies using a decent monitoring utility, e.g. HWInfo64. It shouldn't vary more than +/- .6V, from 19.4-12.6V, as the processing load changes. But even if a little bit under that, although apparently 'in spec', is terrible performance for modern PSU's.

But i'm somewhat with you: 750W is way overpowered for your system, so even though it may be a poor unit as long as it's relatively new it should work OK. But it's a big unknown, and it may not last long.

At any rate, I'm not sure what you've done to set up your system. If you haven't, do a CMOS reset and don't fiddle with anything for the time being, let it run with default settings.

It would probably be good to also uninstall and reinstall your video card drivers and chipset drivers. In fact, uninstall the AMD chipset drivers, uninstall the video card drivers and run DDU (display driver uninstaller) from a safe mode bootup. Run it for both AMD video card as well as Nvidia. Then re-install the chipset drivers and video card drivers.

Also be sure Windows is up-to-date, so run Windows Update. And run 'SFC /scannow' from a command prompt with admin rights.

Couldn't do the voltage check because my bios is stupid and I can't enter it (probably set up in legacy and not uefi, so not even doing special startup from windows 10 works).

For the rest, I've uninstalled and reinstalled drivers thoroughly like you said. Windows is up to date and sfc /scannow didn't detect any errors.
For the moment, it hasn't happened again! I hope it is fixed now. If it doesn't crash in a week, I'll flag this post as answered with your comment as solution. Thanks!
 
Couldn't do the voltage check because my bios is stupid and I can't enter it (probably set up in legacy and not uefi, so not even doing special startup from windows 10 works).
....
OK...that's interesting. If you can't get into safe mode, it bespeaks of basic problems with Windows. It's available whether in UEFI or CSM mode of operation, that should not matter in the least. Since you've recently upgraded hardware there could be a relationship.

You won't like it: in these cases a clean install of Windows is in order. When you do the re-install, make sure your BIOS is in UEFI mode and let the installer delete all old partions and create new ones so it will convert the disk to GPT and install everything right. Of course, you should take a backup of important files and some screen shots of installed apps and desktop if you're complex with your customizations won't hurt.
 
  • Like
Reactions: vcc98554vn
OK...that's interesting. If you can't get into safe mode, it bespeaks of basic problems with Windows. It's available whether in UEFI or CSM mode of operation, that should not matter in the least. Since you've recently upgraded hardware there could be a relationship.

You won't like it: in these cases a clean install of Windows is in order. When you do the re-install, make sure your BIOS is in UEFI mode and let the installer delete all old partions and create new ones so it will convert the disk to GPT and install everything right. Of course, you should take a backup of important files and some screen shots of installed apps and desktop if you're complex with your customizations won't hurt.
thanks it helped me my computer now work properly tks very much man