Question Absolutely stumped by random freezing, crashing and BSOD

Mar 4, 2019
2
0
10
Hello Tom's Hardware,
I don't know if this is specifically a GPU problem, but it's my best guess at this point.
Recently I built a new computer (my first one) and started suffering random crashes. I first replaced my refurbished gtx 1080 with a rtx 2070, and after that didn't work, I switched out the motherboard for an msi gaming x470 plus, and because we accidentally bent a pin, we got a new Ryzen 7 2700x. Long story short, we're still getting the same crashes.

Crashes are usually freezing then blackscreen, followed by sometimes a restart, sometimes manual.
There are several other oddities:
  1. Occasionally, I will be thrown a BSOD. This will usually be a VIDEO SCHEDULER ERROR and rarely something else
  2. Thinking it might be a corrupted drive, I changed both the SATA cord and the SSD for the boot drive, but not for the storage drives. My dad is under the impression that a corrupted drive that isn't a boot drive doesn't cause these sorts of crashes.
  3. Updating the video driver with GEFORCE experience made the computer crash very fast after startup, but the Windows Update display drivers don't have the same effect.
  4. Also reinstalled Windows several times, so I'm running windows with bare minimum of software (basically, browser + steam + Warhammer II for testing purposes)

I also tested the memory (2 gskill 8 gb modules) by removing each one and running the computer and both did nothing to stop crashes.
My Power Supply is 650 watt powerspec gold.
Crashes occur while gaming (usually after 30 minutes) or while doing mundane tasks like typing out this forum post, as it did.

I'd be grateful for any advice. I don't usually ask for help until I'm at my wit's end
 

Ralston18

Titan
Moderator
First, with a new build go back inside the case (power off, unplugged) and double check that all connections are fully and firmly in place. May look that way and even feel that way but no harm in checking.

If the computer is up and running you can use Reliability History/Manager and Event Viewer to look for error codes, warnings, and even informational events that correspond with the time of the crashes.

You can also use Task Manager or Resource Monitor to observe your system's performance.

First watch while not gaming. Let the computer warm up and stabilize: no updates, backing up, apps "phoning home" etc..

Then slide the window to one side and leave open. Then game and continue to watch performance. See what happens when a problem occurs.
 
  • Like
Reactions: nationality

Karadjgne

Titan
Ambassador
The powerspec 650w Gold uses the same FSP base platform as the Evga Supernova Nex 650w. Not exactly a stellar model of psus..

I'd check in windows event viewer, look for the red flagged critical errors.

It's a Ryzen. Have you updated the bios and installed all the motherboard chipset drivers, audio, network, storage, USB etc.

Almost sounds like your psu 12v rail is overheating and shutting down, the gpu being the biggest draw is going to be almost always the first to suffer.
 
  • Like
Reactions: nationality
Mar 4, 2019
2
0
10
Thanks for your responses so far.

There is a bizarre twist to this story. Suspecting hard drive corruption, my dad and I took out the storage (drive E:) SSD, and it refused to boot, saying it needed a proper boot drive. This is confusing because windows is installed on C:
For the sake of science, we took out drive C:, what was supposed to be our boot drive. We got a BSOD, of course, but said nothing about boot drives.
Our best guess is that windows was installed across two drives, which sounds insane to me. I didn't know this was possible.
Anyways we took out all drives but one and reinstalled windows. I hope this solves our issue, but I haven't really tested things yet.

"Almost sounds like your psu 12v rail is overheating and shutting down, the gpu being the biggest draw is going to be almost always the first to suffer."

Do you think I should get a new power supply? If so, do you have any recommendations? The PSU is one of the few things that has stayed constant in all my builds so this seems quite plausible.

"If the computer is up and running you can use Reliability History/Manager and Event Viewer to look for error codes, warnings, and even informational events that correspond with the time of the crashes."

Yes, this is a great suggestion and I will try it and get back to you.
 

Karadjgne

Titan
Ambassador
Me, I am biased. I've been around pc's for over 30 years and have seen my fair share of disasters caused by bad psus. So I personally always recommend getting the best psu you can afford, that fits the build. I mean no need for a $300 titanium psu on a $600 grandma surf machine, but no 80+ on a $2000 gaming rig either. That psu isn't bad, not great or even good, but not trash. There are plenty worse.

First, fix windows on a single drive. Unplug every storage device you have except for what you will place the OS on. This includes optical. You want windows to fully install on that single drive. After windows is done, updated, finished then plug in the rest. Re-download and install all the motherboard chipset drivers since the clean install of windows just wiped them all out. You'll have to dump most of what's on E: other than documents and family stuff, anything looking like windows files will create loops and possibly crash the pc again.

Then test it out. Asus RealBench is good for that, it'll punish the entire pc without driving temps through the roof. If there are instability issues, you'll find them. Run it aimed at 4 hours. That's well past any 30 minutes crash. If it survives that long, it kinda takes the blame off of the psu. If it does crash again, then the psu is the only constant, try a different one large enough to support the gpu.
 

TRENDING THREADS