Sign in with
Sign up | Sign in
Your question

Help! Dying power supply or video card or something else?

Last response: in Systems
Share
October 20, 2010 1:46:34 PM

Hi all, need your expert advices! But first let me apologize for the long read...

Problem:
Recently my computer has been acting weird. I tried to play fallout 3 (patched to latest version) and I kept getting random pc shutdown and restart. This only happened with fallout 3 though, as nothing like this happened when I played all other games, Crysis included.

Google told (sort of) me that quite a few people were also experiencing this and this is one hell of a buggy game. Fine I thought, I'll just find other games to play then.

But here comes the real problem. I have no idea if this is related to my Fallout 3 "fallout" (ha :sol:  ), but now when I run anything that requires a bit of GPU muscle, the whole screen would become a red/green/blue static mess and my computer would freeze. Hell, even Win 7's Aero would cause this.

On top of this, sometimes booting into Win 7 would appear fine at first, but when the actual desktop is about to show up the whole screen will turn into a random bright color. When this happened, pressing the restart button wouldn't do a thing and I have to hold the power button to force shutdown.

Funny thing is (and I don't mean that as "haha" funny, but more like wtf is going on) perhaps 1 out of 10 times I can successfully boot into Win 7 and run games without problems at all. Well, except Fallout 3...

Whatever the problem is it seems to be deteriorating... now whenever I run anything that requires GPU, the red/green/blue static mess of death would show its happy "I want to punch it" face. I can't even open a 640x320 video file recorded by my digital camera, not to mention my huge collection of por.... err never mind :whistle: 


What I did so far:
I'm pretty sure it's not the cpu or ram as I tested with intel burn test maximum stress, as well as OCCT. Nothing went wrong and temps are all good. Occasionally I could also run games and 3D mark vantage at extreme settings without any problem. Other times the freeze would happen as soon as the GPU is put to the task.

I thought since I could boot into Safe mode it could be a driver or OS problem, but nope, I tried everything I could think of: Cleaned the whole computer inside out, update driver, reinstall driver, reinstall OS... Didn't work, problem still persists.

Although I've targeted the PSU or the video card as the possible problem, I don't have access to replacements so there's no way for me to test which one is faulty. The only current option for me is to buy one, which is why I'm asking for your expert opinions on the most probable cause before I surrender my money...


What I have in my pc:
Power supply: Gigabyte ODIN Pro 800W
Motherboard: Gigabyte EP45 UD3R
CPU: Intel Core 2 Duo E8400
RAM: G Skill DDR2 2Gb x 2
Video card: Asus GTX 280 1Gb


So, back to my question:
Which one is more likely, a dying power supply or video card or something totally unrelated?PLEASE HELP!!!!


Moderator Edit for inappropriate content
a c 103 B Homebuilt system
a c 670 ) Power supply
October 21, 2010 9:02:18 PM

Either GPU or PSU. Have you monitored temps on the GPU? Is the cooler on the card clean? Can be heat related!
October 22, 2010 12:33:22 AM

Yes the GPU cooler is clean, and the temps are as they always has been - idle at around 50c and during gaming heats up to around 85c, depending on how graphically intensive the game is.

On a side note, OCCT graphic test heats it up to 100c in about 10 minutes and I didn't dare letting it go further...

But yeah, I don't think it's heat related as the os locked up just when the GPU is tasked to do something. I believe either the video card is fried or the PSU couldn't supply sufficient power any more. I'm leaning towards the video card but want to be more certain before I go out and buy the replacement.

Related resources
a c 103 B Homebuilt system
a c 670 ) Power supply
October 22, 2010 2:50:23 AM

It is hard to tell, could be either. If you have a chance to test one or the other in a different computer it would help finding the culprit.
October 22, 2010 3:34:11 AM

I would test the GPU first. See if you can get your hands on a healthy computer and switch out the GPUs. I would avoid testing any other good products (ie putting another GPU in your rig) if a faulty PSU is still possible as you don't want to fry any good components. If you could get a hold of a PSU tester, or even a multi meter you should be able to get some sort of Idea about the state of your PSU. I'm just very cautious w/ that as I've mixed and matched GPUs and wound up w/ a bad PSU destroying 2 GPUs.
October 23, 2010 2:18:30 AM

I'm so kicking myself in the nuts now!!

I was using version f12 of the bios, which for god knows why always lock up on POST if I have a wireless mouse receiver plugged into a usb hub (which plugs into the mobo's usb). I decided to roll back to f9 which worked fine before the upgrade, and incidentally this was about the time the problem started happening. I re-flashed the bios again, and now things seem to be back to normal. It didn't occur to me before but could it be just that I skipped loading the optimal default after flashing the bios which caused all this drama?

So far things kinda went back to normal. GPU is not giving me the cursed screen (for now) and I can play games fine. I even played Fallout 3 for about an hour before the pc shutdown and restart! Hopefully this really is the cause so I don't have to replace anything :D  But we'll see soon enough...

[UPDATE]
Nope what wasn't it. After I shut down the computer and came back to it hours later, the problem is back. Safe mode is fine, booting normally shows the cursed screen. Ah god damn it! :fou: 
October 25, 2010 9:05:51 PM

Well, when testing cpu, ram and video cards individually everything is peachy, but when I run a game, lock up after few minutes.

Individually, fine. Game, dead.

I think this may very well be the PSU that's causing the problem... thoughts?
!