Any way to test PSU stability?

[arkk]

For over a year now at random, but mostly while gaming, my PC will go black retain sound for about 1 second, and then, the monitor will kind of stay blank until I hit the reset button. It almost looks as if I rebooted it, but that obviously isn't what happened. I've ran all 3 tests in prime for hours and memtested it and all comes back fine.

x2 3800 @ 2.2GHz
2gb RAM
Abit K8N SLI
Geforce 9600 GT
BFG 650watt PSU

Is there any way I could stress test the PSU to rule that out? It seems to happen while playing WoW at least a few times a day and even at times that aren't heavy use. Very random

 

[venm11]

Depending on the PSU's design, stress testing it could cause it to fail permanently. What makes you suspect the power supply? I would suspect the graphics card.

Sounds like what you need to do is test with another power supply and see if that helps the problem. Can you borrow one? Or, if you're unethical, you can buy one and return it.

 

[arkk]

Depending on the PSU's design, stress testing it could cause it to fail permanently. What makes you suspect the power supply? I would suspect the graphics card.

Sounds like what you need to do is test with another power supply and see if that helps the problem. Can you borrow one? Or, if you're unethical, you can buy one and return it.

I thought of that too. Would really the only viable way of stressing the video card be something like 3dmark on a loop?

I thought of the bold for both the PSU and Graphics card, lol. Trying to rule out as much as I can with what I have atm.

 

[Big Lar]

I would uninstall the vid card drivers, then boot into safe mode and use drive cleaner or such and clean out the vid drivers with that, then reboot and install newer or the same drivers over again and see if that clears the problem up.

Also it wouldn't hurt to clean the vid card, blow out the fan with compressed air, etc.

I would also download Memtest HSI for windows and run either 2 or 4 instances of that for about 400% to rule out your ram. One would think if it was the PSU it would happen more often.

Larry

 

[venm11]

Yeah I agree with whatever Big Liar says. jk. Big Lar. Good ideas.

Does it only happen when you're gaming? I've had many graphics cards that would do exactly that when they overheat.

Try touching them when running games or 3dmark, see if the heatsink is too hot to touch. I have a utility (rivatuner? ask in the gx forum) which will graph temperatures over time and record them so you can see what was going on when it crashed. One thing that worked wonders for me was to install a PCI exhaust fan right under the card so hot air wouldn't recirculate.

 

[arkk]

Yeah I agree with whatever Big Liar says. jk. Big Lar. Good ideas.

Does it only happen when you're gaming? I've had many graphics cards that would do exactly that when they overheat.

Try touching them when running games or 3dmark, see if the heatsink is too hot to touch. I have a utility (rivatuner? ask in the gx forum) which will graph temperatures over time and record them so you can see what was going on when it crashed. One thing that worked wonders for me was to install a PCI exhaust fan right under the card so hot air wouldn't recirculate.

I agree it totally sounds like heat, but both my CPU temps and Geforce temps never get crazy according to speedfan. I can add like 5 degrees to each of them too if I was just running Prime, but never any crash.

I hate issues like this where its so hard to narrow down the problem

 

[venm11]

Hm... well try and radically drop the temperatures anyway. Can you leave the case open and blow air into teh case with a big (like a vornado) fan? That way you're cooling everything in the case and can at least see if *something* is overheating if it doesn't crash when it would otherwise. Plus, a component could be mildy defective or damaged and less tolerant of even +5 deg.

Ultimately though... you're going to want to try swapping the video card, or testing your card in a friend's system.

 

[arkk]

Hm... well try and radically drop the temperatures anyway. Can you leave the case open and blow air into teh case with a big (like a vornado) fan? That way you're cooling everything in the case and can at least see if *something* is overheating if it doesn't crash when it would otherwise. Plus, a component could be mildy defective or damaged and less tolerant of even +5 deg.

Ultimately though... you're going to want to try swapping the video card, or testing your card in a friend's system.

Yeah I messed with the cables to give some more air flow and am leaving the side open to see if it helps. We shall see, thanks for the help.

I am due for an upgrade soon anyway, just not sure if I wanna take the i7 plunge.

 

[venm11]

Yeah I messed with the cables to give some more air flow and am leaving the side open to see if it helps. We shall see, thanks for the help.

I am due for an upgrade soon anyway, just not sure if I wanna take the i7 plunge.

Leaving the side open may make the airflow worse. I'm saying find a way to blast tons of air into the case to get all the little components.

In fact, given that it's winter, you could open the case and prop it up on the sill of an open window. Cold air.

 

[arkk]

Leaving the side open may make the airflow worse. I'm saying find a way to blast tons of air into the case to get all the little components.

In fact, given that it's winter, you could open the case and prop it up on the sill of an open window. Cold air.

It dropped my temps by like 10 degrees for my video card, and it seems to have alleviated the problem. But my temps were never too high, maybe my card is defective...or maybe I just have awful case airflow.

Oh well, I am prepping for an upgrade anyway.

 

[venm11]

Are you certain at this point that it's the gx card and not another component?

Which card do you have, anyway?