Lekro

Honorable
May 1, 2012
5
0
10,510
Hello,

I'm fighting with a problem for a month now. I've asked a lot of people for this and managed to decide what is the problem. But I really don't know what is the cause.

The setup:

ASUS P5K SE
Core2 Duo E6750
Gigabyte GTS 250 1GB
Gigabyte Superb 470W

Fan setup:

Front intake-fan at the bottom.
Exhaust fan below the PSU.
Side fan pointing at the CPU HSF.
PSU as another exhaust fan. (At the top)

Temps: Picture1
Temps without case(not just side panels off, the whole mobo and stuff separated from the case on a desk):Picture2

(Sorry for the Hungarian language pictures Alaplap actually means the mobo)

The problem is:

First occured during playing WoW a month ago. The screen just flashed and there were some random texture errors too, so overheating symptons. After that during the POST and in the BIOS there were some random lines and stuff everywhere, after that Windows 7 booted with artifacts and in a 4-bit color mode with 640x480 resolution. Ok VGA died. I replaced it with another new one. I also replaced my old PSU which was a 450W noname stuff with the new Gigabyte Superb 470W just to be sure. I put the VGA in, it lasted a day. Same thing happened the next day while gaming. Ok probably a faulty VGA again. I RMAd it. New VGA, same thing again. Lasted one day with gaming and such. The only difference is that this card didn't die completely. It still can run games smoothly, no artifacts, resolution okay etc. But whenever it reaches about 56-57C in temp it shuts down the VGA driver, sometimes BSOD because it can't recover. Now, when I take the mobo and all the stuff out from the case(you can see the temps above on the pic) everything is working perfectly fine even with this (now)defective VGA(I did try the VGAs out in another computer after the error occured in mine and the same error occured in the other computer too so the cards got defective becouse of overheating). The GPU only reaches 51C while running Crysis 2 at hardcore graphics settings, 3Dmark and MSI kombustor extreme burn in also works fine with no crashes. But as soon as I replace the mobo and stuff back to the case, no matter if side panels open or not, the mobo temps rise up 10degrees(and the GPU too).

So please if you can, tell me what is the problem with my system since the airflow is okay...

Thanks,
Lekro
 
Solution



Sounds like you've got a case grounding issue. Finding the grounding issue is going to be a little tricky though.
You need to isolate the following:

1. Is the GPU is shorting to the case
(Test the whole PC in the chassis without the GPU installed and see if it's stable)

2. Is the motherboard shorting to the case
(Test the whole PC in the chassis with a different motherboard)

3. Is the PSU shorting to the case
(Test the whole PC in the chassis with a different power supply)

If none of these end up being your problem, you probably need a different case...

flank21

Distinguished
Dec 10, 2007
329
0
18,810
the problem may be that your video cards (although new) may be defective. just because something is new in box doesn't mean it will be without errors.

ex. I just bought a new video card for my bro, installed it, tested it with pc doctor and found it to be defective. the next day it was on its way back to the retailer for a replacement.

I would recommend that you get yourself a older copy of hirens bootcd (one that contains pc doctor) and do all the video card diagnostic test that pc doctor has to offer. If the card is faulty try to rma it again and when you get your new card test it with pc doctor again but in your other computer. If it passes all the diag's with no errors then go ahead and install the card in your computer that is having the issues. Now you have verified that the card is fully functional before using it in your computer with the issues.

If the issues start to occur again, then I would say something in your computer is indeed "eating" your gpu's. I had a similar issue happen with my friends computer. We tested the video card in my computer, confirmed there were no errors, put them in his computer and a few days later he had all kinds of artifacts and a bunch of errors in pc doctor. I suspected that his PCIe bus was bad and since he only had a single PCIe slot we replaced his mobo and the problem vanished.
 

dannoddd

Distinguished
Apr 14, 2010
241
0
18,760



Sounds like you've got a case grounding issue. Finding the grounding issue is going to be a little tricky though.
You need to isolate the following:

1. Is the GPU is shorting to the case
(Test the whole PC in the chassis without the GPU installed and see if it's stable)

2. Is the motherboard shorting to the case
(Test the whole PC in the chassis with a different motherboard)

3. Is the PSU shorting to the case
(Test the whole PC in the chassis with a different power supply)

If none of these end up being your problem, you probably need a different case as it has somehow become "un-grounded".
 
Solution