PC Issue - Help Isolate the Underlying Problem

Cheers OCForum,

I've been an avid reader of the site, coming here for bleeding edge information (and even help overclocking my own rig), but just registered to post for help with my computer.

Occasionally my computer will black screen, then shut down with no warning or error messages. I'm trying to isolate the problem so I can remedy it with the least amount of out-of-pocket expense.

Info:

1. About a month ago, my computer started not recognizing the boot disk I have installed. I had three hard drives internally connected, and the computer would only pass the "INSERT SYSTEM DISK" screen on startup after I had disconnected the other two drives. I thought originally that this was a problem with a dying drive, but after replacing it with and SSD and sometimes receiving the same message, I'm led to believe it's a MOBO problem.

2. When playing MoH multiplayer beta last week, my computer would instantly shut down in the middle of a match after only playing for a half hour or so. I thought this was a video card problem at first and lowered the graphics quality, but after monitoring temps and seeing it shut down again while my CPU was at 60 Celcius and my video card at 70 Celcius, I determined that it probably wasn't a video card issue.

3. The second point is reinforced by my AV scans today. I was running AVG, MSE, MalwareBytes, and Registry Mechanic at the same time when my pc instantly shut down. Checking the temp logs show the CPU and video card at lower temperatures than when playing MoH.

4. I had read up on this issue and wanted to verify that it was not a problem with my PSU. I did the "paperclip test" and it passed, with the PSU and my peripherals turning on. My voltage values are running a little low, however: 3.3V (3.22V), 5V (4.92V), 12V (11.84V).

5. My PC specifications are as follows:
Thermaltake W0116RU 750W PSU
DFI BloodIron P35-T2RL LGA 775 Intel P35 ATX Intel Motherboard
Intel Core 2 Quad Q6600 Kentsfield slightly OC'd to 2.8GHz
OCZ Reaper HPC Edition 4GB (2 x 2GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400)
G.SKILL 2GB (2 x 1GB) 240-Pin DDR2 SDRAM DDR2 800
EVGA 512-P3-N871-AR GeForce 9800 GTX+ 512MB
Windows 7 Professional

Any help is greatly appreciated; although my PC is functional, I don't want anything to get out of hand too quickly and me find myself without a working desktop. I have Everest and will be more than willing to provide any insightful information I have left out. Thank you all for your help and consideration.

Bump, I really need some help. :-\

When are you taking those voltage readings? They look fine, within a couple percent is cool if memory serves me.

It's a good PSU you have. Could be a PSU problem, motherboard problem, ram problem... Have you ran memtest yet? That would be a good start to rule it out quickly. Any caps on the board bulging or looking nasty?

Anything extra you can disconnect and see if the problem disappears? Maybe remove power from any extra fans or hard drives, leave the side off the case for extra air, and see if the problem persists? Interested to see if the problem continues if there is less power draw.

I.M.O.G. said:

When are you taking those voltage readings? They look fine, within a couple percent is cool if memory serves me.

It's a good PSU you have. Could be a PSU problem, motherboard problem, ram problem... Have you ran memtest yet? That would be a good start to rule it out quickly. Any caps on the board bulging or looking nasty?

Anything extra you can disconnect and see if the problem disappears? Maybe remove power from any extra fans or hard drives, leave the side off the case for extra air, and see if the problem persists? Interested to see if the problem continues if there is less power draw.

Click to expand...

Thanks for the reply. I've measured those readings at load and idle, there isn't a difference. I disconnected everything from the PSU and ran a paperclip into the green and a black molex connectors on the mobo connector and the PSU's fan turned on, so Thermaltake claims it isn't a PSU problem (and thusly, won't award me an eRMA). Nevertheless, I'll disconnect unnecessary hard drives and fans and give it another go (atm I have three 3.5" drives, two 2.5" ssds, graphics card, four 120mm fans, and a 200mm fan on the top (Antec 900)).

Thanks for the tip with memtest, I'm downloading the auto-installer for USB key for it now and will post the results.

I've given the mobo a once over tonight with everything still attached, nothing looks grimy or bulbous, but will remove components and do a more in-depth inspection in the morning (it's pretty late here, class in ~6 hours).

Great, sounds like a plan. I'll be tuned in for updates tomorrow.

I'm not expecting the problem to be any different with lower power draw, but removing extra variables is a good troubleshooting step anyways so I would nevertheless try it like you mentioned.

Interested in your memtest results. I doubt its the PSU.

Ran MEMTest for a couple hours, one pass returned no errors. I've disconnected the unnecessary components and still managed to produce the same black screen, shut down result. If you have any other ideas, I would absolutely love to try them. And thanks for the help!

Edit:
Just shut down again, in the middle of playing Fallout: New Vegas for ten minutes. Last night I played without hiccup for six hours.

Second Edit:
Reproduced my first test (running MSE, AVG, MalwareBytes, and Registry Mechanic) and closely watched all my temps. My cpu hit 82Celsius before the computer shuts down, and I'm assuming that's the problem. Testing again with a huge fan blowing directly on the internal components (especially the heatsink). Sigh.

Third Edit:
I've been running all four tests simultaneously now for about ten minutes, seeing temperatures from 74 to 80 Celsius, but not as high as 82. If this is the problem, I just need to figure out why my computer was not getting this hot before. I"ll try reseating the heatsink and checking obstructions. I didn't change the internal components or move anything around until it started crashing, though, so I wonder why the temperatures are getting so high now. I see forum posts around the web indicating that 80 is sometimes operable for enthusiasts, so I won't rule out other issues. Heat is certainly a problem, though.

Keep us posted, 82C when running that stuff seems awfully hot for the CPU. Nothing that CPU intensive I would think.

Could be something that nobody ever thinks about...."whiskers" I forget the technical term, but some solder joints will grow "tin whiskers" that can short out components.
One way to keep these from borking your setup, is to use a semi stiff brush on your traces where you have a lot of solder..couldn't hurt.

Re-seating everything will also keep whiskers from growing, not to mention making you sure that you have all your ducks in a row....and keeps contaminants from eroding your foil traces.
Could it be that your thermal compound failed? get some Arctic Silver or similar TIM and re-apply the cpu H.S.