Completely random BSOD's are driving me insane. Any suggestions?
 

I got a new system a couple months ago. 3 weeks after I had it, the motherboard crapped out and I sent it back to the manufacturer for a replacement. Ever since I got it back, I've been getting random BSOD's anytime I fire up a game. (it was just randomly rebooting, but I changed the 'System Failure' options to show a BSOD instead). I've tried a number of games (Oblivion, AoE 3, Quake IV, Heroes of Might & Magic V, NWN) as well as 3DMark 2006. I can usually play for 10-30 minutes before the system crashes and I get a BSOD. FWIW, everything works fine with normal applications, and only has problems running games.

Problem is, I keep getting different BSOD error messages every time it crashes. A few of the more common ones have been:

- General 'Stop' message with no referenced drivers. (this one is the most common)
- 'nvmcp.sys' file (audio driver in the nforce chipset)
- 'nf4_disp.dll' (Geforce video display driver)
- 'driver_irql_not_less_or_equal' error (chipset error again)

If I could just get the damn thing to crash consistently with the same message, at least I'd have a place to start. But it seems to be different every time. Here's what I've tried so far:

- Reformatted and reinstalled Windows (there's nothing on the system now except Windows, Firefox, and a couple games)
- Updated nforce, video, DirectX, and audio drivers. I've used the most recent versions, beta versions, as well as a couple older revisions.
- I've swapped out the power supply, video card, and RAM with brand new replacements. Still get the same crashes.
- Updated BIOS on mobo & video card
- Disabled Cool & Quiet and FanEQ in the BIOS
- Reset BIOS back to factory defaults
- Disabled onboard audio and ran the games with video only
- Disabled audio acceleration
- Checked all temp sensors. All seem to be in acceptable ranges: CPU: 43 C (from BIOS), Case: 107 F (from temp readout on the case), GPU: 57C (from nvidia control panel)
- Ran Memtest and SiSoft Sandra benchmarks with no problems

I'm seriously tearing my hair out here. I've got this brand new gaming machine that I can't do any gaming on! Unfortunately, the shop I got it from has gone out of business between the time I bought it and today. That's what I get for buying from a small manufacturer, I suppose.

If anyone has any suggestions of other things I could try, I'm all ears. I'm starting to think I'm pretty much screwed, but maybe you guys can think of something I havent' already tried.

System specs:
- Abit AN8-V mobo (nforce 4 chipset w/onboard sound)
- Athlon 64 3500 CPU
- Geforce 7600 GT video card
- 1 GB RAM
- 160 GB WD hard drive

Thanks!

Well, from what you've provided us, it looks like its the mobo, and it sounds like it might be defective. You should be able to send it back to the manufacturer for some sort of help, repair, etc. Doesn't it have a warranty?

Gonna second this. First guess was overheating, but you posted temps that are well within an acceptable range. Since the motherboard was defective and replaced once, I'm assuming they just swapped it out for the same one. I would recommend an MSI or ASUS board to you instead.

It probably still has a manufacturer warranty, so I'll see what I can do there. It's a discontinued model and the place that built the system didn't send me any of the other stuff (manuals, discs, box, etc.), so I may be out of luck. I'll see what they say.

FWIW, the original board that crapped out was a Gigabyte, and they replaced it with this Abit board. Seems odd that two different boards would fail, but it's certainly not unheard of.

Thanks for the tips!

Could you try and increase your memory vCore setting in BIOS by a notch or two? A lot of problems related to incompatibilities between the mobo and RAM modules can be overcome by doing that. I know I did it on some computers with random BSOD issues and it worked wonders.

your nv4 BSODs could very well be a corrupt display driver. Would recommend a fresh install (also using Driver Cleaner Pro in Safe Mode) before assuming otherwise, unless those GPU BSODs occured right when the rest started to. =/