Old NVIDIA card; Display driver BSOD, monitors black on reboot

[Kronos51]

Hey guys,

So I hate to be the guy popping in to ask a question out of nowhere, but I'm not the troubleshooter I once was.

My desktop display driver just BSOD'd on me and when I turn the computer back on my monitors don't recognize the computer (as in, they stay asleep and don't think they're connected to anything).

The rig's about 4 years old now, and is an old MSI P7N SLI Platinum motherboard with a single Geforce 8800 GT. Nothing overclocked. I've got dual monitors as of about a month ago (previously just had a single). An Asus is my main monitor with a Panasonic secondary.

Unfortunately I didn't have time to write down the details of the BSOD before the display cut out, but what I saw was that the display driver had crashed and failed to reinitialize. It happened while playing DOTA 2. It was moving along fine without any noticeable difficulty and then I got some artifacts, the screen froze, 5 seconds later I got a BSOD on my secondary monitor (primary monitor went black), 5-10 seconds later both screens went black. I reset the computer and the screens stayed black. The computer turns off at the touch of the power button (which usually requires a few second hold to hard shut down) so I'm guessing it's not booting into Windows either.

Tried letting it sit for 10-15 minutes to see if it needed to cool down, but I get the same result.

This happened once before either last night or the night before but the computer restarted just fine with no noticeable problems.

I tried searching the web but all I could turn up was people getting repeated BSODs. I can't even get anything to show on my monitor, so I'm at a loss of what steps to take next to diagnose the issue. Is my vid card just shot?

Thanks in advance for any help, and let me know if I need to give more info

 

[J_E_D_70]

Can you borrow an old video card from someone?

 

[Kronos51]

Looking into it. I've only really got one friend here that might have one, so fingers crossed.

 

[Kronos51]

Update: Tried turning it on just now and now I get something on the screen:

It shows the beginning of the boot sequence with red text saying "Warning !!! The previous overclocking had failed, and system will restore its default setting. Press any key to continue..."

However, it won't recognize my keyboard. And again, I did not have anything overclocked.

Edit: Further update. Reset it since it wouldn't recognize keyboard (didn't think about plugging a PS/2 keyboard in first) and I'm back to getting nothing. Also tried resetting CMOS on the mobo and got nothing.

 

[Kronos51]

Further update:

Craigslisted a cheap video card (which was still an upgrade, heh) and it appears to work fine. Noticed some hiccuping in games still, but not positive if it's internet lag or video lag. So it may be solved as the video card dying.

 

[J_E_D_70]

At least it's a start! Hope that's all it is.

 

[Kronos51]

Further update: Works mostly fine, but I got another display driver BSOD yesterday. This time all of the colors were flipping randomly to other colors before the BSOD. I vaguely figured that might be a memory issue and ran a windows memory scan from the BIOS but it turned up clean. Not sure what might be the deal now. Also, I updated to the most recent NVidia drivers.

For reference, I'm running Win 7 x32 and am also considering that the fact that I only have 3 GB available RAM might be a problem too. Task manager shows about 40% memory usage when idling with a browser open. Also, the new video card was a GeForce 8800 GTX.

 

[J_E_D_70]

What is the brand, wattage, and age of your PSU? 8800 is highly power hungry. Also, have you used compressed air to blow out your PSU and the CPU heat sink/fan?

Should also download a temperature monitor program and watch your CPU and GPU temps on the secondary monitor.

 

[Kronos51]

It's a SilverStone 650W that is about 4.5 years old at this point I think. I built this system in summer of 2008 and haven't changed anything since (except I switch out RAM very quickly after I built it because something was wonky with the RAM I started with).

I wasn't too worried about underpowering because when I bought it I was planning to OC but never did, and I made sure to leave plenty of extra power (or so I thought). I can't recall specifics but I feel like I had something like 150-200W spare.

I've got a multimeter hanging around if I need to be looking at the PSU outputs (and have an electrical engineering background, so with some idea of what a normal baseline is I think I can figure out if it's working properly).

I have not blown out the fans in a good while, and definitely need to.

I've just downloaded HWMonitor and will be using that to try to keep an eye on everything, although I'm not positive what counts as too high of a temp. During idle (that is just running my web browser and background programs) my CPU is showing about 42 C/107 F and 46 C/114 F on its two cores. GPU is showing 60 C/139 F. Will test with a game running and watch it on the other monitor.

I also ran Memtest86 just to double check the RAM and after 9 hours and about 7 passes it showed 0 errors, so I guess that's something.

Still getting occasional crashes.

Thanks so much for your help so far!

 

[J_E_D_70]

By today's standards that is insanely hot idle but a quick google shows thats the norm. Wow!

That should be plenty of power too - even good psu's can go bad sometimes.

Running short on ideas here - are the capacitors on the mobo exposed or can you pull the heatsink and check to see if any of them are in bad shape?

 

[J_E_D_70]

Ya definitely check the psu if u know how

 

[Kronos51]

There are a bunch of electrolytic capacitors exposed on the mobo, but none of them are blown or otherwise looking abnormal.

Unfortunately I don't know what exactly to look for in the PSU. It seems that is not discussed too much on the net due to the danger of blowing yourself up, hah.