What the Hell is Wrong with Me? [Serious PC issue]

The Dude With Herpes

Alright, so yesterday my primary PC starting having a bizarre crash issue.

Randomly throughout each day my PC will just go black. I don't know how else to describe it. What happens is that as far as I can tell, from the sounds coming from the box, the PC is still on, but all video is gone and if there is audio playing through the speakers it will "buzz" until I physically shut off the PC. When I reboot I get no error message other than the usual "Your PC was not shut down properly" boot message. No "recovered from a serious error" when I get back into windows.

My initial thoughts was that my video card was overheating. But there's not any real reason for that since I'm just playing WoW when it happens (dunno if it happens with other games, haven't played anything else since it started) and that isn't exactly taxing on a 460 gtx. I also thought maybe my CPU is biting it. I didn't really have anything to base that on except for the weirdness of the "crash" and the messed up audio leads me to think that it's maybe not my video card. Either way I know it isnt' happening when the pc is idle, so it is something when it is under some level of load. I also noticed it seems to be happening when I'm afk for a minute and the screen has just been sitting there not moving.

Well, I opened up my box, took out all the lint (there was very little, I do this periodically) and also took apart my video card and processor and redid the thermal paste as it appeared it could use it. The paste job on the video card was shit, sooooo much thermal paste, I'm not sure what EVGA was thinking there. And on the Processor it looked like some lint and stuff had gotten all up ins under the fan and onto the processor.

Neither part looked bad. There was some slight discoloration on the fan for the CPU, but not around where it made contact with the CPU, so I just took it for dirtiness and maybe burnt lint, since it wiped right off with alcohol. The video card was perfectly fine, no discoloration, no warping of the board or anything that would indicate overheating.

Anyway that went fine, booted back up fine, everything seemed ok, and well, it just did it again when I was playing WoW. So I'm suspecting it isn't overheating.

Well, I'm at a loss. it wouldn't be the PSU going out since the pc stays on fine through the crash. There's no graphical artifacts in games on on my desktop to indicate it would be the video card or the CPU. I'm left with RAM or the MOBO. RAM failures usually don't result in that sort of crashing, so maybe the MOBO? I don't really know how I would test that. Maybe I need to reseat the heatsinks on the mobo and paste them?

For reference, as I said it's a GTX 460. The processor is a (getting pretty old-ish) Core2Duo 2800ghz and the mobo is an asus uh...p5ql or something like that. I don't remember. My drivers are up to date on everything except the 460 because the latest nvidia whql drivers are complete shit and constantly made my video card crash.

Help! Please!

EDIT: Gonna play something else for a bit and see if I can make it crash in anything except WoW.

EDIT2: Played Orcs Must Die for a few hours and no issues at all. Everything ran silky smooth.

It occurred to me that I've been playing WoW and using the Steam overlay a lot. I wonder if it could be related to that at all. A memory leak or something. It has done the crash sometimes when I have the overlay up, and sometimes when not. Maybe I'll play the next few days without it and see if it crashes still.

The Dude With Herpes

Ok, its not Steam. Got the crash again after about an hour of WoW. I was thinking that I had just updated my addons a couple days ago, before this started. I can't imagine an addon could do this, but maybe I should go through and check them out.

I did want to clarify though, that when it happens, it's not just a black screen, all video to the monitor is lost entirely.

grouch993

Had very similar symptoms with an 8800gtx. At first it was certain titles that would cause it to happen. After a while it was any graphics intensive program and it eventually failed.

Still under warranty? I have been told EVGA has an advance replacement program.

The Dude With Herpes

Yeah I think I got the lifetime warranty version. I don't recall if I ever registered it; if that even matters. I'm still not convinced it's the video card though. EDIT: I guess I keep hoping it's the mobo/cpu so I have an excuse to upgrade them.

Is there somewhere in Windows 7 that I can look up an error log that might give me a hint at what happens when it crashes?

The Dude With Herpes

Windows isn't tracking the error; or at least it's not tracking why it is crashing. Every instance of the crash is just listed as a "Windows did not shut down properly" error with nothing else tagged to it. So that's not useful.

I'm updating my video card drivers to the current beta ones but I can't imagine that has anything to do with it; I didn't change anything in the drivers when this all started a couple days ago.

I don't have another video card to swap out to test either I don't want to go through the process of contacting EVGA if I can't reproduce the error. It seems to be totally random, but it's always WoW. Ugh.

The Dude With Herpes

Welp, slapped on a hardware monitor to log the temps of my CPU/GPU and it's definitely my GPU. It's hitting 100C+ when it crashes. The fan is working fine too so I don't know what is up.

Submitted a ticket, hopefully I can get the thing RMA'd soon; it's the lifetime warranty version. EDIT: Pretty damn glad I got an EVGA and paid a bit more for the lifetime version. Now I hope they honor it.

EDIT: Hmm...last crash it was only at 99 when it finally gave up. The fan was going nuts though. It had hit 103 a couple times prior and didn't die. It seems kind of arbitrary.

I was also noticing in WoW, I can pretty much make my GPU fan go wild by finding an area with water, but more importantly, lots of things for that water to reflect.

This has never been an issue for my card in the past. Well, either way it just has me more and more sure something is broken with my graphics card. bleh.

The Dude With Herpes

Got an RMA #. But now I have to wait for the RMA to be approved. They have an odd way of doing things over there.

I imagine they have good means to test the hardware that I can't possibly do. I'll be annoyed if they come back and say they couldn't reproduce it and I'm stuck with a malfunctioning card. Also I hope the whole process doesn't take to long. I think I have an old 4850, upon thinking about it, that I can probably put in for a few days, but going back to that will be painful. Its good enough to do my WoW holiday stuff and then I can just play the 360 or something.

I'm also going to get a new CPU I think, just to be safe. I'm debating if I want to shell out $320 for the core2quad that is the highest cpu my mobo can take; which is still more than 2x as powerful (according to various benchmarks) than my current core2duo or if I want to pay another $150 or so and get a cheap i5, new mobo and ram, which would be probably even more powerful than the core2quad, with the ability to upgrade to a i7 down the road if I wanted to.

Ugh. I've only had this PC for 3 years, that video card for 1; shit should last longer than this. Though, I did go a pretty "cheap" route in the first place when I built it, even when I knew better. Live and learn I guess. I certainly won't be skimping when I get around to building my next PC.

The Dude With Herpes

So, got the RMA process all complete and approved.

It appears they'll be sending me a replacement card with a box to mail mine back to them.

So, that's cool? Anyone have any experience with this stuff specifically with EVGA, but maybe other card companies too?

Will it just be "all good" when I get my new card and mail mine back in? Is there a chance they're going to test the card and be like "dude is a dipshit, give us our crap back?" or will they just refurb it and resell it?

I'm asking because I've been debating getting a second 460 and if I can be pretty sure I'm in the clear here for getting this replaced cleanly, I will go ahead and buy the second one whenever Newegg gets it back in stock (if).

mts

hmm, i sometimes get something similar on my 5770. what did you use to monitor the temps?

The Dude With Herpes

Something called HWiNFO32. It's free; not sure how "quality" it is but it was more than enough for what I needed.

You can find it here: http://www.hwinfo.com/download32.html

The log it creates is an XLS file; not sure if you can change that though. I just needed something that was logging as it crashed since windows wasn't. either way it told me what I needed to know.

Dhalphir

I had the same thing with a 9800GTX playing Starcraft 2 - Ended up buying a GTX460 to replace it and had no problems after that.

grouch993

The Dude With Herpes wrote:

So, got the RMA process all complete and approved.

It appears they'll be sending me a replacement card with a box to mail mine back to them.

So, that's cool? Anyone have any experience with this stuff specifically with EVGA, but maybe other card companies too?

Will it just be "all good" when I get my new card and mail mine back in? Is there a chance they're going to test the card and be like "dude is a dipshit, give us our crap back?" or will they just refurb it and resell it?

I'm asking because I've been debating getting a second 460 and if I can be pretty sure I'm in the clear here for getting this replaced cleanly, I will go ahead and buy the second one whenever Newegg gets it back in stock (if).

S3 used to do this. Advanced replacement. They would put a hold for the amount on your credit card and clear it when the return showed up.

I haven't had to return anything to EVGA, but a friend worked with them on a couple cards. Nothing but praise from him. Sort of why I went with the lifetime warranty on my latest card. His were both true failures, so I don't know if they would have tested it and came back with questions.

On the new cpu vs cpu/mb/ram. Do the latter. It will cost less and carry you further. Having to reinstall/rebuild the system will be a pain, but long term a better solution.