Sign in with
Sign up | Sign in
Your question

Sudden crash issues> Overheating?

Last response: in Graphics & Displays
Share
June 3, 2008 7:17:01 PM

Over the last two weeks my PC has started to occasionally crash. iirc no new software had been installed prior to it developing this issue.

Two diffrent types of crashes. One is a spontaneous reboot (turning off WinXP's autoreboot to try and see an error message didn't change this, it still rebooted). The other crash is just the screen hanging in the last frame and any audio getting garbled (probably the last "frame" of audio?).

I suspect an overheating videocard being the issue due to the timing of all my observed crashes, namely during playing TF2 or playing a divx movie.
I've installed RivaTuner and Speedfan and checked temperatures.
I saw temps up to 98+C in RT and at least 110 in Speedfan.
System/cpu temps in speedfan were always in the 30-40 range.

I'm using an Asus Geforce 6600LE that is passively cooled. It has performed without any problems for years.

Any other suggestions on how to troubleshoot this? Could my PSU be dying instead of the gfxcard overheating?

Thanks in advance for any help/advice!
June 3, 2008 7:32:28 PM

One note:

Only two crashes have taken place since I took off the sidepanel of the casing. Seems less frequent.
no crashes while using RivaTuner to log the temperatures.
a b U Graphics card
June 3, 2008 7:40:28 PM

2 Possibilities:

1. Your GPU is overheating
2. Your PSU is dying

Please list full specs (including PSU)
Related resources
June 3, 2008 7:59:20 PM

^agreed. it seems like gpu issue. which either caused by the GPU or the PSU.
June 3, 2008 8:08:47 PM

so do I but also check for dust in gpu fan and psu
June 3, 2008 8:13:43 PM

PSU:
Enlight GPS-350EW (340W max)

System
AMD Athlon64 3200+
Geforce 6600LE 256mb
1 gb ram
Creative Sounblaster LIVE!
Asustek AMD939 mobo. not quite sure onthe exact model, probably A8V-VM.
June 3, 2008 8:29:45 PM

Look for a better PSU. The Antec Earthwatts 380wt would be a good one for your hardware. The GPU may be on its last dieing gasp, though. Likely as not, the passive cooling has failed and the card is now failing due to overheat. If you do end up getting a new video card, it would likely as not overwhelm the old PSU anyway, so count on replacing both.
June 3, 2008 8:35:52 PM

Looking at the spec on that power supply, I would suspect that its marginal at best. Its definitely not a high quality unit, and the electrolytic capacitors in it are probably degraded at ths point so that its not capable of meeting the requirements of the system. However, the temps on your video card are also too high, I'm assuming that they are in C and not F. Try another power supply first, a good one, and then see if your video card is still overheating.
June 3, 2008 8:35:57 PM

sailer said:
Look for a better PSU. The Antec Earthwatts 380wt would be a good one for your hardware. The GPU may be on its last dieing gasp, though. Likely as not, the passive cooling has failed and the card is now failing due to overheat. If you do end up getting a new video card, it would likely as not overwhelm the old PSU anyway, so count on replacing both.


Yeah I'm anticipating on a new PSU if I get a new videocard. I'd probably get a nice one with hdmi output for with out shiny new hdtv.


If it's just the PSU i might get a cheap replacement and hold out until I need to replace the entire system.
June 3, 2008 8:38:30 PM

rgsaunders said:
Looking at the spec on that power supply, I would suspect that its marginal at best. Its definitely not a high quality unit, and the electrolytic capacitors in it are probably degraded at ths point so that its not capable of meeting the requirements of the system. However, the temps on your video card are also too high, I'm assuming that they are in C and not F. Try another power supply first, a good one, and then see if your video card is still overheating.


They're C temperatures. And yeah, the PSU managed to work fine all this time but it's pretty old for a cheap brand.

What puzzles me is that I've never had troubles before. I can't think of any reasons (beside dust, which there isn't much of on it) that it would start overheating.
June 3, 2008 8:45:19 PM

el_loco_avs said:
They're C temperatures. And yeah, the PSU managed to work fine all this time but it's pretty old for a cheap brand.

What puzzles me is that I've never had troubles before. I can't think of any reasons (beside dust, which there isn't much of on it) that it would start overheating.


If the capacitors have degraded due to aging, common issue with electrolytic capacitors in power supply circuits, it may not be supplying the required voltage. When the voltage drops in a circuit, ie to the video card, it will draw more current to compensate for the lower voltage. This will cause an increase in heat production in the circuit. This is one possible reason why your video card is now overheating.
June 3, 2008 8:54:36 PM

Thanks for all the replies so far guys.

Is there a way I can eliminate one of the options?

Would the voltage show a spike or something while under heavy load? I think i can log voltage with Speedfan.

*goes to check*

I've started a log with the GPU temp and voltages. I'll try to give it some load, see what happens.
June 3, 2008 9:00:06 PM

el_loco_avs said:
Thanks for all the replies so far guys.

Is there a way I can eliminate one of the options?


If you can check the video card in a different computer, you could verify whether the problem followed the card or disappeared. If it follows the card, then you know the card is the problem. If it disappears, then the PSU is the next likely suspect.
June 3, 2008 10:24:42 PM

sailer said:
If you can check the video card in a different computer, you could verify whether the problem followed the card or disappeared. If it follows the card, then you know the card is the problem. If it disappears, then the PSU is the next likely suspect.


I'll have to see about being able to check in an other computer. especially since it's hard for me to replicate the problem at will on my own computer.

It remained stable right now under load while going up to 109 for a while. Voltages were all consistent with their regular levels.

I'll see about trying in a friends computer tomorrow, and i'll test out some more.
a b U Graphics card
June 3, 2008 11:32:25 PM

Also Download and run ATITool's "Scan Artifacts" for maximum stress for GPU.
June 4, 2008 7:18:36 AM

Shadow703793 said:
Also Download and run ATITool's "Scan Artifacts" for maximum stress for GPU.


Alright. I'm assuming that works on Nvidia cards just fine right?
June 4, 2008 7:53:28 AM

Hm. My system crashed within 15 minutes or so with Scan Artifacts running.

Speedfan logged a rise in temp from 75c to 97c before the crash.

One voltage reading I'm not sure about. My 12v line showed 11.99v or 12.04v early on. Later it started dropping into 11.93 more and more and eventually stayed there. No lower readings though. Seems like a small diffrence to me, but might be significant?
June 4, 2008 8:03:45 AM

The crash was diffrent than other ones before. Everything stopped and the screen was garbled with the sound getting "stuck".
a b U Graphics card
June 4, 2008 11:14:40 AM

If ATITool crashed the PC you can almost be certain that GPU overheating is one problem. Your +12v rail voltages look fine. And, yes, ATITool works on 99% of the nVidia cards.
June 4, 2008 6:03:33 PM

Well... I've been running Scan Artifacts over 23 minutes now with a big fan blowing into my open case. GPU Temperature is stable at 56C.

No crash so far... I'll go as far as 45 minutes before stopping the fan and seeing if it crashes at around 95+C again. I'd consider that an overheating issue *proved*. I'll report back.


I might be able to try a friend's 7300LE in my case on friday. Also passively cooled.
June 4, 2008 7:13:15 PM

*chuckle*

I turend off the fan after it stayed stable at 56C for 45 minutes.

Now I've done 45 minutes without a fan. I'm at 104C and rising. STill no crash O_o
June 4, 2008 7:52:24 PM

topped out at 106C. Put the side of the case bak on. Temp dropped to 102 now. Still stable and running after 85 minutes.


I'm stumped. Is the crashing just random behaviour? Or is it just because heat-crashes do not always happen?
a b U Graphics card
June 4, 2008 7:59:22 PM

Ok now try this: Run both AtiTool and Prime95 at the same time. This would simulate full 100% system load, and if PSU's week for that demand, it should re-boot. AtiTool only stresses the GPU.
June 4, 2008 9:16:55 PM

Shadow703793 said:
Ok now try this: Run both AtiTool and Prime95 at the same time. This would simulate full 100% system load, and if PSU's week for that demand, it should re-boot. AtiTool only stresses the GPU.


Actually the system crashed about 15 minutes after the case was closed up. Temp dropped a bit and then it rebooted when I wasn't paying attention. So I'm assuming it was the videocard overheating for now.

I'll try doing AtiTool/Prime95 both tomorrow, see what happens.
June 7, 2008 8:30:12 AM

Right. Now I've had some crashes while only using msn/winamp/firefox. Windows rebooted (when i told it not to...) and said it was a device driver issue.

Kinda weird.
a b U Graphics card
June 7, 2008 6:19:26 PM

^Hmm... in that case your PSU should be a cause. Windows usually don't reboot unless it's a complete hardware issue.
June 8, 2008 6:00:29 PM

Shadow703793 said:
^Hmm... in that case your PSU should be a cause. Windows usually don't reboot unless it's a complete hardware issue.


Yep. I'm gonna look around for a good quality replacment (ie. one that should last me at least through the next build).

June 10, 2008 12:51:20 PM

Well. I've not had any crashes since last week now. I've run AtiTool and Prime95 and all that. Nothing. all Smooth.


All I did was clean out the dust thoroughly (one crash after that) and run Trend HouseCall virusscan which only detected minor spyware things.

I'm puzzled... but not complaining. No excuse to buy new stuff anymore though.



Thanks for everyone's thoughts and help on the matter though. I appreciate it! :) 
a b U Graphics card
June 10, 2008 7:26:12 PM

^So overheating looks like the problem.(?). I still do recommend getting a good PSU, at least a Tier 3 (ie. Antec,Thermaltake,etc)
June 11, 2008 8:36:27 AM

Shadow703793 said:
^So overheating looks like the problem.(?). I still do recommend getting a good PSU, at least a Tier 3 (ie. Antec,Thermaltake,etc)


I think it was overheating indeed.

And I will definately get a quality PSU when the next round of upgrades takes place. But that will be a while probably.

Thanks for the suggestions! :hello: 
a b U Graphics card
June 11, 2008 11:16:43 AM

Cool. Good luck.
!