Computer freezes in all my 3-D games - possible heat problem

[PolymerTim]

OK, this is my first post here so bear with me and let me know if I've left anything out.

My computer freezes in both "Black and White" and in "Unreal Tournament" at random times into gameplay. Sometimes I can play for an hour and sometimes only 5 minutes. This hasn't happened in any of my other games. I tried reducing the game detail to minimum and I think that extended gameplay some, but like I said it is pretty random. I've tried running the games in Win2000 and Win98 on the same machine with 2 different video cards and have had lockups in all cases. My case cooling probably isn't what it should be. I can feel one really hot spot on the motherboard, I think its the Southgate (it has a heatsink and is at the top of the MB, eclipsed by the slotA CPU's heatsink). Any ideas, this is starting to drive me crazy.

Specs:
Main video card - Excalibur nVidia GeForce2 MX 400 with 64MB SDRAM
Alternate video card - 8MB nVidia Vanta TNT2
Motherboard - MSI 6167
CPU - Original Athlon 500MHz - Slot A
Case - Mid-tower with 1 intake fan, power supply fan, and 1 slot-cooler placed directly below the video card heatsink.

 

[xospec1alk]

same problem here, i disabled acpi, reinstalled windows so my vid card and sound card are on different irqs, it was on the same one before, and it worked great....i put in service pack 2 for win2k, next thing u kno its locking up again, im gonna try taking out sp2 tonite and see what happens...i have an intake fan, and i plan on getting a slot exhaust fan, but for the time being, i have a 80mm fan blowing right on my vid card cuz i suspected it was overheating

epox 8kha+
radeon 8500 64mb
athlonxp 1900+
win 2k
sblive value

 

[PolymerTim]

Just a little something I recently noticed:

I didn't used to think my video card was having a heat problem, even though its one of the ones with a massive heat sink, but no fan. I noticed some other brands of teh same card do come with a fan, and that got me a little worried. Still, I can feel the heatsink during gameplay, and even right at a freeze-up, and it is only warm to the touch. Then one day on a hunch, I touched the top side of the board, directly opposite the GPU and ouch. I'm gonna have to guess it was running about 75 - 80 degrees C (Just an estimate based on feel). Now that's got me wondering just how effective the heatsink is on the other side to have about 40 degree difference between the sides of the board. As a bit of a drastic measure I tried spraying compressed air (OK so its tetrafluoroethane - the organic chemist in me won't let me just call it air) as a liquid directly onto the opposite side of the board. In short bursts every few minutes, this did manage to keep the temp down, but I ran out of gas within 30 minutes. Like I said before, sometimes I can play this long without freezing, its kind of random.

I imagine this is not normal temperature for a video card, anyone else know about this?

 

[xospec1alk]

bump bump

anyone have a clue?

 

[RudeBoie]

Well you could run Prime95 and see if that causes any crashes.

Then you'll know if it's non-vc related.

If there's something next to the VC (slots wise) try removing it for more airspace.

Case cooling shouldn't be too much of a problem unless it's really warm in your room- it's more important for oc'ing not running at defaults.

Also, if you run Motherboard Monitor, you can directly observe CPU/SYS/PSU temps.

A BIOS update may also help problems.

 

[PolymerTim]

Well you could run Prime95 and see if that causes any crashes.
Then you'll know if it's non-vc related.
If there's something next to the VC (slots wise) try removing it for more airspace.
Case cooling shouldn't be too much of a problem unless it's really warm in your room- it's more important for oc'ing not running at defaults.
Also, if you run Motherboard Monitor, you can directly observe CPU/SYS/PSU temps.
A BIOS update may also help problems.

Unfortunately I don't believe my motherboard has any built in heat sensors, and I don't know of any other way to get a temperature readout (I've been wanting to get one of those infrared thermometers). But I have searched out the hotspots by hand (literally) and can tell you if something is cool, warm, or hot. It shouldn't be processor related as I haev been running dnet for 2 years now (and more recently UD also) with no probs. The processor (being only a 500MHz Athlon) is barely warm to the touch as it has a more than adequate HS and fan. As mentioned before I have a slot cooler in slot directly below the card, but directly above the card is about an inch of free space before running into the processor HS, which as I said doesn't radiate much heat. The top side of the video card, directly opposite the GPU is very hot. Since my original post I've been reading about the GeForce2 GTX with a similar heat problem. Also looking around at other brands of my video card, I see that mine is one of the only ones without a fan on the HS. Also, one of the motherboard chipsets is very hot (either the northbridge or southbridge - whichever one is right next to the CPU). It also has a HS on it.

New update - last night I was able to play for 3 hours before crash (much more than average). I also took note that the room was at least 5 degrees cooler than usual (I live in a dorm with almost no control over the temperature), At this point my best guess is that I just need another case fan. I never realized one case fan could make all that much difference, but I've only got one intake and one power supply fan and the slot cooler. I know almost nothing about cooling, but I think I've heard that you need more pulling out than pulling in.

I also saw on someone elses post some interesting ways to cool the video card, I'm almost tempted to try taping an old CPU fan on top of the card. Any comments, ideas, general rants are all welcome.