I think I should RMA the video card, but I want to be sure

[InternetDuder]

I got overheating-like syndromes in games with my Leadtek 256mb 6800GT. Everything is stock clocked. The game would run fine, then stutter every now and then, getting more and more often, then it'd do one of 2 things:
-Bluescreen, giving an error about nvidia. When windows (xp) came back, it'd blame it on my video card.
-Hang for like 30 seconds, then go back to desktop in really crappy resolution (640 iirc), 16 colors, with an error saying something involving "a device has failed".

Seems like a case of video card overheating, but then I ran prime95 and it failed. I redid the CPU/HSF contact, and prime95 passed twice after that (20 hours each time), so I get to blame the video card again.

Memtest86 was also flawless.

I also was told it might be Dual channel messing up, so I tried one stick of ram instead of 2 (tried each stick) and the stuttering still existed.

My temporary solution is to blow a table fan on the open case, which solves the problem.

The case is well ventilated, and should not pose a problem to the video card.

NView (which has lied to me about voltages) claims the temperatures are in the normal range. I don't remember the temps exactly, but in a previous topic, I was assured the temps are normal.

I underclocked the processor multiplier, and the problem stayed.

I could not underclock the FSB.

So here I am. I am ready to RMA the video card (its newegg, I assume the policy is good), but I do want to be sure about it. If anyone thinks there is anything else I should try, please let me know. Thanks.

Leadtek 256mb 6800GT
AMD Venice 3200+
DFI Lanparty UT Ultra-D

 

[Tu13erhead]

Did you check the temperature on the card? Try this: Display Properties, Settings tab, Advanced, GeForce 6800GT tab. There _should_ be a temperature readout somewhere there.

You could also try re-seating the GPU HSF with Arctic Silver 5. I'd check temps first though.

 

[InternetDuder]

Temps read fine, and I would have to buy thermal paste to reseat.

 

[otispunkmeyer]

Originally posted by: InternetDuder
Temps read fine, and I would have to buy thermal paste to reseat.

cheaper than buying a new card.

i have a leadtek 6800gt and while its totally fine, i may just do the same...i may order some artic silver. its summer now and its the hottest ever. (very very few homes in the UK have AC so were just sweating) my card is still stable but getting very hot.

to date riddick has been the most demanding pushing my card to nearly 80degrees...i reckon playing riddick now would just burn the gpu up.

 

[Spacecomber]

If you have WinFox installed or have Coolbits enabled, maybe trying underclocking the card to see if this makes a difference.

I would think that running an open case with a large house fan is telling of some hardware overheating, but it still leaves open the possibility that the video card is adding enough heat to the case (during games) that some other component becomes unhappy (for example, chipset). This might explain why Prime95 and MemTest are fine when run by themselves.

You might try playing a game and running Prime95 in the background, but I don't know if this will leave enough CPU cycles for Prime95 to really do much testing, since it runs at such a low priority.

By the way, what sort of power supply do you have?

Space

 

[InternetDuder]

I have an antec 450w PSU.

The case right now is open, and I have a 120mm in the front (intake). Ventilation should not be an issue. The room is very cool, and the problem is far too apparent to consider natural (happens in less than an hour of play)

I will try underclocking

 

[blckgrffn]

DO NOT remove the stock heatsink! Don't void your warrant! Just RMA the card and be done with it!

My $.02...

 

[InternetDuder]

Originally posted by: blckgrffn
DO NOT remove the stock heatsink! Don't void your warrant! Just RMA the card and be done with it!

My $.02...

indeed, i have never applied a gpu hsf, and I could get a painless and free RMA right now. I would much rather be safe.

 

[JokerRulez]

I had to RMA two of the Leadktek GeForce 6800 GT cards. I quit on Leadtek. Never again.

I agree with RMA'ing this card. It is the smart, safe thing to do and it only costs you: 1. Time and 2. Shipping.

Good luck,

Joker

 

[grimlykindo]

Yeah I RMA'd a Leadtek 6600gt with the exact same problem. They were out of stock so I refunded it and got a x800 xl.

Do the SAFE thinkg and RMA it!!

 

[InternetDuder]

aww and the Leadtek was so big...

I guess i'll try a different brand. Any brands to watch out for? I am thinking AOpen

 

[Rising Phoenix]

Originally posted by: InternetDuder
I got overheating-like syndromes in games with my Leadtek 256mb 6800GT. Everything is stock clocked. The game would run fine, then stutter every now and then, getting more and more often, then it'd do one of 2 things:
-Bluescreen, giving an error about nvidia. When windows (xp) came back, it'd blame it on my video card.
-Hang for like 30 seconds, then go back to desktop in really crappy resolution (640 iirc), 16 colors, with an error saying something involving "a device has failed".

Seems like a case of video card overheating, but then I ran prime95 and it failed. I redid the CPU/HSF contact, and prime95 passed twice after that (20 hours each time), so I get to blame the video card again.

Memtest86 was also flawless.

I also was told it might be Dual channel messing up, so I tried one stick of ram instead of 2 (tried each stick) and the stuttering still existed.

My temporary solution is to blow a table fan on the open case, which solves the problem.

The case is well ventilated, and should not pose a problem to the video card.

NView (which has lied to me about voltages) claims the temperatures are in the normal range. I don't remember the temps exactly, but in a previous topic, I was assured the temps are normal.

I underclocked the processor multiplier, and the problem stayed.

I could not underclock the FSB.

So here I am. I am ready to RMA the video card (its newegg, I assume the policy is good), but I do want to be sure about it. If anyone thinks there is anything else I should try, please let me know. Thanks.

Leadtek 256mb 6800GT
AMD Venice 3200+
DFI Lanparty UT Ultra-D

This is interesting...

I had to do the same as you. My Leadtek 6800GT went completely out. So, I RMA'd it. I did notice that the card was very warm - hot to the touch. My thought, "Must be the heatsink fan went out."

So, in the time being, I reinstalled my old ATI 9700 Pro card. Now here's the kicker...THAT card is doing the same thing! Only the difference is that for a little while, it would lockup with the video still showing but locked up and the sound going for a few seconds then, blank screen and some weird sounds. Hard reset, the computer is going again. But, if I started to play BF2 again or just play WoW for that day, same problem happens again but now, it is happening in WoW when I NEVER had that happen before!

Theory: Heat.

My question is, this is my old card...it didn't do this ever before. The only difference is that I have BF2 demo installed and the problems are rolling in.

I have read about some posts about battery backup and power ratio's. I have a Antec True Power 430. My APC is rated for 350 mA.

I have a A64 3200, 1 gig of mem, 2 WD 80HD (one SATA and one ATA100), DVD Rom and a DVD burner, Creative Aud Gamer, Winfast TV tuner card and now my old 9700 Pro.

Any ideas?

 

[InternetDuder]

I don't know if power is an issue. I have one DVD burner, one SATA drive, a sound card and a video card, all on an Antec 450w power supply. Have you tried unplugging some components (dvd drives) and see if that does anything?

I know my problem is heat with something, because blowing a fan on the case solves the problem! But its a temporary solution that doesn't addrses the culprit. I don't see how insufficient power could lead to overheating being a problem.

 

[Rising Phoenix]

Originally posted by: InternetDuder
I don't know if power is an issue. I have one DVD burner, one SATA drive, a sound card and a video card, all on an Antec 450w power supply. Have you tried unplugging some components (dvd drives) and see if that does anything?

I know my problem is heat with something, because blowing a fan on the case solves the problem! But its a temporary solution that doesn't addrses the culprit. I don't see how insufficient power could lead to overheating being a problem.

Well, in all this, I FINALLY found the problem...and it was a heating problem. BUT, it wasn't something I did. The Antec 430 TruePower PSU have a thermal power switch. When hooked up to the motherboard, it reads the temps off the board and itself and regulate the fan speed according to the need of cooling.

It looks like the temp senser is out because, I did notice that the case fans were running low and never did speed up. Then it hit me to monitor it. Yup, the temps got higher in the case but the fans never went up. Matter of fact, the case fans are running at 1400 rpm and they should be running much higher because of the heat then kick down to keep the temps down. Anyhow, I disabled the thermal switch and now the CPU is at 36C and the motherboard is at 25C. One crip...the sound is louder now. I have a ASUS board with "Cool and Quiet" tech. Has anyone tried this and is it good or should I just live with the "jet engine" I got now?

Thanks

 

[jelifah]

Holy Crap!!

Maybe that's my problem Rising Phoenix. I recently had to an RMA an Antec Power Supply and instead of returning the SL-350 I sent, they gave me one of the 350 True Power 2.0 Power Supplies. I was getting the same sort of sound stutter and then Blue Screen with a Machine Check Exception, or a reboot that launched the Microsoft website nVidia driver error', while playing WarCraft 3 or MLB2005

You mentioned a thermal power 'switch'. Where was it located, and how did you disable it to make the fan spin all the time?

 

[Navid]

Don't use the temperature reading in the display properties. By the time you get there, the temperature drops a few degrees. Use RivaTuner to log the temperatures. Then, you can see how the temperature varied while you were playing.

 

[Rising Phoenix]

Originally posted by: jelifah
Holy Crap!!

Maybe that's my problem Rising Phoenix. I recently had to an RMA an Antec Power Supply and instead of returning the SL-350 I sent, they gave me one of the 350 True Power 2.0 Power Supplies. I was getting the same sort of sound stutter and then Blue Screen with a Machine Check Exception, or a reboot that launched the Microsoft website nVidia driver error', while playing WarCraft 3 or MLB2005

You mentioned a thermal power 'switch'. Where was it located, and how did you disable it to make the fan spin all the time?

Just unplug the small 2 wire plug that runs from the PSU and bundle it up with the wires so it is out of the way. It will run like a normal PSU then. Also, plug the case fans into the available fan connectors on the motherboard.

Hope this helps.

 

[Rising Phoenix]

Originally posted by: Navid
Don't use the temperature reading in the display properties. By the time you get there, the temperature drops a few degrees. Use RivaTuner to log the temperatures. Then, you can see how the temperature varied while you were playing.

Will do, thanks Navid!

 

[Frown66]

What are the safe temperatures in the 6800GT's? I've read anywhere from 70's to 100's (Celsius). Even 70's sound scary to me. My Leadtek 6800GT goes up to the high 70's I think, but I don't experience the problems the original posted stated (yet). Granted, I've only have it for about 2 months.

 

[rise]

phoenix, so you had your case fans on the fan only line?

for instance my tpII has the 2 pin monitoring wire but its only to read the rpm of the fan through the mobo. i think its only when you use the fan only connectors, if you have them, that will regulate the case fans rpms to the psu's.

 

[Rising Phoenix]

Originally posted by: rise4310
phoenix, so you had your case fans on the fan only line?

for instance my tpII has the 2 pin monitoring wire but its only to read the rpm of the fan through the mobo. i think its only when you use the fan only connectors, if you have them, that will regulate the case fans rpms to the psu's.

You are correct. I was just typing fast and not thinking. For full speed, just connect the fans to a main feed or to the motherboard.

Here is a link to the same topic for BF2 about cooling and my testing of a theory:

http://forums.anandtech.com/messageview.aspx?catid=31&threadid=1612047

Hope this helps!