Video cards keep dying?

[kungfu5150]

Greetings,

My siutation is as follows. A couple years ago I had a evga GeForce 6600GT, eventually the card crapped out on me from heating related issues. I RMA'd the card and got a new one. That one lasted about 6 months, then it died as well. as to why I honestly have no idea. I blamed it on crappy refurb.

Enter present day.. about 4 months ago I got a new power supply - Enermax Liberty ELT400AWT 400W - and picked up a new Connect3d Radeon X1900GT 256mb. I decided to go with the new power supply to support my new card. Great power supply.. Card.. otherwise. The card just recently ate it. Not entirely, but mostly. Esentially where it stands right now is, games run but I get artifacting.. not heavily, but random streaks, etc. For the most part, its playable just annoying. My GPU temps never exceed 71C. If i run artifact scanner with AtiTool, it fails immidiatly. If i run the artifact scanner with Tray tools VPU recover kicks in right away. Ive tried this with a different, older ATI and they both run teh artifact scanner fine. A while back, my GPU was getting up to 90C. I added some new thermal paste and fans, and now it never goes above 71c as stated.. but im affraid its too late and I fried the memory or GPU.

Im tired of going through graphics cards. I've put in a RMA request to Connect3D but its been 4 days and they havent answered any of my emails.

Im starting to to think something else is wrong with my hardware, but everything checks out.. +12v rail is 12.68v usually.. I think cooling may have been the death of my cards, but i have a hard time beliving it was the cause of all of them going.. I live in Canada, its not exactly hot here.

What do you all think?

Edit: Im not overclocking anything at all

 

[tungtung]

To me it looks just a very very very bad coincidence.
I assume those video cards are brand new right (at least the one you bought).

Could it be bad electricity? I don't mean the PSU but the power from hydro, do you have a good surge protector and such?
All this time are the rest of your other components stuck the same?
I mean the other possibility is something wrong with the graphic connector (that AGP or PCI-E connector) on your board.

I would say for now when you get your new card, try get better ventilation on the video card area. More fan or better fan arrangement?

 

[kungfu5150]

Originally posted by: tungtung
To me it looks just a very very very bad coincidence.
I assume those video cards are brand new right (at least the one you bought).

Could it be bad electricity? I don't mean the PSU but the power from hydro, do you have a good surge protector and such?
All this time are the rest of your other components stuck the same?
I mean the other possibility is something wrong with the graphic connector (that AGP or PCI-E connector) on your board.

I would say for now when you get your new card, try get better ventilation on the video card area. More fan or better fan arrangement?

Good advice, thanks. Brand new cards yes, except the one RMA I did.. i imagine that was a refurb they sent me. I do know to keep in mind clean power vs dirty power. For the life of my two 6600gt's i had the system running on a APC UPS 750VA.. so should have been no issues at all, then i moved. Im not using a battery backup anymore, but i do have a decent APC power surge protector coming off my wall socket with everything in that. None of my other components are acting up though, so its probably not dirty power. I've had the same processor and ram the entire life (2+ years) of the system, no problems.. upgraded my board once but only because a buddy of mine gave me a better one then i had.

So youre probably right, just baddddd luck. The current card in question was always a little quirky.. For example, when i play WoW in windowed mode, if i use 4x anti-aliasing, and i hover my mouse over something in my start bar, when the tool-tip tries to pop up the system and game lags horribly.. its bizzare.. (with clean windows/driver install)

Im going to try to test the card in my girlfreinds rig and see if i get the same artifacting with ATI tool.. i imagine i will.
Aside from that, anyone have any expirience dealing with Connect3D's RMA dept? Im having no luck..

thanks for the advise.

 

[lotharamious]

Make sure that you're using power that's GROUNDED. I've had too many weird issues at a friend's place where he has no grounded power, and stuff just decides it's gonna die, i.e. psus, video cards, hard drives, even a crt monitor. We all decided that it had to do with his ungroundedness, so we grounded everything using unorthodox methods, and the issues went away.

 

[tungtung]

Well if your card already give you the artifact and whole crap I doubt trying it on different computer will eliminate the problem, unless the problem is not with the card but somewhere else. However the only thing I could imagine that will damage a video card other than dirty power would be heat. I mean those software as far as I know only monitor the GPU temp, and not the rest of your video card parts (memory chip, etc). So it is possible the heat is what is killing your card. I'd say if you can maybe try running your system without the side panel (open case), and see if you have any problem (with another card obviously).

Regarding grounding problem, well it is entirely possible, however given the fact you said that you moved between the first incidents and the most recent ones, kinda doubtful that grouding is the problem. I mean what are the chances if having two houses with bad grounding, but then again what are the chances of 3 video cards dying.

Honestly I am clueless on this, but my gut is telling me now it could be heat damage on your card.

By the way did you experience any problem with the system before the whole artifact problem shows up (or before the other card died). Problem like your system freezing while playing games, unexpected slowdown in games, etc.? Cause that is some early indication of heat problem. I know my friend's old 9500Pro almost got burned to crisp when the fan died without warning, the only warning he got was slowdowns and random system freeze.

 

[kungfu5150]

Thanks for everyone's input on this issue. I believe ive nailed it down to a heating issue.

I used another card, same model. Everything works beautifully.

The problem is, the dead card's puny freaking heatsink and fan. The new card i borrowed and tried in my system has a nice beefy heatsink with seperate exhaust that takes up an additional expansion slot. The card never gets above 64c in full load with fans at around 55%. Artifact scanner ran for 3 hours @ 70c tops.

Before I realised my old cards problems, it used to get to 90c.. And thats after I discovered those temps. I ran it months like that, and the Cheapo Connect3d couldnt handle that.. nor could any card for that matter. It was probably defective out of box, as I dont see how any card should run that hot right out of the box. I'll try to get an RMA on it.

so moral of the story... dont buy cheap cards. Or if you do, make sure they have a nice big heatsink+fan on them. I have no idea what crack Connect3d was smoking when they designed this card with such a whimpy HS&F. Also make sure you have adequate case cooling as well... dont let this happen to you