GTX 690 not recognized after overheating

[kkarsaj]

Short and simple - today I was idiot (more than usual).

Subject: Dell/nVidia 690 GTX

Story - got this beautiful piece of engineering for low price in great condition and got distracted during cleaning & changing paste ending up with overheating after 3-4 min of testing and ... it's dead Jim. Beeeeeep bip bip bip bip.

Tried to troubleshoot it (stopped at shorts testing, couldnt bare the sadness and anger), but I decided to seek for some wisdom here.

I've marked mosfets and "pex8747 ba80gbc" SLI thingy on the attached photo where I forgot to put thermal pads. I would really appreciate if someone drop advice or two on what to test; if not, lets this be testament of "multitasking, what could possibly go wrong".

[cyberwillis]

Hello...

What exactly you did by distraction ? Just forgot to put pads ?
Have you tried to check the Fuses ?

[cyberwillis]

I don't think for 3 - 4 min it would cause your board just die because you forgot the thermal pads.
To me appear that the PCIEx have bad contact or the Fuse get blown.
Why I am asking you take a look on the Fuse F1 (close to the M3 memory)? I was used to deal with boards without turning off the power supply cable I was just taking off boards and putting again carelessly until last year I get a HBA board to deal with, and damn they are SO SENSITIVE. If you have a HBA board and your machine is with your power supply cable attached, 99% sure the two fuses will blown and you will not notice that and your HBA board will not even turn on. So even if you had stressed your GPU board and believing it was fine why a overheating of 4 min would make your GPU die ? Don't make sense. There is other part of story you not telling ? Like forgot to turn on the Fan or something else ?

[cyberwillis]

Also, I started to notice in your picture there is too much flux around the PLX chip (PEX8747), somebody tried to reball it? May be if that was the case, and the ball used was poor quality may be because overheating the process was undone. Because that area should be clean from flux.

[kkarsaj]

Hi

tnx for reply. I did check F1 and it's OK and the flux, that was me as well (just a quick heat up, not reflowing or reballing).

It happened just as I described - i forgot to put pads on mosfets and PLX, everything else was OK. Card worked for - how much time is needed to boot up and finish 50% of 3DMark Vantage (less than 5min) - then died (screen went black and after restart - beep bep bep bep). Tried on few different MBOs, same thing and GTX 580 works fine in that system where this one died.

Now im waiting for new PSU (other one with 2x8pin cables is in my sons computer) to continue...

One more thing - now, when powered on, even without fan and metal parts, its completely cold as much as i can feel (don't have thermal cam). Also, fan is spinning when connected and the logo LED has power

[cyberwillis]

Yeah if another GPU works fine in that system is not related to the MOBO itself nor the PCIex.
The power supply failure is possible, just a note this GPU needs 300W for it alone so preferable use a PSU that has 450W to Up. At this point is hard to tell if is something got damaged on the power connector or not. Can you verify if there is shortcircuit in line 12V?

[kkarsaj]

Yeah, no shorts found. I used Antec 750W when this thing died (same PSU is running happily with AMD FX8150 & GTX 580 now).

tnx for now, i have to wait for PSU before i continue with troubleshooting.