Need help with GTX 285 repair.

[mdnelson09]

Hey everyone,

This weekend i got a GTX 285 and the board prevents my mobo from turning on when a six pin populates the green circled plug (or both at the same time). if it populates the other plug, it will let the mobo power on and signal that it has no video output. Has anyone ever dealt with something like this?

Also i have a resistor on the board (like the one circled red) that doesn't look like its in the best condition, and I'm curious if should replace all of them, (if that could be causing the issue, or if its the next part.)I'm also reading up on how to test the mosfets on the board to see if any of those are bad.

The image below isn't my card, but mine looks like it.

Thanks for your help

 

[BarbaricSoul]

Hey everyone,

This weekend i got a GTX 285 and the board prevents my mobo from turning on when a six pin populates the green circled plug

http://img.techpowerup.org/130127/5.jpg

Sounds like you have a short in that connector somewhere that is causing your system not to boot.

 

[McSteel]

The components circled red aren't resistors, they're chokes (inductors). And as for the malfunction you're seeing, something is definitely failed-shorted, and most probably it is a power-MOSFET. Obviously, you can try checking the pins of the PCI-E connector itself for continuity, see if there is an immediately detectable short. If not, it's most likely somewhere after the filter capacitors...

 

[mdnelson09]

The components circled red aren't resistors, they're chokes (inductors). And as for the malfunction you're seeing, something is definitely failed-shorted, and most probably it is a power-MOSFET. Obviously, you can try checking the pins of the PCI-E connector itself for continuity, see if there is an immediately detectable short. If not, it's most likely somewhere after the filter capacitors...

Thanks for the info..
Not sure where to start first. I guess i could test the mosfet's with a multimeter. I just got a GTX280 in the mail (ebay) with some other issues. Ive put a few days in trying to research this stuff on both cards but sure enough, all i get are "Just RMA the card" or "buy a new card" threads online. I keep getting this feeling that I wont get the answers I'm looking for unless i change my major to electrical engineering.

 

[McSteel]

As i said, you could start by checking for shorts on the PCI-E connector pins first. The top row is all GND, and the bottom row is 3 x 12V + 1 x Sense/GND. You want to check if there's an immediate short because you'll want to know if any of the VRM input capacitors have failed. This could mean they were defective, or it could mean there was a feedback surge due to a failed MOSFET that knocked one or more caps out. If there are no immediate shorts, then the problem lies elsewhere, and could easily be a failure of a driver IC, a PWM control IC, or even of the GPU itself.

What problems does the GTX280 have?

 

[mdnelson09]

As i said, you could start by checking for shorts on the PCI-E connector pins first. The top row is all GND, and the bottom row is 3 x 12V + 1 x Sense/GND. You want to check if there's an immediate short because you'll want to know if any of the VRM input capacitors have failed. This could mean they were defective, or it could mean there was a feedback surge due to a failed MOSFET that knocked one or more caps out. If there are no immediate shorts, then the problem lies elsewhere, and could easily be a failure of a driver IC, a PWM control IC, or even of the GPU itself.

What problems does the GTX280 have?

The 280 will display fine, but spool up to 100% as windows is loading and then kill the display. my onboard debug shows that the system is booting fine, but the gpu isnt displaying anything. I have a hunch this card overheated once or twice in its life. Thinking about baking it.

 

[McSteel]

Are you sure the heatsink is making good contact with the GPU? Is there a sufficient amount of thermal paste between the two (but not too much)?

 

[qubit]

It looks to me like that card was killed with overclocking.

As you're having trouble even identifying what the components are (resistors for chokes) I think you'd be on a hiding to nothing; it's just not worth it, what with all the frustration you're likely to get from this.

It's quite possible that the GPU is dead and there's no way to test it without specialist equipment.

 

[majestic12]

oven trick maybe?

 

[mdnelson09]

I talked to the guy on ebay and since the condition of the 280 was what it is, he gave me a partial refund but let me keep the card. Good news is i can now try a bunch of things without feeling too guilty.

Are you sure the heatsink is making good contact with the GPU? Is there a sufficient amount of thermal paste between the two (but not too much)?

Paste looks like stock but at the same time it doesn't seem like it would overheat THAT fast. Ill try repasting it and running again tonight. Maybe it isnt that bad of a problem. XD

It looks to me like that card was killed with overclocking.
...

Probably...

oven trick maybe?

Ill try this and see if it works (after a re-paste)

Thanks for all the tips everyone

 

[mdnelson09]

[update]

ssooooo......

I went to bake the cards today after a re-paste the night before... this happened.

It separated from the card when I was removing the paste applied last night. Arctic ceramic might as well be glue. It was a clean separation for the most part but the paste was disgusting under the block so i cleaned it up and put my stuff on.
Needless today, my day was full of anger and lulz.
and then I made a GTX cake (cooked @ 195c for 10 minutes)
Fast forward 50 minutes:

Win.

The 285 is still shorting my board out so I'll try replacing the chokes and mosfets later down the road, but hell, my 280 works