8170 GTX485M overheating issue

Hi,

I have a Sager 8170 with a GTX485 video card that has started misbehaving, shutting down the laptop whenever it's under any sort of load.

I have pulled the case and vaccumed all the cat fluff out of the heatsinks etc and reapplied thermal paste but it's still overheating. The fan is clean, turns on and runs correctly but the large silver heatsink plate attached to the video card gets too hot to touch very quickly and the laptop still shuts down.

GPU monitoring software says my GPU temp is 45° while running on the PSU with a monitor plugged in and no LCD on but if I start TF2 it quickly jumps to 96/97 for a while before shutting down after I stop paying attention to it!

I've tried different driver versions but the problem persists. i'm at the stage now of looking for a replacement card but am reluctant to pay for a new one until i'm certain it's the card.

Any other ideas to try? Is GPU overheating a common problem if your graphics card is on its way out?

 

Could be a heatpipe 90 the heatsink itself is fault if the paste job is good.

 

I did think about that, but how do heatpipe / heatsinks go bad if they have no moving parts?

 

They're filled with a Phase Change fluid, from what I was told, that instantly turns into a gas when it gets hot and helps the transfer of heat from the heatsink core to the fin-array.

A good way to test to see if the heatsink is bad is to do the hot water maneuver. Remove the heatsink and get a pot of hot water, then dip the part of the heatsink that attaches to the graphics card. It should instantly grow almost as hot as the water if it's in working condition.

It's somewhat similar to the Vapor Chamber technology in the GTX 690 desktop card.

 

thanks i'll try that!

https://onedrive.live.com/redir?resid=8DF089A0F6D6EC83!495&authkey=!ANi6aAYVyeQokew&ithint=file,.txt

Link to the temp log for the GPU, it was working fine this morning..played a bit of TF2 (90 degrees) then sat and did a bit of work (45 degrees) then started up TF 2 at lunch time... shutdown almost as soon as the card is taxed.

I zapped the heatsinks with an IR temp gun after the last sutdown, GPU heatsink, 60 degrees, heat pipes at 51 and 52 degrees GPU temp according to the monitor was 130

 

You could try repasting it. Too much or too little will cause high temps, you want to put about a pea size amount in the middle and let the heatsink spread it by pressure when you put it back on.

 

If we consider the actual size of a typical pea, that's generally too much paste for a mobile chip. A "BB" is sometimes used as a reference for the amount you should apply. As long as you have good, even spread, and it's not excessive, you can use any number of application methods.

 

that sounds about right if you had the temp monitor set to Fahrenheit and not Celsius like the heat gun.
52 degree Celsius = 125.6 degree Fahrenheit

 

If its ok then bad it could be the heatsink mounting has an issue.