So my GPU just lit on fire

[BallaTheFeared]

http://oakridgetoday.com/2013/03/13...y-cray-re-soldering-supercomputer-connectors/

 

[Cookie Monster]

Lead free solder, read about nvidias bumpgate. Overheating computers.

Regarding bumpgates... its not as simple as you'd like to think. By that I mean it isn't simply "high-lead" bumps vs Eutectic bumps something that Charlie was trying to portray the problem as.

Plus the only advantage I can think of for "lead-free" solder is the.. ROHS compliance hah.

Anyway, OP show us the pics!

 

[notty22]

Do they have more solder problems? Do you have a link to whatever you are referring to?

Is Kidsurf.com not working for you?

 

[wand3r3r]

Avoid the issue, typical nvdiot (focus) response. Is your sig out of date or is that all you get from nv now?

 

[wand3r3r]

http://oakridgetoday.com/2013/03/13...y-cray-re-soldering-supercomputer-connectors/

Ok.

Nice signature.

 

[tweakboy]

Its ok man spend 120 dollars get a 460 1GB it will be 1000x faster then the cheapo 9600 GT. thx gl

 

[Silverforce11]

http://oakridgetoday.com/2013/03/13...y-cray-re-soldering-supercomputer-connectors/

This is absurd, why would they cheap out when they are building supercomputer parts for millions $$? WHAT THE HELL?

 

[lavaheadache]

This is absurd, why would they cheap out when they are building supercomputer parts for millions $$? WHAT THE HELL?

cheap out? The issue was that there was too much gold and it was affecting the solder joints.

 

[cbrunny]

Anyway, OP show us the pics!

Will do. Have to take of the heat sink, which I've never done before (little bit excited to do this). I'll post the pics soon.

 

[BallaTheFeared]

cheap out? The issue was that there was too much gold and it was affecting the solder joints.

Probably 14k walmart gold :hmm:

Does anyone still use lead? I hope not!

 

[lavaheadache]

lol, fools gold

 

[BallaTheFeared]

lol I typed that out first, but I didn't want anything to think I was saying AMD were fools (somehow), this place is always ready to pounce!

 

[cbrunny]

Here is the close up shot with the heatsink off.



Uploaded with ImageShack.us

Full card shot with heatsink off. Looks like the thermal paste turned into dried up chips. It provided zero resistance when taking off the heatsink.



Uploaded with ImageShack.us

This is my dog. He did not light on fire. But he was happy that it blew, because it meant I couldn't waste time sitting at my computer. He got a nice walk out of the deal.



Uploaded with ImageShack.us

 

[BallaTheFeared]

Pretty sure that's a vrm that popped.

 

[cbrunny]

Pretty sure that's a vrm that popped.

haha ok. I know basically nothing about the components on a card.

 

[lavaheadache]

I'm actually surprised to see that it was bare too. esp. on a card of that caliber. It was bare right, as in not cooled by the heat sink?

What a funny looking design btw

 

[lavaheadache]

lol I typed that out first, but I didn't want anything to think I was saying AMD were fools (somehow), this place is always ready to pounce!

It's all in the delivery. I will admit, you've come a long way. You seem a more reserved. More like BallaTheRefined

 

[Shmee]

Cute dog. Sorry about the card though :C

 

[cbrunny]

I'm actually surprised to see that it was bare too. esp. on a card of that caliber. It was bare right, as in not cooled by the heat sink?

What a funny looking design btw

No, there is a heatsink. I had to remove it to get a decent picture. The heatsink is one piece though and covers the entire card. It's a passive card: https://www.asus.com/Graphics_Cards/EN9600GT_SILENTHTDI512M/

 

[lavaheadache]

lol, I realize that the card has a heatsink, but does the heat sink make contact with those vrm's? That was what I was saying. I'm inclined to think not by the layout with those large caps in the way

 

[cbrunny]

lol, I realize that the card has a heatsink, but does the heat sink make contact with those vrm's? That was what I was saying. I'm inclined to think not by the layout with those large caps in the way

Oh sorry, I misunderstood.

No there is no contact between the heatsinks and the vrm. The heatsink extends over it, but it stands above by probably 3/8ths of an inch.



Uploaded with ImageShack.us



Uploaded with ImageShack.us

 

[lavaheadache]

And that is why we had failure.... Those vrm's likely ran very hot. That was a fan less card? If it was, that was an idiotic move on Asus's part not to have any kind of cooling on that power delivery.

 

[coffeejunkee]

Yeah, a little 1$ piece of aluminum on those vrm's would have been neat. There are even holes for the screws so it's really stupid not having a heatsink there. But what about case airflow?

 

[Cookie Monster]

Looks like one of the fets blew on one of the VRM phases.. I don't know if its due to there being no heatsink (usually bare designs like that means thermals are OK and its heatsinked onto the PCB board where they is alot of copper planes surrounding the power stage area).

Then again maybe it is.. depends on the case air flow I suppose. Heat could have been "trapped" there leading to the imminent failure. Im surprised the fuse didn't blow (unless theres none on that board) because something made the fet pop due to excess current..

 

[lavaheadache]

You will not be able to convince me that a card of that magnitude is ok long term to have bare vrm's with a passive heat sink design.