How to remove Itanium "thermal grease"?

[SVSKM5]

Does anyone know how to remove the "thermal grease" from an Itanium I microprocessor? The "grease" used appears to be a gray silicone type material. I believe it is referred to as a "phase change" thermal compound. It is very hard when cool and supposely softens when the chip heats up. The "thermal grease" acts as a glue and it is nearly impossible to get the built-in heatsink off the chip. I've tried all the "normal' solvents that you would use with "normal" grease, but no luck. I have even tried heating it up to get it to melt, but can't get that to work either. Any ideas?

[JAC]

try freezing it, then mechanically assisting it off with a blunt object... you can try a freezer, but I think you might have better luck with that rapid cool down spray you can get for diagnosing electronic faults... it typically goes down to -50 degrees C. ( just over twice as a cold as a good deep freezer ).

[FDIV]

I have had a great deal of experiance with the itaniums. The biggest problem when removing the heatsink is the fact that it is so broad and flat that the grease creates a vaccume whenever you try to pull it off and so it holds fast. Usualy I slide a utility knife between the heatsink and the processor (make sure to angle the edge twords the heatsink as the heatspreaders are made of soft metal and will scratch). This breaks the vaccume and the heatsink comes right off (I'm assuming you removed the screws attaching it first.) As for removing the pesky compound I usualy first scrap the majority of it off with a wooden scraper (a sharpened paint stir stick) I then use goo-gone on a paper towel to wipe up the rest. The goo-gone does disolve the thermal grease quite easily. As you mentioned it is interesting stuff though. I scrap these itaniums at my local scrap yard outside and the hot sun sometimes completely liquifies the grease.

Hope this helps-
fdiv

[SVSKM5]

Thanks for the responses. I will give both methods a try.

On the method of using a knife to break the seal between the heatsink and process, have you had problems with the cache chips still "sticking" after the processor has been freed? or does the heatsink's freed angle just cause the other two chips to break free?

[FDIV]

Oh, I see what your trying to do. Your ripping appart the actual chip packaging to expose the processor dies. I did this once with one that was already mangled (taking it apart this far will mangle the package and destroy its collectable value.) The biggest problem you will encounter is the 5 sided torx screws with a post in the middle. I made a makeshift bit using a dremmel to grind down a larger hex wrench. Once you get all the screws out and the little clips on the bottom off its just a matter of prying. The heatspreader will probably get bent and theirs a small sealed end hiding in the plastic casing you have to be carefull with if you don't want to break it but really you just pry and eventualy it comes off. After this use a wood scraper and goo gone.

good luck.

[SVSKM5]

Yes, I'm trying to free the processor card for use in a display. I would like to keep the card from being damaged (operational in the sense, that it could be repackaged, theoretically). I was able to do it once by prying, but failed the second time. The chips came up with the thermal grease (amazing that the "grease" is stronger than solder). I haven't had any luck with goo-gone at least when there are lots of that "grease". I don't want to damage any more cards, so am I looking for the "magic" solvent that dissovles the "grease", or some way to get the heatspreader off. As an aside, I can't believe how heavy those things are