Sanded down the IHS of a Q6600 to the silicone.

[CryptiK]

IMO that would generate way too much heat and if you cut too deep, it's all over in a fraction of a second. This kind of thing is much better off to be done by hand.

[STEvil]

reminds me back when we were lapping Athlon XP cores lol

[Brian y.]

That is one amazing looking chip.



My hats off to you sir for your patience

[largon]

By the way a lot of people seem to think that the core and the IHS are soldered and this is simply not true. They use some hind of very strong silver adhesive that is no fully solid.

That "very strong silver adhesive" is the proverbial "solder". It is a metal alloy containing tin and indium (probably) and it has a melting point around 100ºC.

Here's a pic I took that shows how it looks:

[Ruslan]

That "very strong silver adhesive" is the proverbial "solder". It is a metal alloy (probably) containing atleast tin and indium and it has a melting point around 100ºC.

Here's a pic I took that shows how it looks:

I stand corrected. I didn't think it was solder because it is not completely solid.

[TheKarmakazi]

Awesome work! Bet your arms were sore the next day

[Ruslan]

Awesome work! Bet your arms were sore the next day

My muscles were fine but my blisters are still healing. No pain no gain

[Boogerlad]

largon, didn't you say that jewlery polish eats it off and gives it a mirror finish?

[eternaljammer]

pretty awesome mate!

[-n7-]

Impressive stuff.

[xdrift0rx]

so are you saying you WOULDNT reccomend doing this with a waterblock or that the waterblock just needs to be lapped too after lapping the processor? i lapped my Q6600 and my one core is still a higher temp than the other. this is my temperatures as i type this right now..




i would love to do this if i could. i think i could get a nicer clock out of it.


forgot to mention my waterblock is the swiftech storm rev 2.

[JaVa_Az]

Depends on the water block .Some of the newer ones are bowed so they mount better with Intels Bowed IHS`s. Not sure about that swiftech you have, but im sure with some google searches you could find out if it was made bowed or not. I would not worry to much about the idel temps on the cores. If your Q6600 is like mine the temps under load will even out. I would say everest is wrong on your core temps too. Your running a old version Of everest At that time they believed the TJmax was 95c when it is now believed it is 100c for a Q6600 Go. Still nothing official from Intel on 65nm CPU's so it is still just a guess. Newest everest release is 4.60.1500

[Boogerlad]

storm is not bowed. However, swiftech storm doesn't like mcm cpu's. A d-tek fuzion v2, EK surpreme or a swiftech apogge gtz would be much better in my opinion.

[xdrift0rx]

that one core just gets so much warmer than the rest. it sucks.



here it is with a load.



are you saying the storm doesnt like MCM's because of the jet design? maybe i didnt lap my cpu flat enough? also my VID sucks. i used core temp to check the q6600 i use at work and it gets 1.250..im tempted to switch it out some day...xD who would tell teh diference??

[largon]

largon, didn't you say that jewlery polish eats it off and gives it a mirror finish?

Yup.
I have some ancient silver/jewelry polishing paste that has the ability to sort of decay the tin-indium-solder. I just apply it on the stuff and the solder literally disintegrates into black dust in seconds leaving the die with a shiny mirror finish.

[xdrift0rx]

hold on, so does that mean you could cut open a slit in the rubber surrounding the processor and drop some in there and then be able to pop off the IHS? or would it mess up the core itself in some way? just wondering..

[JaVa_Az]

hold on, so does that mean you could cut open a slit in the rubber surrounding the processor and drop some in there and then be able to pop off the IHS? or would it mess up the core itself in some way? just wondering..

You can get the IHS off people have done it but they allmost all have ended up cracking the cores. without the shim in there the cores get just a little bit too much pressurewhen mounting the heatsink they will crack. I believe there is a post in the beginning of this thread showing pictures of failed cores taking off the whole IHS. I would say with a water block you could get away with it if your careful but any kind of decent air cooling heatsink is going to have too much weight to it. I think if you did pop it off and glue you some shims juat a tiny bit higher then the cores back onto it. then lap till you even would be the fastest easyest way to do this mod.

[Boogerlad]

Yup.
I have some ancient silver/jewelry polishing paste that has the ability to sort of decay the tin-indium-solder. I just apply it on the stuff and the solder literally disintegrates into black dust in seconds leaving the die with a shiny mirror finish.

Do you have a link, know the product name or the composition of that jewlery/silver polish?

[largon]

Boogerlad,
The tube looks like it's like 30 years old the instructions are in german 'n' there's no composition listed on it. Brand is "Gildenglanz".
:\

[reno 456]

This is a great idea...I wonder if it would work for a P4? Or a dual core?

[Bandito]

I might try tomorrow to achieve the same outcome by etching or even electro-etching. I'll just mask the sides and back with tape. I'll try on a celeron first, as I think the silicon will also etch. Hopefull will reduce effort, as only the sides will need to be sanded.

[Alexontherocks]

You can get the IHS off people have done it but they allmost all have ended up cracking the cores. without the shim in there the cores get just a little bit too much pressurewhen mounting the heatsink they will crack. I believe there is a post in the beginning of this thread showing pictures of failed cores taking off the whole IHS. I would say with a water block you could get away with it if your careful but any kind of decent air cooling heatsink is going to have too much weight to it. I think if you did pop it off and glue you some shims juat a tiny bit higher then the cores back onto it. then lap till you even would be the fastest easyest way to do this mod.

I kept both a e6600 IHS less and a q6600 IHS less with a ultra 120 for respectively 4 and 9 months. Many remounts, never a problem with neither of them.

I think that if the mounting is even in therms of pressure, nothing should happen. Then again I only speak from my persoanl experience

[Bandito]

Damn thats way too much work! Spent the last 2 days on it. Tried etching, electro-etching, 1.5 hours of sanding. This is a celeron 420. I think next time, I will pop off the whole ihs, then glue on a 0.5mm thick copper ring and lap that back to the height of the core.

[Commander_HK47]

Disregard this post, as it was made with out investigation on my behalf.

[Bandito]

Later this week, I'm going to try to dissolve the glue that sticks the ihs on to the circuit board by soaking in Methyl Ethyl Keytone (MEK). If it does, it could make it easier to remove the ihs.

I am going to do this to a E8500 or E8600, does anyone have a pic of the inside of one of these?

And would doing this help when using a phase cooler?