FEATURED De-Lidding for LN2

[Johan45]

I'm posting this here and anyone can chime in with their experiences/techniques. I have never had good luck with Intel delids and LN2 I always seem to end up popping the TIM. I came across this post at ROG forums and looks like I might be doing it wrong. I never have tried using this much paste. If anyone has some tips on keeping a good contact I'm all ears. Pretty sure that's a pic of Dre8auer

With initial data obtained, CPUs are de-lidded to replace the stock thermal interface material (TIM). TIM choice is critical because the compound needs to be malleable enough to cope with the contraction and expansion of surfaces during extreme benchmarking. Eventual fractures are inevitable, but some materials perform better than others in this regard. Application methods are also a far cry from the conservative, rice-grain-sized amount required for conventional setups. A liberal quantity of paste is spread over the die and surrounding PCB to reduce the chances of dreaded ‘paste crack’ and to help improve thermal transfer.

 

[EarthDog]

Well that is a first for me to see it this way...

 

[ATMINSIDE]

Sub'd to see how this works, very curious method there.

 

[Johan45]

I know, looks pretty sloppy but at this point I'm willing to try pretty much anything.

 

[EarthDog]

That is a scary noise, that pop, isn't it?

 

[Johan45]

That is a scary noise, that pop, isn't it?

To me it's more disheartening than scary, I have wasted so much time/LN2. I got so pissy the last time I haven't benched Intel on LN2 since.

 

[jiccman1965]

I use the one that was posted by Dancap, was looking for it but can't find it. I put a very thin coat and make sure it goes over the sides of the die. That's how I do it and it has worked great. After the cpu is mounted again put another thin coat on IHS. After I have mounted the pot I heat it up to 70C then cool it back down and it has worked 99% of the time.

 

[jiccman1965]

De-lid

Hay founds some pic's of the de-lid, hope this helps.



After you apply TIM put into socket then put on the IHS, if you leave the old glue on it helps hold it in place.

 

[Alaric]

Subbed for knowledge.

Common theme seems to be smother that puppy. So that when the extreme cold shrinks it, it's still transferring heat effectively from the entire die's surface?

 

[Johan45]

Subbed for knowledge.

Common theme seems to be smother that puppy. So that when the extreme cold shrinks it, it's still transferring heat effectively from the entire die's surface?

IT's not so much the cold that's the issue, it's the heat produced by the die under heavy loads like XTU or CB15, which heats the TIM faster than the IHS can pull it away. The TIM expands and breaks the contact between the die and the lid. This was even further exaggerated now that Intel has made the PCB thinner which allows more flex so it's easier for the TIM to expand, less resistance from the PCB tohold it flat.

 

[jiccman1965]

Yes it can be touchy for sure trying to balance out the heat and cold, But the way I suggest above has worked well for me. I have been also using the KP thermal paste also.

 

[Johan45]

Yes it can be touchy for sure trying to balance out the heat and cold, But the way I suggest above has worked well for me. I have been also using the KP thermal paste also.

I got a bunch of that from Marc0053 when he went on hiatus. Man that stuff is super thick

 

[jiccman1965]

Yes it is but has worked well for me.

 

[Dolk]

[spolier]That die is going to crack[/spoiler]

 

[EarthDog]

Do tell....

 

[Johan45]

[spolier]That die is going to crack[/spoiler]

Why so mysteriously ominous, that's so unlike you

 

[MetalRacer]

I've been using the Dancop application method also but haven't tried heating it up to 70c before cooling it down. I'm going to try a Venom and a Trex pot to see if they work better than the FI Dark.

 

[Johan45]

I got one of those venoms from Marc0053 too. Haven't tried it out yet though

 

[Lochekey]

At the Philly party last summer a lot of the guys were using techniques similar to this and seemed to have decent results. I found what was working for me was to apply tim similar to what Jiccman had and then mount the IHS and apply the paste and mount the pot and only slightly tighten everything down. I was then running cinebench and would finish tightening the pot down and would look for a sudden temp drop on the cpu to guide when I had the right pressure and then woould go another half a turn on the thumbscrews. This seemed to work pretty well overall and only had minor issues throughout the weekend.

 

[Dolk]

To be honest, I'm surprised that the chip can be directly weighted down with pots as heavy as venoms. But again I'm always surprised by the resiliance of our chips each time someone pushes it further than any engineer ever thought could happen. Its like being a DM in a D&D campaign where your BBG gets killed by something as simple as flung dung.