Interesting read. If i understand correcly, the H100 cooler goes direcly on top of the inner CPU without its (IHS?) or cover?

Interesting read. If i understand correcly, the H100 cooler goes direcly on top of the inner CPU without its (IHS?) or cover?

 

Ofcourse not.. you need to re lid after applying the liquid metal.

Ofcourse not.. you need to re lid after applying the liquid metal.

 

A question about that (from an obvious newcomer to delidding) - is it necessary to use an adhesive to reattach the lid to the CPU, or does the liquid metal serve that purpose?

A question about that (from an obvious newcomer to delidding) - is it necessary to use an adhesive to reattach the lid to the CPU, or does the liquid metal serve that purpose?

 

 

I have not done this and I had asked the same question... and my understanding is, you don't need any adhesive... you put the lid back as it was and place it on the MB and then lock it with that CPU holder pin.

 

I am sure they could answer this better.

 

MAnny

A question about that (from an obvious newcomer to delidding) - is it necessary to use an adhesive to reattach the lid to the CPU, or does the liquid metal serve that purpose?

 

No, you do not need to reattach the lid to the cpu. Proper sequence would be something like this:

 

1. Make sure the case is horizontal on the desk, if you have a vertical case (normal tower case)

2. Apply TIM to cpu die

3. Place cpu into cpu socket on motherboard

4. Place IHS on cpu BUT make sure it's not centered. Place it a bit towards the locking handle side of the socket. The reason for this is that when you lower the locking handle, once it makes contact with the IHS, it will slide the IHS forward just a bit until it's fully locked. By offsetting it about 1-2 mm, the IHS will end up centered on the cpu once the handle is locked.

5. Once locked, apply TIM to IHS and install waterblock on IHS.

6. Done.

No, you do not need to reattach the lid to the cpu. Proper sequence would be something like this:

 

Thanks yet again. Related question - is it best to use the same TIM (Coolaboratory Liquid Metal Ultra) between the top of the IHS and the waterblock? Or would Arctic Silver be better? I'm sure I've read the answer somewhere (maybe even in this thread) but can't recall.

Thanks yet again. Related question - is it best to use the same TIM (Coolaboratory Liquid Metal Ultra) between the top of the IHS and the waterblock? Or would Arctic Silver be better? I'm sure I've read the answer somewhere (maybe even in this thread) but can't recall.

 

I recommend LMU on both DIE>>IHS and IHS>>waterblock... Just make sure the waterblock is NOT aluminum. LMU eats through aluminum.

Efrain Ruiz,

 

although it will not make that much of a difference, personally i wouldn't go for polished heatsink surfaces. I think manufacturers do that for a reason. The contacting surfaces are larger with raw material. The drawing below (have been awhile since i used Photoshop :) shows what i mean. Would there be no such things as cooler pastes, manufacturers would probably polish and apply more pressure with metal to metal contacts to get rid of the heat. But since there is a film of TIM between them, the surfaces better be raw.

 

I recommend LMU on both DIE>>IHS and IHS>>waterblock... Just make sure the waterblock is NOT aluminum. LMU eats through aluminum.

 

Thanks. Am feeling pretty much up to speed now. Time to go look for parts and equipment...

 

The contacting surfaces are larger with raw material.

 

Interesting. Makes sense in theory, but I'd be interested to know if others agree. I will say that I got a much better overclock the second time around, when I lapped the heatsink. But it wasn't an apples-to-apples test - I upgraded at that point from the 965x to the 980x and while thermal performance was drastically better, that might have been because of the particular chip, not the lapping.

I recommend LMU on both DIE>>IHS and IHS>>waterblock... Just make sure the waterblock is NOT aluminum. LMU eats through aluminum.

 

Thanks. Am feeling pretty much up to speed now. Time to go look for parts and equipment...

 

I don't think LMU is a good idea on the IHS-waterblock unless you are sure you will never take them apart.

The LMU bond is pretty strong from what I have read.

 

gb.

 

I don't think LMU is a good idea on the IHS-waterblock unless you are sure you will never take them apart.

The LMU bond is pretty strong from what I have read.

 

gb.

 

I'll let you know in 6 months.

I don't think LMU is a good idea on the IHS-waterblock unless you are sure you will never take them apart.

The LMU bond is pretty strong from what I have read.

 

gb.

 

That might give me pause - wouldn't mind having the option of moving to a revised-stepping IB at some future point if there were advantages. Though I suppose I could budget for a new waterblock as part of the upgrade...

Flat mating surfaces are best. See the Anandtech thread linked earlier and the detailed posts by "IDontCare" on the matter of Ivy Bridge thermal characteristics. Also one of the members here has lapped his IHS and H100 and is running similar voltage at 4.9GHz and has lower temps than me - and I have a full blown water cooling setup.

Flat mating surfaces are best. See the Anandtech thread linked earlier and the detailed posts by "IDontCare" on the matter of Ivy Bridge thermal characteristics. Also one of the members here has lapped his IHS and H100 and is running similar voltage at 4.9GHz and has lower temps than me - and I have a full blown water cooling setup.

 

Make that 5.0Ghz, Tech... ;-)

Make that 5.0Ghz, Tech... ;-)

 

Tech we most Lapping like Efrain :good: