So I delided my 8700k....

[martin-w 3,235]

Just now, HiFlyer said:

I thought about it....... but it kinda felt like cursing in a church... inappropriate.

And yes, I guess I am strange, that way. 😋

 

😀 You're as bonkers as I am.

[HiFlyer 9,359]

So after thinking about it a bit, I went back in, yanked the lid of and tried again. The first delid was my learning experience, and this time I felt more confident. The change was slight, but I got all cores (HT On) under 80c (realtemp/Coretemp/cpuid) after two hours of Prime95

I think I'll leave it alone now.

First Try

 

2nd try

[lownslo 592]

1 hour ago, HiFlyer said:

I think I'll leave it alone now.

Indeed!  Temps look very good now!

Greg

[HighTowers 1,434]

I like the pad idea too. May consider it. Thermal Grizzly Kryonaut is quite good yet its thermal conductance is only 12.5 W/mk  vs that IC Graphite pad at 35 W/mk. Not sure how that translates in terms of temp differences but thats halfway to liquid metal which is usually around 70 W/mk

Just looked at some charts. Seems the Thermal conductivity numbers are negligible.  Application and its composition are what matters.  

Good thing the pad doesnt slide out as its electrically conductive as well. 

Edited November 13, 2018 by HighTowers
Clarification

[martin-w 3,235]

17 hours ago, HiFlyer said:

So after thinking about it a bit, I went back in, yanked the lid of and tried again. The first delid was my learning experience, and this time I felt more confident. The change was slight, but I got all cores (HT On) under 80c (realtemp/Coretemp/cpuid) after two hours of Prime95

I think I'll leave it alone now.

 

 

 

What did you do different this time, or nothing? 🙂 I have heard that some people get lower temp when NOT applying to underneath of IHS, whereas others say the opposite.

[HiFlyer 9,359]

1 hour ago, martin-w said:

 

 

What did you do different this time, or nothing? 🙂 I have heard that some people get lower temp when NOT applying to underneath of IHS, whereas others say the opposite.

My first try, I did not completely remove all of the previous factory gasket material. I simply pared it down a bit, figuring I would be replacing lid in the same spot so it would make little difference. 

The second try. I got a razor blade and removed every last bit I could.

My first try, I used a slight bit more liquid metal, because one bead seemed a bit to little.

The second time I stuck to the one bead no matter what.

The first time I applied gasket material around the entire edge of the lid (except for a tiny gap to let heat escape)

The second time I applied just to the "wings" of the lid, and spread it with the edge of the razor into a super thin coat.

[martin-w 3,235]

17 hours ago, HiFlyer said:

My first try, I did not completely remove all of the previous factory gasket material. I simply pared it down a bit, figuring I would be replacing lid in the same spot so it would make little difference. 

The second try. I got a razor blade and removed every last bit I could.

My first try, I used a slight bit more liquid metal, because one bead seemed a bit to little.

The second time I stuck to the one bead no matter what.

The first time I applied gasket material around the entire edge of the lid (except for a tiny gap to let heat escape)

The second time I applied just to the "wings" of the lid, and spread it with the edge of the razor into a super thin coat.

 

 

Yep, pretty much what other have found. I might try de-liding again but with no liquid metal on the underneath of IHS. Some  have discovered its cooler.

[Ted Striker 338]

2 hours ago, martin-w said:

I might try de-liding again but with no liquid metal on the underneath of IHS

Are you saying bare contact or another paste instead of liquid metal?

Ted

[martin-w 3,235]

36 minutes ago, Ted Striker said:

Are you saying bare contact or another paste instead of liquid metal?

Ted

 

 

No Ted. The usual method is to apply liquid metal to not only the die, but also to the underneath of the IHS. Some find this better for temp and it's the method most used. However, I have seen reports from some people that have achieved slightly lower temp by just applying the liquid metal to the die. In fact it was briefly mentioned on Gamers Nexus, they found it to be the case with one of their CPU's. 

Bare contact would be absolutely disastrous and result in instant overheating and shut down. 🙂

 

[lownslo 592]

2 hours ago, martin-w said:

...try de-liding again but with no liquid metal on the underneath of IHS. Some  have discovered its cooler.

I'd be curious how that works out.  My current 8086K had a high corner on the die!  I've delidded quite a few CPU's over the years but that's a first for me.  Had to grind the top of the die flat... tricky work!  I put Coductonaut between the die and IHS (the latter was very flat on the inside surface) but I had thought about not using any TIM between the die and IHS when it was time for reassembly.

Greg 

Edited November 14, 2018 by lownslo
Martin replied with an explanation for Ted while I was posting.

[martin-w 3,235]

8 minutes ago, lownslo said:

I'd be curious how that works out.  My current 8086K had a high corner on the die!  I've delidded quite a few CPU's over the years but that's a first for me.  Had to grind the top of the die flat... tricky work!  I put Coductonaut between the die and IHS (the latter was very flat on the inside surface) but I had thought about not using any TIM between the die and IHS when it was time for reassembly.

Greg 

 

No Greg! As I explained above, you would be bonkers to use no liquid metal, it would shut down instantly due to overheating. Please don't do that!!! 😀

 

See below... Or my previous post.

 

Quote

The usual method is to apply liquid metal to not only the die, but also to the underneath of the IHS. Some find this better for temp and it's the method most used. However, I have seen reports from some people that have achieved slightly lower temp by just applying the liquid metal to the die. In fact it was briefly mentioned on Gamers Nexus, they found it to be the case with one of their CPU's. 

 

Edited November 14, 2018 by martin-w

[HiFlyer 9,359]

8 minutes ago, martin-w said:

The usual method is to apply liquid metal to not only the die, but also to the underneath of the IHS

I tried that, but no dice. The underside of my ihs completely rejected the liquid metal. It beaded on top and absolutely refused to stick. I thought there might be oils or other contaminants causing the issue, but scrubbing the heck out of the ihs with alcohol showed no change. Conductonaut just ran off of it like water on glass treated with RainX.

[martin-w 3,235]

Just now, HiFlyer said:

I tried that, but no dice. The underside of my ihs completely rejected the liquid metal. It beaded on top and absolutely refused to stick. I thought there might be oils or other contaminants causing the issue, but scrubbing the heck out of the ihs with alcohol showed no change. Conductonaut just ran off of it like water on glass treated with RainX.

 

That's weird... no trouble with the CPU's iv'e delidded. And not seen anyone report that before. It is a bit tricky to get it spreading, but it flows after a bit of manipulating. 

[lownslo 592]

8 minutes ago, martin-w said:

Bare contact would be absolutely disastrous and result in instant overheating and shut down. 🙂

 

I agree not worth it... kind of like those folks who delidded and then mounted the CPU cooler to the bare die.    In the end I cleaned up the proud corner of the die and relidded with Conductonaut.

Greg

[HiFlyer 9,359]

13 minutes ago, martin-w said:

 

That's weird... no trouble with the CPU's iv'e delidded. And not seen anyone report that before. It is a bit tricky to get it spreading, but it flows after a bit of manipulating. 

I had a few black Q-tips that came with the liquid metal, and while it took a while to finally get it to stick and spread across the cpu die, it did work.

Not so, for the ihs. I rubbed if back and forth, swirled it, went side to side, smushed it...... nothing. I was just pushing a little quicksilver bead around, and probably annoying it.