Modding the IHS?

[Pierre3400]

Hey guys,

I have just ordered myself a few liquid ultra's. I will be delidding my 3770K.

I have been reading more and more success stories about delidding these cpus.

I have come to one conclusion. It works! But i have noticed that people doing this, spend a lot of time cleaning the CPU's, the glue that has been used to hold the IHS in place, and this is not replaced.

I noticed, that the inside of the IHS of course houses more space then the cpu chip needs, in my mind this is pure air.

If there was a possibly of air to get out from inside the IHS, would this not improve the cooling improvement further?

Here is my idea explained, with my mad paint skillz.

Lid untouched.

This is where i would use a very tiny drill (1mm/2mm), and make 4 holes. From my point of view, this would help heat get away from the chip. It my not be big, there may not be an airflow, but in my mind it would let heat get away.

I plan on trying this on my 3770, the lid will also be lapped. Since the Liquid ultra isnt cheap, and i have no idea if i can take the chip apart after using it and testing, i am unsure if i can run a delidded and lapped test, and then a delidded, lapped and modded test after.

What are thoughts on this?

 

[briansun1]

I think you mikght want to take a look at this thread http://www.overclockers.com/forums/showthread.php?t=726373

 

[ozan]

Dafuq? You cant de-lid 2500k. De-lidding is pretty much only for ivy bridges this generation. If you are going to de-lid the ivy, I recommend using a shaving razor's blade, since thats how me, the coplete noob did it on the first time, without screwing it up.

 

[Woomack]

In this case better take off IHS and try to put cooling directly on the core. Some users had worse results without IHS but that's probably only because of not best cooling placement as core is not exactly in the center.

About SB, try lapping it so much till you see the core I saw that somewhere but it's easy to damage cpu.

 

[Pierre3400]

Im sorry i do not understand what you are trying to explain by posting that link.

 

[briansun1]

I don't ether ? what was I thinking???

 

[Pierre3400]

Dafuq? You cant de-lid 2500k. De-lidding is pretty much only for ivy bridges this generation. If you are going to de-lid the ivy, I recommend using a shaving razor's blade, since thats how me, the coplete noob did it on the first time, without screwing it up.

Dafuq! Im sorry, didnt know about sandy bridge. But will do it with my ivy then or buy a cheap ivy, and why on earth you tellinh me i have to use a razor, thats like asking what way to the south pole and u point north, 100% besides the point, this is about drilling holes in the IHS while its delidded, if i can get that far then do you really think i need advice on how? Theres 5 million how tos out there, and im pretty sure i wrote "i have been reading"

Please stick to topic. But thank you for telling that sandy dont delid. But it dont change the common goal on the idea. And its not a 2500k its a 2500.

Im sorry for outburst, but im getting tired of people ignoring the point of a thread!

 

[EarthDog]

I dont think you can delid a 2500... its soldered.

And why are you drilling holes in the IHS? I dont recall that part of the procedure. You use a razor blade to get it off... unless I missed something?

 

[Pierre3400]

Earthdog, it has zero, nothing, nada to do with removing the IHS, its to give the heat that will closed inside the IHS a chance to get out.

 

[ozan]

Earthdog, it has zero, nothing, nada to do with removing the IHS, its to give the heat that will closed inside the IHS a chance to get out.

After de-lidding, you will probably remove the black goo, which will provide plenty of airflow below the ihs, if that is really needed at all

 

[EarthDog]

Earthdog, it has zero, nothing, nada to do with removing the IHS, its to give the heat that will closed inside the IHS a chance to get out.

What?

If that was a good idea, I would have to imagine it would be done by the CPU MFG. I think you are barking up the wrong tree here for absolutely no reason. Delid the right way and dont make your IHS look like a strainer. LOL!

 

[Pierre3400]

Im just going with basic physics here.

Heat travels upwards.

If you put a hole in the top of the box, hot air will leave the box.

But let it go, seems no one sees a point, so imma do it, and report back when its done.

What?

If that was a good idea, I would have to imagine it would be done by the CPU MFG. I think you are barking up the wrong tree here for absolutely no reason. Delid the right way and dont make your IHS look like a strainer. LOL!

Im sorry ED, but that just seems closed minded.

Im very sure that just cos no one has done it before, that doesnt mean it doesnt work. Im very sure that history has proved this over time, even though we are only talking about an IHS.

 

[EarthDog]

I dont see a point.. the changes will be negligible at best if not making it worse and you end up with a strainer for an IHS. I also have to think that an IHS IS a heatspreader and would be more efficient at taking away heat that mere air making contact with the TIM. But that is just a guess. YOu are taking away area for the IHS to do its job by drilling holes.

Looking forward to seeing the results though, and laughing at the look of your IHS (not laughing AT you, just the look of the IHS).

Report back!

Im sorry ED, but that just seems closed minded.

Peirre, if I thought it had a snowballs chance in hell at working, I would be behind it. As it stands, Physics doesnt prove anything... that heat that you say is 'trapped' already rises to the IHS so its already whisking it away. All this does it take away IHS for the air to touch TIM. I cant imagine any difference out of this. But I hope, for your time and effort and killing the warranty, something positive comes out of it instead of spaghetti strainer fit for a mouse.

Make sure you:
1. Note ambient temps for your baseline and current testing.
2. Make multiple mounts and TIM applications so you can minimize that potentially huge difference.

 

[Luis_GT]

Making holes on the IHS will not make any difference... the fact that the IHS already has a small gap when its siting on top of the die... ie. Non of the edges are sealed since the die is thicker than the IHS

 

[King107s]

I'm betting this attempt will destroy the CPU... Lapping it for better contact and thermal conductivity is the only thing you should be trying here. On that note, Intel has gotten A LOT better at manufacturing their IHS's and the days of Prescott cores with crazy concave IHS's are pretty much gone. I did not lap my i7-2600K and my temps are uniform across all cores and very low (on water).

If it aint broke don’t fix it!

BTW I'm pretty sure the brilliant engineers at Intel have put quite a bit of thought into the design of their processors and the IHS that covers them...

 

[Pierre3400]

Odds are i will end up selling my 2500 and buy an Ivy bridge, mainly for testing this.

I still dont, and possible wont ever understand your view on this.

Are you on mobile where you cant see pictures? The 4 holes wont be on the top surface. That will only be lapped.

 

[EarthDog]

@ Luis GT - the die is sealed isnt it? Its sitting on a pile of glue or whatever? Unless that medium isnt airtight, its sealed. Right? s2011 CPU's has a hole in it, but not for cooling purposes.

@ Pierre, I know where you want to drill them, now, and my thoughts do not change one bit. This wont change a thing.

 

[Pierre3400]

@ Luis GT - the die is sealed isnt it? Its sitting on a pile of glue or whatever? Unless that medium isnt airtight, its sealed. Right? s2011 CPU's has a hole in it, but not for cooling purposes.

It honestly sounds like you think im going to drill holes in the top lapped surface of the IHS.

 

[EarthDog]

Read my edit.. I NOW know and it doesnt change my conclusion in the least. A+ for the idea, I just think it wont yield a darn thing. As I said, the warm air trapped inside is already in contact with the IHS.

 

[Luis_GT]

@ Luis GT - the die is sealed isnt it? Its sitting on a pile of glue or whatever? Unless that medium isnt airtight, its sealed. Right? s2011 CPU's has a hole in it, but not for cooling purposes.

The bottom part, where the IHS is cut in has no glue at all leaving a small gap.

If you delid and remove the glue, when you place the IHS on the die it'll free spin on top of the die as non of the edges of the IHS make contact with the CPU board.