3770K delid: reaffix lid or not?

MiG29TangentBoy

Limp Gawd
Joined
Nov 2, 2004
Messages
377
I'm going to delid my 3770k, and I have been researching the process. My only question is if I should bother re-securing the lid afterwards? I see some people use silicone sealant to keep it back on, where others don't bother and just say to be careful when locking it back into the socket. Would that infinitesimal amount of space the glue was taking up under the lid's edges make that much of a difference in regards to the space between the IHS and the core itself? I'll be using Coollaboratory Liquid Pro on the core itself, and have no real intention of reopening the lid back up once it's done. I was thinking of using a few small dots of epoxy just to keep the lid from wiggling around after, but I wasn't sure if I should make an effort to properly seal it up more than that. I know this is probably a silly nitpick, but I'd rather ask first since this is my first delid effort. Thanks!
 
This is a good question.

I will never even attempt this as I have been present a bare tbird core crushings. You never forget that creaky silicon crush sound as you man handle those old HSF units on with a crowbar.
 
with coollaboratory stuff between the die and internal heat spreader if you attempt to open it again you will most likely find that you have a broken/scratched CPU.. so once the heat spreader its set just keep it with a little pressure the coollaboratory liquid metal will solidify and keep the heat spreader firmly attached to the die and PCB so no worries trying to add glue, epoxy or anything else.. what I generally recommend its just put a cup above the heat spreader to keep the pressure for 45minutes - 60 minutes and done you can set the CPU to the motherboard..
 
So the TIM will be enough by itself once set? That's good to know. Thanks!

schizrade: I feel that. I remember having to invest in those 462 copper shim covers to protect my old Barton chip under a very heavy Thermalright SP-97 heatsink. Those were the days. But since I'm not going bare core on this one, I'm not so worried (I ain't that brave, probably as a direct result of those Socket A cracked core memories!).
 
Last edited:
This is a good question.

I will never even attempt this as I have been present a bare tbird core crushings. You never forget that creaky silicon crush sound as you man handle those old HSF units on with a crowbar.

You're not the only one, I crushed a Athlon 1.4Ghz tbird chip with a Thermaltake Volcano 7, which was an enormous heat sink with the "conventional" single-prong bent steel clip. I still have nightmares! Well, not really but I'm not de-lidding anything any time soon.
 
Do they even make shims that fit newer CPUS? I remember they were pretty popular back in the Athlon Socket A days. I'd feel more comfortable having a wider flat surface for a heat sink to contact than a smaller core that can get cracked when tightening it down.
 
Relid it for sure.. Don't shim or use tape etc. I had to bend the brackets on my h100i to apply enough pressure to the CPU to make up for the gap created by the delid process. My PC didn't boot at first, and I thought I had killed it. Until I pushed lightly on the block and it fired right up, bent bracket real well and it was golden ever since, guess the pressure mattered even though without the bending it was still pretty secure. The biggest gain from delidding is getting rid of the gap, don't forget that.
 
The purpose of delidding is twofold not only better tim but also it IS the silicone layer which makes temps bad so if you were to use silicone again use a very SUPER thin layer and lay it ONLY everywhere but the mounting wings side [only on top/bottom side not sides then mount it into socket squeezing it to set few hours at least before installing heatsink OR an hour or so hot set THEN few hours cooling airing it out]

I used a little drop of super glue in the center of each of 'the wings' between the underside of ihs and pcb **placing it just seconds before laying the ihs lid back onto the cpu itself close to original place as possible
-- it was placed carefully into the socket after applying coollab stuff so I could lay ihs on and clamp it down immediately while holdin ihs with my finger in place best I can or glue might dry THEN I cleaned the top again while mounted before placing final tim for heatsink
Yes the finer details of these steps are important lol .. I know acetone can damage pcb layer hope superglue isn't anything remotely near it. .
Really dunno if it is holding but will find out lil later in the week I used the gel type didnt want it flowing freely but guess you could use a paper clip dabbed in some if you only have regular superglueskii
 
Last edited:
This is a good question.

I will never even attempt this as I have been present a bare tbird core crushings. You never forget that creaky silicon crush sound as you man handle those old HSF units on with a crowbar.

And that is why you use a nut driver instead of a flat head screwdriver.

Not only do you not have to worry about slipping and gouging the motherboard, you can also be a lot more precise with the pressure you are putting on the clip.

That being said, I broke a socket-A chip into 3 pieces when I was tightening down a home made waterblock.
 
Yes, it's recommended to place the IHS back onto the die once you've applied CLU/CLP. The difference in temperature between bare-die and IHS are negligible.
 
I went with the ek supremacy evo so I could use their naked ivy kit (works w/ has well as well), no need to worry about crushed core and no need for heat to transfer through the ihs.
 
Perhaps the primary purpose of the heat spreader is to maintain a gap between cpu die and the heatsink. When you lock down the heatsink the heat spreader keeps it from crushing the die. Some heatsinks that have curved base reportedly perform better than flat-base heatsinks. I can imagine the curved base is able to press the heat spreader more and bridge the gap between cpu die and heatsink even more and thus the better performance.
 
Back
Top