DIY Intel CPU Relidding with no Delid Tools
ฝัง
- เผยแพร่เมื่อ 29 ต.ค. 2024
- We show you step by step, how to go about doing a great job a RElidding your Intel CPU without using any special die tools or jigs. We also show you CoolLabs Liquid Ultra application close up. Then we take it all apart to see how good of a job we did...or didn't. YMMV.
HardOCP - hardocp.com
HardOCP Patreon - www.patreon.co...
HardOCP Twitter - / hardocp
HardOCP Facebook - / hardocp
Thank you for all the good work. We appreciate the time you put in to show everybody all the details.
Kyle, you just convinced me to redo my delid on my 6700k...this time clamp it down with foam support. I pinched her back together after it tacked up, and popped it back in my socket and clamped it down. When returning home 2 hours later, I put the water block back on and started using PC as normal. My temps went from 63 at full load to 50. I wonder if I can get it cooler with clamping it back on. Thanks for the vids man!
I would use RTV silicone that says "ultra" in it.
(Ultra Black, blue, red, etc.)
That means it is oxygen sensor safe and fume free.
This means that when it gets hot,
no fumes are emitted that can harm electronic devices.
Regular RTV silicone will emit fumes, more when hot.
there is no ultra red..
10:21 - he spills some, good job mate :D
kyle, the heatspreader is maybe a little bit far "above" the die (little gap)
it is possible to scrap down some 0.02 -0-04mm off the heat spreader: at the perimeter. so it will be lower.
maybe also scrap off some 0.02 off the thickness
Good video. 14:50 Why not just put a weight on top rather than use a clamp? Pressure would be uniformly spread. 16:40 From everything I have learned, it is better not to reseal the IHS to the PCB. Just tiny outside dabs of gunk at the corners if you want to secure the IHS, otherwise don't seal at all. Some people use tiny dabs of Crazy Glue. Apparently it is desirable to reduce the distance between the top of the die and the IHS, with lower temperatures being reported as a result, although some people suggest the IHS may be riding entirely on the die then. Also, maybe a good idea to paint the top of the PCB and tiny components with suitable (non-toluene) clear nail polish or a professional compound such as MG Chemicals 4223F polyurethane protective conformal coating), to prevent any shorting by the liquid metal if it flows out (the other MG compounds 419D, 422B and 4225 are no doubt also OK).
awesome vid man! just in time for me. thx!
I was inspire by this video to give my own delidding a try and I can attest to being able to do this without buying a single thing. Just takes patience and nerve.
Fortunately, I have a 3D printer and was able to design, and successfully use my own, custom made delidding tool. Worked like a charm, though it destroyed itself after the single use.
I went with copper high temp sealant instead of the red, but that was simply because that's what I had on hand; red would work just fine and copper's only difference is a higher temp rating which is way over anything a CPU should ever see and color.
I also 3D printed my own, custom, relidding assembly and used some medium pressure clamps I also had on hand. Cured for 20 hours, there abouts.
I used a syringe and needle tip I had on hand from an ink refill kit and clamped the end to get a finer bead almost exactly like HardOCP did.
Didn't do any fancy benchmarks, but can already tell temps are way down. I used the CPU-Z's built in stress test (before delidding) to see temps hit 70C+ temps, now (after delidding) same stress test yields test no higher than 61C.
I had concluded that I didn't get the best chip of the lot and experienced the same temp spikes as other users experienced. I also realized this CPU will probably never see stable 5Ghz, but that's not a problem for me as I don't usually OC; was just tired of seening wild temp spikes simply by opening a browser and such.
To me, overall, it has been a big difference and well worth the effort, though my hands are still shaking from the attempt. :) Thank you HardOCP for your many experiments.
Note: I would recommend is either working on an anti-static pad or wearing an anti-static wrist or leg strap. When relidding I also used a cushioned material similar to HardOCP's technique, but also had that wrapped in anti-static material. CPUs and RAM are highly sensitive to ESD and precautions should be taken to not damage the components. This includes making sure to neutralize any tools being used.
I ve changed 7700k my cooler to EVGA clc 240 because those temp spikes. Today I learn this whole thing . Water cooling didnt help much , almost whenever I open mozilla , cooler fans whines cause temps reach 60-70 ish degree Celcius. I even thought selling cpu and board and move Ryzen 2. Because those kits + sealents + liqiud metal thing + shipping will cost 60-70 euro.
Can someone tell me the practicality of removing the lid of a CPU? I assume that they would be extremely difficult, to impossible to modify, or maintain/repair.
We generally apply a much more conductive Thermal Interface Material, which brings temps way down under load and provides us with better overclocking results. You can see some results here: www.hardocp.com/article/2017/10/11/overclocking_intel_8600k_delidded_on_air_water/
@@HardOCPTV
Oh, gotcha'. Thanks for the quick explanation!
I'm SOOoooo early to this video I am having trouble not typing FIRST. On a more [H]ard note is there any benefit to resealing the lid? i delided a 6600k and used Liquid Ultra and just placed my lid on top when reseating.
Obviously I didn't finish watching the video before commenting. However i did a 'not normal thing' when I re- lided mine. I painted the entire non-die area with a ceramic TIM just encase I used too much Liquid Ultra. In hide sight my guess work and fumbling probably kept everything kosher.
Addressed in the video and I am putting the review article together now. Writing it as we type.
Thanks for the sealant tips/tutorial boss. When I remount mine in the future I will definitely be referencing your article/video. Thanks for the video I am definitely the kind of person that needs to see a thing done in order to copy it. Great step by step.
Side note, when did your beard go ALL grey!?!?
Hehe, beard went gray a long time ago! I will update the video with the link to the article in a bit. It is short and to the point.
You don't look like this anymore? www.kylebennett.com/files/hfpics/photokb/96_kyle-the-cat.jpg
Hi. Why did you leave a gap of the sealant? was it to reopen the lid??
Allows for expansion of gasses under the IHS through heating and cooling cycles.
@@HardOCPTV oh thanks I didn't know that
ive delided about 40 cpus in the last few years... started with haswells. Haswells used to be great. clamp them into a vice, take a block of wood and smack em with a hammer... these days Skylake and Kabys PCB so damn thin had start using razors... still only takes about a min but the first 6700k i smack with a hammer i cracked the pcb in half lol. Just keep getting thinner and thinner =D
You would be correct sir. I took some time to get out the caliper and go over some old favorites. hardforum.com/threads/naked-intel-core-i5-7600k-kaby-lake-how-to-h.1924043/page-2#post-1042803724 Since Haswell, the substrate has gotten tremendously thinner.
Now you gotta hit em with a heat gun
My Cpu will not post - code 00 - This is the second 6700k to break after a delid. I will probably just replace with ryzen 3000 and not delid anymore.
Why do you leave the gap when resealing the HIS? Wouldn't the air get pushed out as long as it's clamped strong enough?
That leaves a gap for pressure changes as it heats and cools.
I've seen videos of people using super glue to relid the ihs. Is super glue or gasket sealer better for relidding? Thanks for your video btw.
I am not a fan of using "super glue." It sets up fast and if you do not use enough clamping force, you can negatively impact the Z height of the IHS.
Makes sense. Thanks again!
Super glue is really bad.
Man I ain't doin any of this shit. I'm replacing my 2600k with Ryzen 1700. I was so excited of ugprading my rig to 7700k but then all these delid trends came up because Intel used a shitty TIM rendering high temps on AIOs even without overclocking.
wimpy
Itbankrock Ryzen for gaming is not upgrade from 2600k and 2600k is soldered so it's same as having decided cpu
Enjoy the bottleneck LMAO
2600k is still strong
Nice Video - The Sharpie is a "Ultra-Fine" not Fine point - If that was a Fine point the whole cpu would be covered.
I like this
Need 4speed Cpu.
NT-H1 is a bad idea, don't do it unless you wanna take that thing apart every month at least...
is that a working 4770k?
No.
Wayyyyy to much sealant!!!
Thin see through film is almost too much
I just use 4 drops in the corner :p
*Shot* Thank you!
That made me laugh ohandyeahnicevidman
lol, just put cpu in socket and mount cooler. That is ideal pressure.
But what we have found is that without sealing it increases the Z height in the socket due to more substrate flex.
but what I said is with sealing :). So put all as u with glue and put in socket and put cooler and wait till tomorrow. That's it.
Totally agreed! Just do not put a cooler on it while the glue is setting up in the socket.
why not, stock cooler. That is minimal but enough pressure.
Because when you put a cooler on there, you pushing down on the IHS, and the IHS is pushing down hard on that substrate, and the substrate flexes into a concave shape making the Z height greater, which is what you do not want. Setting up the seal with the socket is NOT ideal, but does work. To get the best Z height you would want the substrate fully flat while the seal is setting up. This is why we do not suggest you using the socket. That all said, it is not going to be a night and day temperature difference, but every little bit is what we were shooting for.
I don't care what anyone says. Have you ever watched a video of someone
deliding their processor using one of those deliding tools? These
machined tools send such a tremendous shock through the entire processor
at the moment the lid breaks free. This cannot be good for the millions
of tiny circuits packed inside the chip. No way is that a good idea...
You might as well throw it against a wall, it's about the same shock
level! ... The razor method is the only way to go. It's gentle and non
destructive as long as you make sure to keep the blade absolutely
parallel with the circuit board of the processor, and not let the blade
go in too much as to protect the raised chip in the center from getting
scratched. Otherwise, the razor method is far superior to the shocking
delding tools available today.
There isn't shock like you make it to be. Certainly not even remotely close to throwing the processor.
ive been running a delided 4790k for several years now, 24/7 on an aggressive overclock.
absolutely no issues. I dont know what you're on about.
Same here man, he is probably just salty as he tells himself that he will break his cpu if he delidds it. So he talks down on the fastest and safest method to do it. Do a quick search and look up bricked cpu's using vice and razor blade method then check bricked cpu's using delidd tool. So tell me which one is less risky lol.
CPU's and GPU's have been proven to be very durable. Delidding it wouldn't break it unless you smashed it with a hammer. If you are careless with it, yes you can chip the silicon, then it's as useful as a paperweight, but if you are careful, it can easily be done.