Rockit Cool Rockit 88 Delid and Relid Tool @ [H]

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,424
Maybe this has been answered, but I can't seem to find it :( What TIM did you use? Finally setup my 8700K and thinking of doing this now :p

Typically we use liquid metal like CLU (sub for your pref) between the die and the IHS. You could also substitute a high quality paste. And to top it off we use a quality paste between the IHS and waterblock/cooler. You could use CLU here too of course but it will dry out much sooner making eventual cleanup a pain in the ass.
 

Digital Viper-X-

[H]F Junkie
Joined
Dec 9, 2000
Messages
14,115
I've been using MX-4 for the longest time, is it time to switch ;P?

So either CLU or Conductonaut work?
 

DanNeely

2[H]4U
Joined
Aug 26, 2005
Messages
3,729
Are you getting higher stable clocks after delidding, or just significantly lower temps?
 
Joined
Nov 13, 2010
Messages
759
For anyone interested I delidded for the first time yesterday. Here's some personal tips.

I would always opt for the nail polish. Gives you that peace of mind for 5ish bucks. Sure you might have the hands of a surgeon, but there's always a chance of drip or the liquid metal spreading out to contact points during relidding.

Be easy on the gasket. Gasket makers are very messy to deal with. Just pouring it on via the plastic icing nozzle gives way too much. apply a litle and just spreadint it on the edges of the IHS is enough.

The little toothpick that comes with the delid kit is your best friend.

DO NOT forget to apply liquid metal to the underside of the IHS.
 

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
51,126
DO NOT forget to apply liquid metal to the underside of the IHS.
Not that this would cause any harm, I am not sure where this got to be a thing. You know how many footprints I have checked using single side application of CLU? A lot. Never seen any issue with it. But again, sure, go ahead, not harm in that. Check the pics for my foot print....and I laid it down MUCH thinner than I usually do.
 
Joined
Nov 13, 2010
Messages
759
My initial thought that one layer on the cores was enough. A ton of follow along videos did not show them putting anything on the IHS.

I shot a message to James at Rockit and one of his recommendations was to put a layer on the IHS. That was my first time hearing it. Then after doing the whole delid process I watched a few more videos and one from GamersNexus advised to put a layer on the IHS. Well lo and behold I put in my delidded cpu and the temps were spiking higher than they every did stock. I redid the whole job and put a good layer on the IHS. I am now getting really good temps especially during 100 percent load.

This is all anecdotal though so I'm sure people get great results without spreading some on the IHS but personally I'm going to do that from now on.
 
Joined
Apr 5, 2016
Messages
2,249
My initial thought that one layer on the cores was enough. A ton of follow along videos did not show them putting anything on the IHS.

I shot a message to James at Rockit and one of his recommendations was to put a layer on the IHS. That was my first time hearing it. Then after doing the whole delid process I watched a few more videos and one from GamersNexus advised to put a layer on the IHS. Well lo and behold I put in my delidded cpu and the temps were spiking higher than they every did stock. I redid the whole job and put a good layer on the IHS. I am now getting really good temps especially during 100 percent load.

This is all anecdotal though so I'm sure people get great results without spreading some on the IHS but personally I'm going to do that from now on.
I think it depends on what you use to re-adhere the IHS, how much clamping force you use, and whether you're using a liquid metal TIM or a more traditional thermal grease. Liquid metal TIMs spread super thin compared to others. On my latest relid job, I used CoolLabs Liquid Ultra, and I didn't put any TIM on the IHS, and my temps are fantastic. However, I also used superglue to re-adhere, and a fairly aggressive amount of clamping force, which means my IHS's outer flat is practically sitting on the PCB - the distance between the die and the "ceiling" of the IHS is as small as it can be. If I'd used RTV, and/or less clamping force on the relid, that single layer of liquid metal TIM may not have been enough to make good contact.

Isn't it fun tearing expensive, delicate pieces of electronics apart? XD
 
Joined
Nov 13, 2010
Messages
759
Yeah, I did use RTV so that would account for mm of space that I had to compensate for.

How is using super glue? I've seen someone superglue just the corners but I was scared that the liquid metal would just drip through the cracks.

I'm actually quite done with RTV if there is a better alternative.
 
Joined
Apr 5, 2016
Messages
2,249
Yeah, I did use RTV so that would account for mm of space that I had to compensate for.

How is using super glue? I've seen someone superglue just the corners but I was scared that the liquid metal would just drip through the cracks.

I'm actually quite done with RTV if there is a better alternative.
I've used super glue on every delid. Twice on a 7600k, and three times on a 7700k. I used the corner method your describe. Never had a problem with it.

The liquid metal TIM is "watery" compared to traditional thermal pastes, but once squished between the IHS and die, it won't "run out" or slosh around inside the cavity. Every time I've taken a previous relid apart, I've found the liquid metal TIM right where it belongs: on the die, and on a die-sized rectangle on the IHS.

Super glue is also advantageous in that popping the top in the future is much easier than it would be with RTV. It can be easily cleaned off with acetone and a q-tip; no need for the scrapey scrapey.

It's worth mentioning that I've only ever used the LocTite Gel Control superglue sold on RockitCool's site. I've read horror stories of other folks using superglue and it melting the PCB. I can only assume that not all 'superglue' is made equal, but I can attest to the safety of that particular type.

Be aware: a superglue relid with a stock IHS will result in an overall package that is a fraction of a millimeter thinner than a normal Intel CPU. If you're using a regular CPU waterblock that employs a spring-tension hold down method, this won't be a problem. If you're using a monoblock, however, which has no springs and relies on the CPU being a certain thickness, this may result in a condition where the IHS isn't sitting "tall" enough to make good contact with the coldplate. You can get around this with a custom-milled IHS, and RockitCool will make you one at no extra charge on request, but it's something to know about if you use superglue.
 
Joined
Nov 13, 2010
Messages
759
Re dellidded and used the super glue method

When I opened it up the rtv had spread itself almost to the cores. I did a thorough cleaning job this time and I wasn't as fearful about the contact points. I reapplied the liquid metal and I still did a thin layer on the underside, a preference at this point. This time I used NT-H! instead of Arctic 5.

Glad to say I got ANOTHER 7-10c drop in idle temps and I'm sure max temps will improve as well as soon as I test.

delid2.JPG
 

trick0502

Supreme [H]ardness
Joined
Apr 17, 2006
Messages
5,366
Think it would be worth it to invest in one of these for a 4770? Total cost would be about $80 for all the parts to de and relid.
 
Joined
Apr 5, 2016
Messages
2,249
Think it would be worth it to invest in one of these for a 4770? Total cost would be about $80 for all the parts to de and relid.
$80? Seems like a steep estimate to me, but it's been a while.

How hot does your chip run? Are you overclocked? Is your goal higher clocks? Do you have a different goal? It's tough to gauge worth for other people, moreso when one isn't sure of their purpose in an investment. =)
 

trick0502

Supreme [H]ardness
Joined
Apr 17, 2006
Messages
5,366
$80? Seems like a steep estimate to me, but it's been a while.

How hot does your chip run? Are you overclocked? Is your goal higher clocks? Do you have a different goal? It's tough to gauge worth for other people, moreso when one isn't sure of their purpose in an investment. =)

@4.2 it runs hot. I’m really just trying to get a little more out of this chip. The cost of upgrading to a newer platform isn’t really worth the cost/performance right now.

I’m thinking if I can get 4.5-4.6 (300-400mhz) boost and run at cooler temps it might be worth it.
 
Joined
Apr 5, 2016
Messages
2,249
@4.2 it runs hot. I’m really just trying to get a little more out of this chip. The cost of upgrading to a newer platform isn’t really worth the cost/performance right now.

I’m thinking if I can get 4.5-4.6 (300-400mhz) boost and run at cooler temps it might be worth it.
Do you believe temperatures are inhibiting your clock ceiling? If so, a delid and relid with a quality liquid metal TIM will make a significant difference. Trouble is, it's tough to tell if you're hitting a thermal ceiling, or if you're just at the limits of your chip's binning until you see how it performs with the new cooling. The silicon lottery is in full effect.

The nice thing is that even if you don't see massive gains in this chip, when you upgrade down the line, that delid tool will still be in your toolbox.
 

Aenra

Limp Gawd
Joined
Apr 15, 2017
Messages
191
Even if voltage wasn't an issue? You will never get 400MHz more because you delidded. If that's your goal, save yourself the time, risk and warranty-voiding.

* Because i do not know whom i am addressing, might as well.. and excuse me if this is old news to you: :)
Practically speaking, what you might get is maybe another 100MHz, or a reduction in temps that you'd otherwise have been unable to. For a very, but really very rough estimate:
1) divide the maximum wattage of your CPU at its highest current OC by the thermal conductivity of your current paste.
2) divide the maximum wattage of your CPU at its highest current OC by the thermal conductivity of your LM of your choice.
3) subtract (2) from (1).

So for example sitting at 400Watts with a paste of 15W/mk, you'd get 400/15=26,6C
Sitting at said same 400Watts with say conductonaut (73W/mk), you'd get 400/73=5.4C
Ergo if you changed, you'd gain approximately 21C. Tops.
Assuming that is your remaining cooling setup has the capacity to absorb all of this differential. 99% it won't.. you're talking 4 and 5 rads minimum to see this much. And even then, we'd have to take your waterblock into consideration, ie its own thermal limits; and your coolant; and flow rate.. and so on.
You want to see the real numbers, only one way to go about it :)

But again, don't even think of something like +400MHz.

** Was busy before, neglected to think of the obvious, lol; so in anticipation of your enquiry: Yes, most people do gain around a 10C benefit as you may have heard/read. And even that is under good circumstances, can be less. That's because as stated, optimal circumstances are not feasible for most and most importantly, the above "equation" is as rough as it gets, there are other factors involved in this as well.

TLDR, if you haven't wasted a fortune for your WC gear, my advice is don't. One fuckup and you'll fry everything; unlike pastes, liquid metal is electrically conductive.
 
Last edited:

DanNeely

2[H]4U
Joined
Aug 26, 2005
Messages
3,729
Even if voltage wasn't an issue? You will never get 400MHz more because you delidded. If that's your goal, save yourself the time, risk and warranty-voiding.

* Because i do not know whom i am addressing, might as well.. and excuse me if this is old news to you: :)
Practically speaking, what you might get is maybe another 100MHz, or a reduction in temps that you'd otherwise have been unable to. For a very, but really very rough estimate:
1) divide the maximum wattage of your CPU at its highest current OC by the thermal conductivity of your current paste.
2) divide the maximum wattage of your CPU at its highest current OC by the thermal conductivity of your LM of your choice.
3) subtract (2) from (1).

So for example sitting at 400Watts with a paste of 15W/mk, you'd get 400/15=26,6C
Sitting at said same 400Watts with say conductonaut (73W/mk), you'd get 400/73=5.4C
Ergo if you changed, you'd gain approximately 21C. Tops.
Assuming that is your remaining cooling setup has the capacity to absorb all of this differential. 99% it won't.. you're talking 4 and 5 rads minimum to see this much. And even then, we'd have to take your waterblock into consideration, ie its own thermal limits; and your coolant; and flow rate.. and so on.
You want to see the real numbers, only one way to go about it :)

But again, don't even think of something like +400MHz.

** Was busy before, neglected to think of the obvious, lol; so in anticipation of your enquiry: Yes, most people do gain around a 10C benefit as you may have heard/read. And even that is under good circumstances, can be less. That's because as stated, optimal circumstances are not feasible for most and most importantly, the above "equation" is as rough as it gets, there are other factors involved in this as well.

TLDR, if you haven't wasted a fortune for your WC gear, my advice is don't. One fuckup and you'll fry everything; unlike pastes, liquid metal is electrically conductive.

Thanks for this assessment, it answers exactly what I was asking about earlier.

FWIW I don't have any plans to mess with my current 4790k or to do a new build for another 2-4 years so I was mostly interested in general information. While I don't have multiple rads I do have a quad 140 running my CPU+GPU. It's a bit overkill for my current setup, when I bought it I was assuming SLI/xFire was in my future to power the 5k monitor I was lusting after at the time. Not going to happen now with multi-GPU on life support and moving towards hospice care.
 
  • Like
Reactions: Aenra
like this

cdr_74_premium

[H]ard|Gawd
Joined
Oct 20, 2010
Messages
1,578
I am not asking if it's worth it, because it's probably not.

Say I wanted to delid my puny i3-4170 just for lower temps and a longer life span. Is there a manual about this somewhere? I can google it, but I want something [H]'ers would approve if possible.
 

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
51,126
I am not asking if it's worth it, because it's probably not.

Say I wanted to delid my puny i3-4170 just for lower temps and a longer life span. Is there a manual about this somewhere? I can google it, but I want something [H]'ers would approve if possible.
Would certainly give you lower temps if that is what you want. Longer life? Who knows. Thing will probably last a lot longer than you want to keep it anyway.
 

Aenra

Limp Gawd
Joined
Apr 15, 2017
Messages
191
Is there a manual about this somewhere?

Of course not.. what do you want 'x' manufacturer to tell you? That LM is corrosive, electrically conductive and prone to solidifying over time? They do need to sell :)
There is nothing 'safe' about LM. It's not mission impossible, it's not even all that difficult to handle/apply; but it's not safe. And at least in my experience? It will 101% not prolong your CPU's life span. Best case scenario, you'll be where you're at now (in terms of life expectancy).

But you'll say, what if i was nearing 90s and am now in the 80s? To which the only reply? Honestly? You should never have been in the 90s, or 80s even, in the first place.

Anyway, it's a hobby! Do it, but do it knowingly.
For me? A current example, my 1950X is sitting at 4.1gigs and with voltage to spare, got lucky; my temps are around 65-68, ambient depending, my WC-related gear is tip top; so i'll be applying LM to see if i can hit a stable 4.2. I'd never do it to lower my temps; barring extremes, temps are manageable in a million ways. Including lowering one's OC ^^

* (last) Edit: One day i'm going to make a long post about temps and airflow and aaaall the bad habits i see going on. One day.. because the list is ever growing, lol.
 
Last edited:

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
51,126
Really? The superglue is MORE work to relid?
Well, this is exactly what I have always said, but I decided to experience it first hand. Superglue was 100X more work to relid than RTV. RTV comes right off. Superglue, not so much. And you have to take it off or you are going to impact your Z height.
 
Joined
Apr 5, 2016
Messages
2,249
Well, if Acetone is used to melt the adhesive, I'd be a bit worried about other components that might have some sort of glue / adhesive on them
*shrug* I wouldn't. I've all but replaced Isopropyl with Acetone for cleaning everything. Both the aforementioned processers have practically taken baths in the stuff, more than once each, and come out no worse for wear.

YMMV, of course, but it's definitely caused me no problems.
 

Digital Viper-X-

[H]F Junkie
Joined
Dec 9, 2000
Messages
14,115
I just hit up home depot : P

Liquid Electrical Tape : Check
Red RTV: Check
Gel Control SG: Check

I got both just in case, I'll decide when I do the delid/relid which to use. I'm thinking of doing a practice run on my i5-4590 before I touch the 8700k. Is there enough liquid metal for 2+ applications in the 1 GM Conductonaught?
 

Digital Viper-X-

[H]F Junkie
Joined
Dec 9, 2000
Messages
14,115
Success! though the IHS seemed a bit uneven, when re-installed. there is a nearly 20deg c drop on prime, and a considerable drop on mining temps.

i did go with superglue for this attempt, Had acetone ready just incase, used liquid electrical tape as well. Lets see if I can now get it to 5ghz on not so insane temps!

Before: Prime
before.png

Before: Mining
before_mining.png



AFTER: Prime
prime_after.png


AFTER: Mining

after_mining.png
 
Joined
Apr 5, 2016
Messages
2,249
Success! though the IHS seemed a bit uneven, when re-installed. there is a nearly 20deg c drop on prime, and a considerable drop on mining temps.

i did go with superglue for this attempt, Had acetone ready just incase, used liquid electrical tape as well. Lets see if I can now get it to 5ghz on not so insane temps!

Before: PrimeView attachment 69144
Before: Mining
View attachment 69145


AFTER: Prime
View attachment 69146

AFTER: Mining

View attachment 69143
Excellent work, and congrats on your better temps! =)
 

Digital Viper-X-

[H]F Junkie
Joined
Dec 9, 2000
Messages
14,115
Excellent work, and congrats on your better temps! =)
‘‘Twas a bit nerve racking but the [H] video made it a bit easier.

Only one side of my IHS is flush against the cpu substrate, which has me worried since I ran out of mx-4 and couldn’t redo everything today. But the better temps made me feel better :D
 

rpjkw12

n00b
Joined
Dec 31, 2016
Messages
2
I need to find a source for the syringes... I've just been using a toothpick to apply the RTV to the edges of the IHS (yes, I still make sure to leave an air gap). Works just fine but the syringe / needle method would be a lot less messy methinks...

I already bought a 100 pack of those plastic razor blades Kyle uses in his delid videos. Those things make cleaning the glue off the PCB and IHS a freaking breeze!
I'm, as usual, late to the party, but have you considered diabetic insulin syringes and needles? They could be too small and I don't know if a prescription is required, but there's a possible aid to you. I've used larger syringes in the past, but they're hard to find, too. I even asked my pain doc if I could have the syringe and A needle when he's finished giving me Steroid injections. He looked at me like I suddenly grew an eye in my forehead and said he's required to properly dispose of used medical supplies. THEN he asked why I wanted them, and when I told him he was convinced I should be committed. Maybe he's right. Good luck finding what you need.
 

Raendor

Gawd
Joined
Sep 21, 2015
Messages
845
I delidded my 6700k using Rockit88 back in December 2016. Was a great experience and very easy. Used liquid metal under IHS and ordinary stationary tape to protect the area around the die itself. My temps decreased significantly, but the CPU itself stayed as much of potato as it was in terms of OC sadly. Hope there will be an 8-core CFL or Ice Lake cpu will also be 1151 socket or similar shape allowing to use the tool once more in the future. Not planning to upgrade from my current cpu just for gaming with existing lineup.
 

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
51,126
What type of nail polish do you recommend?
Any will do. Just make sure when you lay it down that you do not paint across substrate where the IHS comes into contact. You don't want a "lump" off paint there.
 
Top