Coollaboratory to Thermal Grizzly Conductonaut

britain4

britain4

britain4

Associate
Joined
6 Jan 2018
Posts
359
When I first delidded my 4770k 5 years ago I used Coollaboratory Liquid Pro (or Ultra, can’t remember) under the IHS which was the best available at the time. I’ve been considering re-doing it partially out of curiosity to see how the old stuff looks after 5 years’ use and partially because I’m getting a 15 degree difference between cores which may or may not be down to the LM under there. I certainly saw improvements from replacing the paste with Kryonaut on top of the IHS.

Anyway the heat conductivity of the Thermal Grizzly stuff is supposed to be more than double that of the Coollaboratory - has anyone made the same switch and seen any results from it?

If not I’ll update the thread once my Thermal Grizzly comes in.
 
Thermal Grizzly Conductonaut = 73 w/mk
Coollaboratory Liquid Ultra = 38.4 w/mk
Coollaboratory Liquid Pro is rated at 32.6 w/mk.
Thermal Grizzly Kyronaut = 12.5 w/mk
Gelid GC Extreme = 8.5 w/mk​

and the list goes on.

but the actual temp difference shows a different picture when used .
Coollaboratory Liquid Ultra = 32.0c
Thermal Grizzly Kyronaut = 32.3c
Gelid GC Extreme = 35.3c with 56 other normal TIM all between 34.8c and 38c​

End result is Coollaboratory and Thermal Grizzly are much harder to apply .. meaning much higher chance of not getting a good TIM print which results in not getting temps as low as with a good TIM print. But most of the top tier normal TIMs are easy to apply and get a good print so good heat transfer and low temps, quite possibly lower than Coollaboratory and Thermal Grizzly give if not having a good print. Also Coolaboratory and Thermal Grizzly are much more espensive.

Boils down to paying way more for a much harder TIM to apply and get a good TIM print that will then only beat normal much lower cost TIM by 3-5c versus using a much lower cost TIM that is easy to get a good mount TIM print almost every time and cheap enough we can remove cooler after first attempt and check TIM print if we are not sure it's good and do it again for a few pence worth of TIM.

3-5c difference really isn't much .. and the top 10 normal TIMs are all within 0.6c if we rule out Thermal Grizzly Kryonaut which is also not as easy to apply and get a good seat.

I'm spend an hour or two checking air temp going into cooler and setting up case fans and fan speed curve to get that air temp going into CPU and GPU coolers is 3c or less above room ambient with CPU and GPU coolers is under heavy load then use a high priced TIM that at best is only 3c better and have no idea what the air temp going into CPU and GpU coolers is with CPU and GPU under heavy load. :p

Here is link to TIM performance graphs http://www.overclock.net/forum/246-...better-cooling-airflow-cooler-fan-data-2.html
 
Cheers for that! I've had a look at that graph and it makes interesting reading. I've used Kryonaut on the GPUs and on top of the IHS and I feel like it must be nearing the limit of the difference a thermal paste can make in those applications. By all accounts Kryonaut is already the best out there so for the sake of 3 degrees I wouldn't consider using the liquid metal on top of the IHS or on the GPUs.

However I'd be interested to see a graph of the effects of different pastes specifically between the CPU die and the IHS... I would think it would look quite different due to the much smaller contact area. I've seen lots of pictures of "effects of liquid metal after 3 months" and the like so curiosity is getting the better of me as to what 5 years will have done to it!

I still need to get hold of a thermometer to check the intake temps of the various coolers but have seen good results from just replacing dried up paste on the GPUs and generic paste on the CPU with Kryonaut. I also test fitted 1 extra fan into the top mount and saw about 5 degrees difference in water temps and GPU temps, so when my 2 extra fans arrive that should make quite a difference too.
 
Last edited:
I don't think it would make much difference. IHS (integrated heat spreader) is really a misnomer. The heat transfer area on IHS is only slightly bigger than the top of CPU die because the thickness of IHS material is too thin to transfer heat laterally away from CPU die under it. The bigger size of IHS is to spread the load out to edges of CPU PCB. This allows cooler to be supported through PCB to edges of socket so there is no flex in area of PCB that has the 1000+ pin connections between CPU and motherboard.
 
That is a fair point - hadn’t thought about it that way. I’ve certainly read conflicting things on it so there’s only one way to find out!

Not necessarily expecting better temps but I can’t remember how good a job I did of applying it and it’s been a few years now so we’ll see what happens.
 
Yeah I'll update the thread with the results, it'd be nice to think it'd knock a few degrees off but I'm not holding my breath, mainly it will get rid of my nagging doubts about the application lol and the "hot" core which could be silicone lottery or could be the way I applied the paste (or it's degraded somehow), we'll see.
 
britain4 said:
~ it will get rid of my nagging doubts about the application lol and the "hot" core which could be silicone lottery or could be the way I applied the paste (or it's degraded somehow), we'll see.
Click to expand...
Was there always one hot core? That's often a sign of poorly fitted heatspreader/TIM, in stock chips. I'd expect you can improve on that with a reapplication. I've also heard CLU is more prone to moving around than Conductonaut, especially if the motherboard is vertical.

If you didn't before, you can gain a little peace of mind by applying tape, nail polish or a similar insulator over the components under the heatspreader. Don't forget to take photos when you crack it open!
 
britain4 said:
Yeah I'll update the thread with the results, it'd be nice to think it'd knock a few degrees off but I'm not holding my breath, mainly it will get rid of my nagging doubts about the application lol and the "hot" core which could be silicone lottery or could be the way I applied the paste (or it's degraded somehow), we'll see.
Click to expand...
I suspect you will get a degree or two lower temps, but I suspect your hot core will still be your hot core. It's quite common.
 
So my Thermal Grizzly stuff arrived this morning and I thought I’d give it a shot.

The die and IHS were practically spotless when I removed the Coollaboratory stuff with some very minor scratching/pitting to the middle of the die. It was still perfectly liquid (as expected) and goes to prove this stuff can be left under an IHS for a long time without needing replacement.

I applied the Thermal Grizzly once and decided to redo it as I thought I’d messed up the application slightly - on removing this stuff from the IHS there was a slight stain where it had been even just for a couple of minutes. The die was still spotless but it was still a little surprising considering the Coollaboratory didnt leave a trace.

Core temps are more in line with each other now but that’s almost certainly down to the way I applied it originally rather than anything else. The first 3 are within a couple of degrees of each other now and the 4th is a few degrees cooler as it always has been. It may have knocked a degree or 2 off overall temps but certainly nothing worth writing home about.

I did take pics and if anyone is interested I’ll post them up but there’s really not a whole lot to see.
 
I'd be interested! Also interested to check out the Conductonaut after multiple years. It can attack aluminium in seconds, pretty reactive stuff at times.
 
I now have a second perhaps more worthwhile update :D

Seems I was taking the instructions to apply a “pin head” of the compound a bit too literally. Only used a tiny blob to get a very very thin but even layer. I had a nagging doubt about whether I had used enough, took the cooler off again and reapplied just a little more so there was an ever so thin pool of liquid on both the IHS and die. Max temps are now down by another 10 degrees on all cores!

I must have made the same mistake first time I did it 5 years ago too - but I still got a decent reduction in temps from delidding at the time and thought it was great! Even improperly applied the stuff made a 15C difference!

I’ll get the pics uploaded now
 
Last edited:
Blimey! I did my 4690k last month and only did a thin scraping too. Some pooling but not a solid layer. I'll redo it when I break down my loop... Which isn't finished yet. Haha
 
Here’s the inside of the IHS - pretty much immaculate (the outline you can see was from the old paste) - there are some very small “pits” near the middle but it’s hard to say what those are from, probably not the LM

kopMKXL.jpg
Here’s what the die looked like - the lighting makes it look a lot worse, barely visible in the flesh
NQBqSCu.jpg


Here’s the effect the Conductonaut had on it after 5 minutes - can clearly see the staining where it was touching the IHS

b322deh.jpg


LuckyBenski said:
Blimey! I did my 4690k last month and only did a thin scraping too. Some pooling but not a solid layer. I'll redo it when I break down my loop... Which isn't finished yet. Haha
Click to expand...

If you had some pooling I would say it should be OK - mine did not, it was a visible but very very thin layer. You do need to use more than it says on the leaflet - but still only a minuscule amount.

Still might be worth reapplying if you’re not sure though!

If your loops not finished I take it you don’t have any temp results right now?

So now that could mean that the temp drops were from 1. Me not applying the CLP properly in the first place 2. The difference in conductivity between the two LMs or 3. The CLP having moved or degraded over the years somehow - or a combination of the 3
 
Last edited:
britain4 said:
If you had some pooling I would say it should be OK - mine did not, it was a visible but very very thin layer. You do need to use more than it says on the leaflet - but still only a minuscule amount.

Still might be worth reapplying if you’re not sure though!

If your loops not finished I take it you don’t have any temp results right now?
Click to expand...
Just with the stock cooler in open air, before and after delidding. Definite improvement, but it seems I have a very "jumpy" chip - temperatures jump 20° the second a load hits it, then go down again. So I get consistent average temps but the peak was in the 90s or 100 before delidding. Now it's more like 65 with split-second 85s IIRC
 
LuckyBenski said:
Just with the stock cooler in open air, before and after delidding. Definite improvement, but it seems I have a very "jumpy" chip - temperatures jump 20° the second a load hits it, then go down again. So I get consistent average temps but the peak was in the 90s or 100 before delidding. Now it's more like 65 with slit-second 85s IIRC
Click to expand...

Not sure if it’s to do with the cooler or not - but I noticed this on my 4770k with my H80i before the Conductonaut - now the temps seem a lot more steady under the same loads and don’t “spike” up into the high 80s way above the average temp like they did before

Just looking at the temp graph in Aida64 it’s way more consistent than before
 
Last edited:
I’d love it if it did work, minty freshness every time you turn your computer on. As long as you don’t slop it all over the place Conductonaut is perfectly fine to work with. Wouldn’t like to spill a drop on my CPU socket though. Easy enough if you take it slow and steady.

Not much point and added danger using it on top of an IHS but direct die it’s fantastic.
 
britain4 said:
I now have a second perhaps more worthwhile update :D

Seems I was taking the instructions to apply a “pin head” of the compound a bit too literally. Only used a tiny blob to get a very very thin but even layer. I had a nagging doubt about whether I had used enough, took the cooler off again and reapplied just a little more so there was an ever so thin pool of liquid on both the IHS and die. Max temps are now down by another 10 degrees on all cores!

I must have made the same mistake first time I did it 5 years ago too - but I still got a decent reduction in temps from delidding at the time and thought it was great! Even improperly applied the stuff made a 15C difference!

I’ll get the pics uploaded now
Click to expand...

Yeah, your 10c lower temps are not TIM w/mk difference, but difference in quality of seat.

But it's nice you were able to lower your temps this much. :D
 
You must log in or register to reply here.
Back
Top Bottom