Coolaboratory Liquid Pro Vs Artic Silver 5 Test results

[Martinm210]

Have heard a mixed bag of good and bad through my searching, but overall it seems like it could be worth a couple of degrees over AS5, so I had to try it myself.

The fact that some people won't touch it saying it's a total PITA to install was almost appealing...ahh a new challenge..

This is the package that you get, a very small siringe with a hypodermic needle. Careful you don't want to stick yourself with this thing, or leave it out where kids can get to.


Installation:
It's a very time consuming PITA, that's for sure, but it's also very important I think that you take your time and rub the stuff onto the surfaces well enough to break the surface tension and gain bonding. Expect a good half hour just to install, and also expect you will have to remove the processor and waterblock to apply properly in good light.

Here is a bead on the IHS:


Spreading the stuff around, working my way from the center out:


IHS done and installed:


Block coverage:


So far, I'm very impressed, 2.5C drop in idle and 3.4C drop in TAT 100% load temps is very noteworthy for a simple TIM change, although it doesn't come without some negatives. I've heard several people complain that the stuff has a tendency to want to glue your block to the chip, or that it's difficult to remove the paste from the chips/blocks. The CLP brochure does come with notes that removal may require a metal polish to remove the paste, and it's very explicit about NO ALUMINIM! coolers period...it will eat aluminum. Regardless, I'm all for the added risk and difficulties involved considering the outcome, that's a substantial improvement IMHO.

Below are water to core temperature differences. Note these temperatures were all recorded using TAT's log function and actual core temperatures. Core temperatures on my 680i are typically around 5C higher than the motherboard temperature. For example at 20.3C ambient that I currently have my MB temperature is 31C, and my TAT core temperature is around 36C. Also I used TAT 100% load which is the most extreme loading tool available and much more intensive than any other program including orthos small ftt, real world loading is normally half that amount. I'm charting water to core temperature because it takes out the variable of radiator performance as well as ambient temperature changes. My air to water temperature at load is right at 4C at TAT 100% load if I let it load for 10 minutes.
Idle water to air temp difference is usually just under 3C. This is with my CPU only loop on an MCR320, D5 variable, with Yate Loon D12SL12 fans on a 1/2" foam fan shroud.

All of this is done with fairly inexpensive digital aquarium probes that are nothing fancy. I try to improve accuracy by taking alot of measurements and I think this gives me a good "relative" comparison, but actual measurements could be off slightly due to the equipment used.

Enough of that, here are the Final results:
(Update 9-23 after 40 Hrs Cure Time)
Data:
Each test case number was done by logging TAT core temperatures for 10 minutes while manually logging air and water temperatures on my digital probes. What you see in the test line is the average TAT logged temperature (usually 100 or so results), and my average of 10 air and 10 water temps.


Graph:


Also testing on my 8800(Stock non-lapped IHS), but I've noticed it doesn't like to bond nearly as well with the nickel plating as it does pure copper, so I'm thinking you see the greatest benefit with a lapped IHS where you end up with a Copper/CLP/Copper sandwich.

[jokester]

What technique did you use to apply the AS5, as it can be very sensitive to how thick a layer you use (thinner being better obviously)?

Jokester

[MegadetHCl]

Good stuff, man! Nothing like trying the hard stuff by yourself. I agree, the T difference is significant.

[Fairydust]

Yup so far the results are similar to mine.

[Martinm210]

Originally Posted by jokester

What technique did you use to apply the AS5, as it can be very sensitive to how thick a layer you use (thinner being better obviously)?

Jokester

I have been using the thin line method for a while and multiple remounts until I get it right. I've done so much remounting lately, I think I'm fairly consistent now.

[Martinm210]

Originally Posted by Fairydust

Yup so far the results are similar to mine.

Very informative..thanks!

[hecktic]

Martin those are some excellent results.

Just a note for clarification, the CL Pads or basically version 2 of CLP will work with Aluminum sinks with no problem at all and the cleanup is much easier than CLP since they provide instructions with their own cleaning kit included with the pads.

[Martinm210]

Originally Posted by hecktic

Martin those are some excellent results.

Just a note for clarification, the CL Pads or basically version 2 of CLP will work with Aluminum sinks with no problem at all and the cleanup is much easier than CLP since they provide instructions with their own cleaning kit included with the pads.

Thanks, suprised me too. I think the few reviews that didn't note much difference were either air coolers or they tried to use traditional paste methods to apply it. I think it's really important to take your time and rub this stuff around until it bonds.

I haven't tried the pads, have you? I was thinking the liquid performs better if applied right and you don't have to purposely melt anything.

This is where I got mine:
http://www.sidewindercomputers.com/colipro.html

But I see they also have the pad version.

I'll probably stick with the liquid, I should be able to get 10 or more applications with one tube.

Not looking forward to it, but I think I'll reset the GPU and chipset blocks with this stuff. The chipset scares me a little with no IHS, but I'm going for it anyhow..

[NaeKuh]

once again martin you amaze me. Great review.

The problem with liquid pro, is its very very dangerous on alu, or anything made of alu.

Also i hear it bonds to the sink over some time.


The difference i would say isnt worth it over shin etsu, seeing how sin etsu gives better temps then AS5.

[nibble]

While the difference is a relatively large margin it's just not worth it, it WILL fuse your block and IHS together over time.

[Jedda]

Originally Posted by NaeKuh

once again martin you amaze me. Great review.

The problem with liquid pro, is its very very dangerous on alu, or anything made of alu.

Also i hear it bonds to the sink over some time.


The difference i would say isnt worth it over shin etsu, seeing how sin etsu gives better temps then AS5.

That's how I've felt about it too.
Shin-e, MX-2 without the 'bonding' problem will do me.
Still it's interesting to see Martin's experience. Early days, so far. Does that stuff cure over time and heat cycle for a bit better performance?

[AndyM]

This 'bonding' issue is pretty much a non-issue IMO. I've had it 'bond' my lapped CPU and block together before also. A simple twist and it snapped apart. It's not like you will need a chisel or something to seperate them. That said, I love this stuff, noticable improvement in temperatures, and can really transfer higher heat loads into your block when running higher vcores and speeds compared to standard thermal pastes such as AS5 I've found. A bit tricky putting it on, you can put too little on at times, but with good 'wetting' without going overboard so that you get a some runoff, LM is awesome stuff. Q-Tips work great for swabbing it on if you dont mind a bit of lint in the interface. Now if I didnt feel like some junkie with that syringe around....

Great review as always!

[TinTin]

I still have many in my home, performance is very good, but some bought from other country are old formulae, it will dry and then the CPU sticked with the waterblock.

Disadvantage is not easy to remove, always has some residual on the CPU IHS and then the words on the IHS can't be seen again

PS: I use it for my lapped CPU, so no problem

[Jedda]

Andy, that sounds more like a vacuum seal?

Is there any clear way to differentiate the various versions of the stuff?
How do you know it's the latest?

[AndyM]

Not really a vacuum seal, the LM was dried out/solidified and not liquid anymore. Since the LM sticks so well to metal, but is itself a 'weak' metal, the film of LM is what is being broken to seperate the parts. I'm not too sure how it changed to that state from how it comes out of the syringe, but the 2 surfaces were 'dry' after taking the block off.

I tried to clean the LM off afterwards with alcohol, but I think sanding is the only way to really clean it off. Thankfully you can layer the LM with the same thermal transfer performance as with a single layer it seems. A fresh coat, quick check of the contact pattern between the block and cpu to make sure I put enough on, and my machine was good to go.

[KaptCrunch]

viscos seal

[hecktic]

Originally Posted by Martinm210

Thanks, suprised me too. I think the few reviews that didn't note much difference were either air coolers or they tried to use traditional paste methods to apply it. I think it's really important to take your time and rub this stuff around until it bonds.

I haven't tried the pads, have you? I was thinking the liquid performs better if applied right and you don't have to purposely melt anything.

This is where I got mine:
http://www.sidewindercomputers.com/colipro.html

But I see they also have the pad version.

I'll probably stick with the liquid, I should be able to get 10 or more applications with one tube.

Not looking forward to it, but I think I'll reset the GPU and chipset blocks with this stuff. The chipset scares me a little with no IHS, but I'm going for it anyhow..

The pad version is a lot easier with a one-time quick cure in period which is completely user controlled. It works on aluminum and you can cut up custom sizes of the pad as needed.

Word of caution though, I would not use CLP or CL Pads on anything that does not have an IHS. This would include chipsets but I suppose if you must try it then go for it but be careful and come back and tell us the results

[Exedy]

Nice comparison

I just need to know how much Liquid Pro better than Thermal pad from Coollaboratory

[Martinm210]

Thanks guys,
It does appear to be curing as i'm seeing slightly better temps that posted above.

I'll see if I can finish up my testing and post results later today.

I also think the bonding issue is not that extreme, sounds like some may get stuck on the blocks and IHS and you'll have to remove it with polish or lap it. I lap both routinely anyhow, so that's not a problem for me.

Heck I was even tempted to try and remove my IHS in hopes of 2C or better, never would have thought I could get that with this stuff.

I went ahead and put it on my 8800gtx and D-tek Fuzion GFX last night, so I'll have a second mounting to compare it to. My 8800 just has a stock non-lapped IHS. I'm seeing gain there too, but maybe not as significant as with the CPU.

So, I'm beginning to speculate this stuff works best on a lapped IHS with direct copper to copper bearing....but that's just speculation..

[hecktic]

Originally Posted by Martinm210

So, I'm beginning to speculate this stuff works best on a lapped IHS with direct copper to copper bearing....but that's just speculation..

Yah maybe....

[Martinm210]

Ok, I've completed my testing and now have included the tabulated raw data and a newly updated graph. I posted all this in the OP:
http://www.xtremesystems.org/forums/...d.php?t=159752

In the end the improvement using CLP was:

2.5C Idle
3.4C Load

Better than I expected..

[leo_bsb]

thanks martin, nice work.

[003]

Originally Posted by hecktic

Word of caution though, I would not use CLP or CL Pads on anything that does not have an IHS. This would include chipsets but I suppose if you must try it then go for it but be careful and come back and tell us the results

I use the liquid (non pad) version directly on the core of my 2900Pro, and it works great.

[Nate P.]

Should this go in the user reviews section? Great job BTW Martin!

[disruptfam]

where abouts does mx2 come in? is it between as5 and liquid pro?