Liquid pro. Anyone use it?

JonJ678

JonJ678

JonJ678

Soldato
Joined
22 Dec 2008
Posts
10,369
Location
England
This stuff. Gallium solution, tends to attack most metals. Highly electrically and thermally conductive. Data sheet warns not to store it in glass containers, but thinks it would be hard to kill yourself with it.

Just ordered a syringe (!) of it, wondered what peoples experiences of it are on here. Cost about £8. I'm going to put some on the intel hsf, find out if it actually attacks aluminium or not. Obviously I also have plans on putting it on a processor

Cheers

The manual makes for some entertaining reading, I recommend it

There is amazing comedy availabe on the web about this. My personal favorite is the review which put as5 on first, cleaned it off with isopropyl then wondered why this didn't work well. A lot of people are upset that it 'eats copper' despite fairly clear evidence that it alloys with the surface. I have absolute faith that this is going to be fine. Thermal paste should fill in the little pores, gallium is going to do an excellent job of that. The manual suggests lapping to remove all traces of it, which is also amusing. I think I'm more excited about the syringe than about the waterblocks
 
Last edited:
It eats aluminum, not copper. lots of video's on youtube of this.
Tends to 'weld' its self to copper more that nickle but works best with very flat interface.
Remove with a twist to break the surface tension/alloys.
Very poor with direct touch heatpipe coolers (but then normal TIM application is not ideal either)

Make sure to prep the surfaces - as with painting, the final result is all about the prep.

More hassle that it's worth tbh - but when used with the right hardware it's the best by miles

Not ideal for watercooling - but the hair dryer 'burn in' works with solid metal blocks
 
Last edited:
Yep, gallium corrodes aluminium. It doesn't do it to copper, which is why I'm amused by the reports that it does based on the alloy layer on the copper surface which is filling all the holes...

I think you're spot on in saying that it's all about preparing the surfaces with this one. Even small amounts of AS5 in the crevices will hurt performance. I have a suspicion that once applied you don't need to remove it, even if you change to a different paste over the top.

The intention is to mill flat and lap the waterclock, lap the processor. Probably both to a mirror shine since this will improve wetting, while I consider it daft with normal paste it may matter with this one. Then clean both extensively with acetone and isopropyl, apply the metal then mount waterblock.

I'm sure heat will help it, especially near the start. It'll encourage it deeper into pores and make it more inclined to alloy with the copper. So I'll run the processor at 70 degrees or so for a few hours after mounting, probably by turning off most fans. This will also encourage it to bind the processor and block together, but I fully agree that rotation is the way to go to release it.

Looking at the block mounting hardware, it's not going to have much give in it when twisted. Probably enough, and worst case I can probably remove the M4 bolts from the back of the board. Might be worth mounting with the screws backwards to make removal easier at the cost of making installation harder.

What's the hair dryer 'burn in' you mention? It suggests turning a hairdryer on it for a bit to encourge the tim deeper into the block. If so, spreading onto the base of the waterblock, heating it, then fitting the block back together may be the way to go
 
Just kiss goodbye to anything you use that stuff on. It's horrendous. I used it extensively until I tried to get anything apart afterwards.

Seriously - it's not worth the hassle.
 
same experience as wja96 here. Used it back on my 4400x2 which basically ended up brazed to the zalman cnps9500. I nearly broke the CPU trying to remove it and the CPU was permanently marked which affected the resale value a bit.

It performs well but absolutely not worth the hassle of applying and more importantly cleanup.
 
I consider myself duly warned, and appreciate this. However this is one (more) mistake I'm going to make myself, it's just too exciting to pass up.

If it were reasonably possible to solder the chip to the heatsink, I think I'd be tempted to. This looks ideal. Lapped block, lapped processor, brazed together. Happy days.

Processor socket is bolted to the board, so probably quite strong. I'll mount it in such a way that I can remove the bolts completely if required, then twist it off. No worries
 
It's very good, got some for testing and under load using my i7 : 4ghz, it beats the Arctic mx-3 by upto 4 degrees!!1

The metal pad however is another story...hard to install and doesn't seem to do much.

I haven't tried to remove it yet but it did come with a cleaning set.
 
Last edited:
I think your experience was misleading. I've found a variety of reviews on this, with results ranging from nearly the same as as5 to significantly better. The common trait to all of those which found it to be barely any better was they did not adequately clean the blocks. Generally they used isopropyl et al, when for this stuff to work properly it needs a completely clean surface. That is, one which has been lapped flat then cleaned extensively to ensure that the pores aren't still full of as5.

However perhaps you're quite correct and it will offer no improvement. Time will tell, though not any time soon. The people I ordered from took money from me (late wednesday) but failed to send a confirmation of order email, it's not arrived yet and should have, and they're ignoring my email asking what the hell they're playing at. So perhaps it will be difficult to source after all
 
People get too hung up on cleaning blocks, it does not make that much difference, if anything AS5 is very sensitive to application method due to it's thickness.
 
With most things I'd completely agree. However this one heavily relies on wetting properties, and a slightly greasy block will hurt this effect. A bit like trying to solder using an iron which has never been kept tinned and is oxidised to hell.

This is my idea as to why some people get good results with it and others get average. I can't see it wicking down into any of the pores of the block if there's semidissolved silicone grease in there. The manual supports this idea.

"Coollaboratory Liquid Pro" conducts heat better than conventional, previously used heat conducting pastes
(improvement by the factor 9 to 150). If these pastes have entered the cooler surface's microscopically small
unevenness, they will form an insulating layer between Liquid Metal and the cooler's surface.

All remains of previously used heat conducting pastes have to be removed thoroughly (!) before applying
"Coollaboratory Liquid Pro". Sometimes, this might not be easy. Often cooler surfaces previously (mis-) treated
with conventional heat conducting pastes need to be polished or ground slightly (1/100 mm).

It is extremely important to clean both the CPU and the cooler - even if they are new. Use a clean cloth tissue and
apply some organic solvent to it. Suitable are acetone or pure alcohol. Not suitable are scrubbing detergents, oil-
or petroleum based detergents or any other substances leaving remains.
Carefully clean both the cooler's and CPU's surfaces chosen for heat transport. The surfaces must be free of dust
and grease (no finger prints!).
Click to expand...
 
Last edited:
Not really, IPA will more than clean a block to the point that what ever is left on it will not have a significant effect on the next application of a supposedly miles better TIM. I would even go as far as saying that just wiping it off with a paper towel probably isn't going to make a substantial difference.

Edit:-
Suitable are acetone or pure alcohol
Click to expand...
^^^
 
I have no data to argue with yet, so I think we will have to agree to disagree for the time being.

This one uses a different mechanism to normal, the idea is that it alloys with the two surfaces. This is probably why it tends to stick things together. At this point any silicone grease in the way is going to play silly buggers with the alloy microstructure, if at the base of every pit there is a small pile of silicone it's going to have an effect.

However with most pastes, it's silicone grease with stuff suspended in it. At this point some silicone in the pores is going to make no difference whatsoever, since you're putting more of the same stuff on anyway. I normally put fairly little effort into cleaning the blocks, I think at present I'm using a mix of as5 and mx2 with no particular ill effects. However I believe this case to be different, as it is not going to be miscible with the previous paste. I'm looking at this as a saner way to solder/braze the parts together, where surface prep is crucial.
 
I was thinking of getting this today but after reading the comments here I don't think I'll bother.

So this is not ideal to be used on a Swiftech GTX?
 
anti5580 said:
So this is not ideal to be used on a Swiftech GTX?
Click to expand...

It is very good indeed as a thermal interface material, but it will make it very hard to get the components apart again. You may consider that worthwhile, ideal even, or you may not.

Personally, I strip my kit way too often to make this realistic for me.
 
You must log in or register to reply here.
Back
Top Bottom