Liquid Metal Thermal Paste

[Sentential]

I saw someone who lapped their IHS and it was copper....anyone else vouche for that?

Ya I can actually. My P4 lapped is copper as well.

[thegreek]

wow that's some crazy stuff....like many of the others I'm sticking to AS5 myself lol...wouldn't want to try and mess with that.

Well I think putting that on an AMD64 IHS would prove to my friend that they aren't aluminum tops If they were aluminum it'd corrode them right? I believe I saw someone who lapped their IHS and it was copper....anyone else vouche for that?

There is no big difference though, about 1-2 C better

[MaxxxRacer]

IHS can gain you nothing, all the way to 8C better depending on the contact the IHS was getting with the core.

[moonlightcheese]

think about it... you're lapping a piece of metal that sits between the die and the heatsink... so really you're just making a thermal resistor slightly less resistant at transfering heat from one surface to the other. but it's still a wasted piece of metal that no amount of lapping or changing will bring you ANY benefit. just take the stupid thing off XD

[Ray_GTI-R]

Fairydust & xtremesystems - thanks for the info. Have ordered some & will test it on a copper-cored Prescott P4 socket 478 heatsink.

Fairydust - what is the broken black stuff in "...lmfinal4.jpg"?

From the tiny info I've found so far on this product I've heard that Brasso is the thing to clean up LM - followed by the usual aclohol based cleaner followed by acetone just in case. The remains shown after "ordinary" cleaning in the pics that Fairydust has kindly recorded are encouraging to me. Logically, it seems that LM has found some pits and filled them?

[Bloody_Sorcerer]

think about it... you're lapping a piece of metal that sits between the die and the heatsink... so really you're just making a thermal resistor slightly less resistant at transfering heat from one surface to the other. but it's still a wasted piece of metal that no amount of lapping or changing will bring you ANY benefit. just take the stupid thing off XD

unless you lap straight through it to your core and only in the center. then it saves you from core-crushing AND you get the advantage of no IHS!

[moonlightcheese]

unless you lap straight through it to your core and only in the center. then it saves you from core-crushing AND you get the advantage of no IHS!

ROFL. you be the first to try it bloody haha!

[chinkgai]

unless you lap straight through it to your core and only in the center. then it saves you from core-crushing AND you get the advantage of no IHS!

roflmao

[Fairydust]

Fairydust - what is the broken black stuff in "...lmfinal4.jpg"?

Must be the camera angle. Despite the black appearance on this picture it actually was all shiny and silvery in reality.

[Nubius]

well I wasn't necessarily talking about temperature benefits of lapping the core, moreso I was just confirming that it was copper

[Ray_GTI-R]

Got some in the post today and I've been playing around with it ever since. This stuff is truly alien technology - it's doing my head in.

Fact:- it's very liquid and the hotter it gets the more liquid it gets
Fact:- as recommended, on aluminium it's very reactive. I'm not talking surface markings, this stuff just eats fresh aluminium
Fact:- put too much on and it will drip, if it gets hot it will drip more. Not a problem if your heatsink is parallel with the ground.
Fact:- at first sight it has a surface tension, just like water - when first applied to a bare metal heatsink it just sits there, you have to "work it" to get a thin film coverage like the ads. When I applied it with a cotton bud I was alarmed by what seemed to be impurities left behind - that's normal, just excess material, not impurities ... maybe the reason for my earlier question.
Fact:- it's very difficult to photograph with intense flash. It'll look black or just completely fool the camera focus, hence the question/reply by Fairydust
Fact:- neither petrol, alcohol nor acetone(!) will completely clean off this stuff. In my case, only Brasso does the trick. Reading the supplied instructions reinforces this. I've double-checked with an 8x jewellers' glass.
Fact:- it's conductive OK. In the course of playing around I obtained a 1" long thin layer with zero electrical resistance along that length.
Fact:- it won't dry out (I was hoping it would "cure" a bit to make a more pliant material). I played a hairdryer over a die-sized area and it just got wetter as the temperature went over the too-hot-to-handle boundary and blobs of it ran together & followed the direction of gravity.
Fact:- it won't shatter glass in the short-term. I wanted to see the results in action so I used a glass sheet to observe the way it spread on my copper core. When I finished I cleaned off the glass (with petrol, then Brasso, then petrol) and the glass was undamaged.
Fact:- it won't cure big imperfections with either die or heatsink but minor imperfections are handled OK. The instructions quote a layer of 0.003 to 0.005 mm. The Intel copper-cored heatsink I used had a huge dip at the centre (no wonder it was cheap - I've never seen anything so bad before, it was obviously a second). Small imperfections are handled very well.

Opinion:- if I were to use this (a big if at the moment) any setup would have to have a motherboard that is parallel & level with the ground. At the more usual 90degrees to horizontal I would have to be very sure that the amount that I had applied was right and that no drips would form when the CPU got hot.
Opinion:- The instructions say use on the heatsink and/or the spreader. I tested it on just one surface (the heatsink) and then on two surfaces (heatsink and a piece of clear plastic, just so that I could see what happens). I'd recommend using it on two surfaces, the joint just seems to "stick" better when applied on two surfaces.
Opinion:- while I've been typing this (took an hour!) I've had a sample in the freezer. Under -24C (limit of my thermometer) and it's still liquid!

Unknown:- does it expand significantly when cooled significantly below ambient? My guess is no???
Unknown:- does it give off a gas when in contact with aluminium? My guess is yes???
Unknown:- what it's made of. My guess is that it's definitely alien technology???

[Fairydust]

What is it made of: The Answer is 47.

The result of your freezer experiment are not what I got, mine frooze solid pretty quickly.

I didn't have any dripping problem during my little test, the surface adhesion and cappilar action held it firmly in place on my vertical testbeds. Finding the right amount initially took some tries, but after that it was pretty straight forward. I guess it's a little harder with the IHS as there is way more area.

Do some more temperature tests, that's all they ever want :P

[One_Hertz]

Do some more temperature tests, that's all they ever want :P

Yes, please do so I am quite interested.

[Ray_GTI-R]

Doing this stuff is all pretty new to me and after 24hrs, I'm still testing the physical side of LM before trusting it on one of my precious PC's (I don't have a testbed PC ... yet!). Also, I don't have any proper test tools/software for testing the temperature so my results will be very unscientific. I've got MBM5, Sandra Pro & Abit's hardware monitoring but that's it. All I can say right now is that Fairydusts results are OK with me and I'd use LM as my first choice for my next project, relegating my ShinEtsuMicrosi to 2nd place. When I get round to it I'll post my unscientific results.

[Ray_GTI-R]

unless you lap straight through [the IHS] to your core and only in the center.

I understand that there is a layer of thermal pad material sitting between the core and the IHS. So this lapping method - as with complete IHS removal - eliminates both IHS thermal resistance and thermal pad thermal resistance. Unlike complete IHS removal, any excess pressure from the heatsink fixings is still taken up by the remaining ☺-shaped surround acting as a physical brace. Which is what it was designed for (badly, I know but we're talking mass-production engineering not "The Best"). Removing unecessary material whilst retaining physical integrity sounds like good, simple engineering practice.

I'd save time by cutting off most of the top of the IHS with my favourite zinc-copper removal machine/s leaving a roughly ☺-shaped surround, de-burr, remove any thermal material from the core then give a careful lap.

Now, seriously. Where can I get a zero-static sealed ZIF 478 socket to load the chip into during all this brutal work? (yes, I've seen the modified static bag method, thanks ).

[fareastgq]

Now, seriously. Where can I get a zero-static sealed ZIF 478 socket to load the chip into during all this brutal work? (yes, I've seen the modified static bag method, thanks ).

I call it electrical tape

[Ray_GTI-R]

I've taken a few pictures of the first 24/48 hrs of my playful experiments with LM (the physical aspects). I recommend first reading the tiny text file of my description of each pic before you download megabytes (just in case you are expecting something else). The text file is here;- http://freespace.virgin.net/ray.w/LM/lm2.txt

There are 18 pics here;- http://freespace.virgin.net/ray.w/LM/

Please note that these pics are around 500K each or thereabouts as I wanted to keep picture quality. FTP'ing them onto your machine first might be an option whilst also reading the descriptions?

Any/all comments welcome. Also feed me back if you have any problems accessing anything I've linked in this message.

HTH, Ray

[Ray_GTI-R]

Just read this elsewhere on the Interweb;-

"an alloy of 24% indium and 76% gallium is liquid at room temperature" &

"[this product] has no cure time.... it'll always be in a liquid state*, it is electrically conductive** ... toxicity shouldn't be a problem."

* I'm guessing that means "at moderate temperatures" like my something under -24C test.
** I agree, it's very conductive.

Also spotted on biomass & soot ... latest cutting-edge graphics card heatpipes with liquid metal in them - would require no mechanical pump, could be driven by magnetism making it very quiet, manufacturers won't say what the liquid metal is.

Do we want to know more?

[Fairydust]

I still don't know why your Metal didn't solidify in the freezer, how long was it inside? -24 C is way below room temperature, for melting points of Gallium alloys check out this site.

In the discussion about the liquid metal graphic card coolers from nanocoolers (Sapphire Blizzard cards), we already knew it was a gallium based coolant. The nanocoolers website has been changed, and liquid metal cooling isn't even mentioned anymore. The Inq has a pretty good article on it. It just isn't better than water cooling in a PC application, but a lot more problematic and expensive.

[jackyl]

I'm very interested in using this stuff "I have a nitroused moped " But It's nothing more than taking your time with it. Like using a phase change setup.

I've see the auction on ebay . I've read through countless articles on this stuff. And honestly I'm sold. But I just want to know if anyone can tell me what the difference between this and the stuff that is being sold on ebay is? If it's the same stuff I'll spend the 30bucks plus shipping and never run out of thermal compound applied with a model paintbrush kicks ass!

[chinkgai]

the 30 dollar pic is like the same size as the pic from ebay...but over twice the cost of the ebay syringe...

how much exactly is the 30 dollar one? did i miss the exact amount? it escapes me.

[jackyl]

Well if it's the same stuff, that glass tube that is between the fingers would fill about 50 syringe's. But that's just from the looks of what you get in a syringe. I'm just wondering if it's the same stuff or what? The main pics are the same. Hard to say I'm just waiting to see if someone knows.

[chinkgai]

from the OP

Gallium Indium alloy with traces of precious metals (probably as anti-corrosion additive)

sounds like its not the same as yours, since the link you sent states:

This element (atomic number 31)

pure gallium vs a mixed alloy

[Fairydust]

chinkgai is right. The $30 dollar container is pure gallium, no alloy.

[freecableguy]

I'm going to bring this thread back to life. I have installed a pair of DD Low Profile Maze4 GPU sinks on an identically clocked, identically volted 7800GTXs. In a few days they go into my new SLI watercooling loop. One has this new Coollaboratory Liquid Pro liquid metal thermal compound on the block and the die. The other has AS5 only. We'll see which one does better. I hope the answer is the liquid metal because it looks like it's going to be a to get off the Maze4 if I need to!

-FCG