Forums | MacLife

Kirk wrote:

You squeezed it out of a tube?  Then it is most definitely not really a metal.  It might be a metal powder filled epoxy.

No, not a 'tube,' a...

Bat wrote:

Kirk wrote:

Interesting, Bat.  There are only four metals that melt at a low enough temperature so they could be used for that heat transfer pad, indium, mercury, bismuth and cadmium.  Cadmium and mercury are nasty toxic.

Interesting indeed. (Note that I got it over a year ago, so it's the 'liquid in syringe' form.

...syringe with a very thin metal tip. (The Java enlargements show it fairly well, maybe the reviews do better). It does have the look of mercury, with a slightly thicker consistency. Most CPU interface compounds are metal-based with additives, from cheap zinc oxide paste to Arctic Silver V (something of a standard), and are applied by squeezing from a tube, the resulting droplet then smoothed out with something like an old credit card, aiming for the thinnest layer that still gives full coverage. This, OTOH, really does need spreading with a brush, and will run if not applied carefully on a level surface.

I'll try to save a job to show Levi with next time he's out. (You are coming by soon, right? ).

Or I can save some for you to see at Levi's next LAN party (also soon, yes? ).

Syringes and tubes are for this detail of no practical difference.  That you can squeeze it out of a syringe at room temperature says you've got an adhesive or wax of some sort there that probably is filled with a powder to give good thermal properties.  Alumina and aluminum nitride are common powder additives. I'm sure its much better than thermal grease in any case.

All metal alloy options that one could squeeze out of a tube or syringe at room temperature would contain both cadmium and mercury and we know they're not going to be using those.

The advantage of a true metal is much much higher thermal properties.   Metals have thermal conductivities of at least 50 Watts/meter-Kelvin and on up to over 1000 W/mK.  Filled adhesives are typically around 0.5 W/mK and never exceed 1.4 W/mK.  Still, that's better than thermal greases with typically under 0.2 W/mK.

Kirk wrote:

All metal alloy options that one could squeeze out of a tube or syringe at room temperature would contain both cadmium and mercury and we know they're not going to be using those.

Perhaps, if they needed an element that melts at room temperature.
I forget the month, but a magazine of popsci has a non-toxic (no Mercury) version of the trick spoon trick.

Bat wrote:

add radiator

Nice! hehehe that happens to be the radiator that Danger Den sells with the kit I bought. I see DD doesn't carry the one I bought but this is close.

Let's see if I can finish this up before falling asleep.

Kirk wrote:

All metal alloy options that one could squeeze out of a tube or syringe at room temperature would contain both cadmium and mercury and we know they're not going to be using those.

Agreed. It's low toxicity stuff and not as runny as a 'pure liquid metal' would be (mercury would run right off the CPU. No way you could handle that). It is fairly runny, tho; not for newbs. So...

Syringes and tubes are for this detail of no practical difference.  That you can squeeze it out of a syringe at room temperature says you've got an adhesive or wax of some sort there that probably is filled with a powder to give good thermal properties.  Alumina and aluminum nitride are common powder additives. I'm sure its much better than thermal grease in any case.

...I think it does, from both theory and behavior, have a powder additive to the base metal alloy/compound, to thicken it. You can argue sematics if liquid metal + metal powder = pure metal or whatever they call it. Arctic Alumina and Ceramique are as their names imply; pastes with aluminum and very fine ceramic particles, respectively. Perform well at a good price.

The advantage of a true metal is much much higher thermal properties.   Metals have thermal conductivities of at least 50 Watts/meter-Kelvin and on up to over 1000 W/mK.  Filled adhesives are typically around 0.5 W/mK and never exceed 1.4 W/mK.  Still, that's better than thermal greases with typically under 0.2 W/mK.

...and you run into diminishing returns. {Background for non-engineering types } The TIC is there, not because it conducts heat better than metal to metal, CPU to heatsink contact, but because those can't be made perfectly flat, hence have gaps, albeit small, sometimes microscopic ones. All the recent CPUs have integral heatspreaders, too, which are 'gappier' than the older, unprotected die types. Enthusiasts sometimes even lap their heatsinks' bases for better flatness, or buy same commercially [example]. Unless you fill those gaps with something having good thermal conductivity, you have air in them- and air is a downright poor thermal conductor. I wouldn't buy 1 CPUs' worth of a REALLY pure liquid metal at, say, $50; let's hope this new stuff hits a price/perf sweet spot, where heatsink performance, not the thermal interface material used, is the main limiting factor.

...

Bat wrote:

[Update, Levi?

Nothing to report. I had a four day weekend iwth the best intentions of working on LEVIATHAN II but I went fishing instead. Caught some nice smallmouths...too purdy to be stuck in the basement tinkering.

Stage Two: For the most part LII is done. Now I need to clean up it's home. The last job I had I finished but left the mess it made...papers and CDs all over my lil corner of the basement. Gotta spend a evening picking that all up. After that I can sit at my desk and start software installs.

Maybe it's the revised audio stack in Vista, eliminating the HAL, but Booksley had a similar problem under Boot Camp.

AC '97? You haven't spent the $10 to get your Audigy working on Vista?