Cooling Carbon-Carbon tiles + Peltier Question

Well heatpipe tech is still relatively new in the enthusiast market. Im afraid you wont find a high power heat dispurcing unit on a laptop. But the ides on some of them havent changed mutch since the p1 days. The idea is removing the heat from one spot and transfering it to a larger area. For you chipset you could implement a heatpipe sink and move the heat towards the rear of the mobo where the exaust fan is (abit and asus are both doing this) and the heat output is not high enough to warant anymore. If you want better temps sandwitch a 20w or a 30w pelt and copper plate and heatpipes with large fins on the end to disperce heat and it will be more than enough for a chipset.

80w pelt can be cooled with air but you need massive amounts of surface area. I have a 320w unit that I powered at 7v and it frezes on one end after a fiew seconds but the heat reaches peak heatsink capacity after a fiew minutes and then the cold side becomes hot and I just shut it off for fear of pelt suicide.

 

I'd like to see the diamond for one of the A64s That with phase change cooling on the back side... sounds good to me, who wants to sponsor me?

 

How about no?

Dude its only going to make a marginal diference anyway. At some point its down to the phase units capacity to handle heat more than it is on the jackets capacity to transfer it. Copper is just fine IMO. If you want better temps the increase offloading capacity insted of trying to increase load.

 

Double-carbon composite is made by layering raw carbon fibers with some carbon dust in-between, and then compressing and heating in an autoclave. I very seriously doubt that one could make this at home (in any normal home workshop at least), and I don't know anything else it is used for.
------
I don't know anything about the thermal properties of the shuttle tiles, other than they can withstand some extreme temperatures. One material that sounds familiar in that regard is monolithic graphite, which is used for making cruciables that are used for melting glass and metal in.
By the looks of prices at this site: http://www.graphitestore.com/
it doesn't appear to be very expensive. A sheet .25 thick and 4 x 4 inches = $12.38, not bad at all. But their spec sheet says the metric thermal conductivity is 83 W/(m^2/K/m) where this page: http://www.engineeringtoolbox.com/thermal-conductivity-24_429.html
says copper is 401, about 5X better. They list aluminum as 250 but that is probably for pure aluminum; silicon-Al is somewhat higher than that but still not higher than copper. (carbon and graphite are close but not quite the same thing) The only material listed better than copper seems to be silver, a bit better at 429 but not much.
------
Synthetic diamonds seem to be the best solid available material now, price notwithstanding: http://www.aip.org/pnu/1993/split/pnu131-2.htm
-however-
carbon nanotubes do lots of odd things too and large-scale manufacturing is only now becoming possible.
~

 

To the best of my knowledge, that being a degree in materials science, alloying materials lowers their thermal conductivity. This is the case with aluminium silicon.

Silicon is added to aluminium to improve its castability. However, it does generally have a higher thermal conductivity than typical engineering alloys.

There has always been a lot of discussion regarding novel materials for waterblocks, particularly the base, but lets be honest, they are expensive and with copper being as good as it is, all you are trying to do is improve the one feature of cooling technology that doesn't need that much improving. By this I mean that getting the thermal energy from the base of the block to the surface in contact with the water is becoming the least significant portion of the total thermal resistance from cpu to water.

Now, interface materials is somewhere where we really could do with some novel ideas. If we consider the sequence of thermal resistances from cpu to water -

cpu to base plate
base plate to block internals
block internals to water

changing the material of the block changes the middle one, and by a multiplication factor, so if we get double the thermal conductivity, we halve the resistance.

However, cpu to baseplate is a pretty darn large resistance in the grand scheme of things, and yet we only attribute 2-5% of the cost of the block to this area.

So anyone with some spare creativity kicking about, please direct it to getting the heat from the cpu into the block itself more efficiently.

8-ball