IS Gold the best Heat Conductor

[rizge]

IS Diamond the best Heat Conductor

Just curious
Because Silver is

 

[Sonny]

As I remember it Gold is one of the best heat reflectors that's why the McLaren F1 road car's engine bay is lined with it.

 

[rizge]

So why don't you use it guys instead of that Arctic Silver?

 

[Sonny]

Because it REFLECTS heat & what you want is the best CONDUCTOR

 

[rizge]

Very Big & High LOL

Sorry didn't realize that
LOL

 

[country_3030]

The reason is due to its density.

 

[AntmanMike]

And it reflects heat. It would be like insulating the CPU from the Heatsink. And BTW, have you guys ever seen the heatsink on a Mac DV? The CPU points into the case, and there is a heatsink built into the case that touches it perfectly. No TIM though.

 

[Tipycol]

Just curious
Because Silver is

Do you mean silver is the best conductor?

And accoding to ATC9001, diamond is the best one:

Originally posted by ATC9001
I was all hardcore into this once...materials that have hgih thermal conductivity. Anyway here's a quick scientific rank order of thermal conductivity (this might be w/mc or something similar I can't remember)


diamond 2000 w/mk
silver 418 w/mk
copper 400 w/mk
aluminum 200-220 w/mk
who cares...but gold is below here, don't confuse electrical conductivity with thermal.

Ok so obviously you can't go and steal the hope diamond and say you just want to make a waterblock. Next best thing is synthetic diamond which can get to 1900+ w/mk. So I go talk to a friend of mine who is a grad student in chemistry with my grand idea, thinking I can make a mold easily I got CAD skills, and if your produced like 5 it wouldn't cost too much...wrong. It costs way way too much money to get synthetic diamond close enough to real diamond, it's dimishing returns too so it may cost 1000 for a 400 w/mk block but would cost 2000 for 500 w/mk etc...the process become insanely expensive.

So that's the scoop on materials....

It's hard to compare heatsinks based on just the metal, because if you made 2 heatsinks identicly outa copper and aluminum, the aluminum may actualy perform better (you'd probably have to try hard to get this but it can be done). It also has to do with the design. The heat spreads differently depending 99% on design and 1% on thermal conductivity. But that 1% (made up figure this is just to show that design is critical) is really important also. I don't know who came up with those percentages of 10-40%...(forgive me if I sound rude, but you can't go around stating that crap with nothing...probably not even a formulated theory in your head) But if you could show me 1 calculation/comparison/scientific principle you used to get that i would truly be impressed. You get small gains going from 200 w/mk to 400 w/mk, you think your gonna get much more going to 418 w/mk....10-40% more still? doubtful. the relation is not linear so don't try to speculate either. I would imagine gains would be so small that when you came to test them, the slightest difference of how you applied thermal paste would negate the results. This is because of dimishing returns and the already small performance increase of copper of al.

silver waterblocks could be done, probably cost around 100 bucks....and you can't just take the best designed copper block...because it'll perform differently due to thermal transfer speeds, although I would take the best copper design over spending the cash to make multiple silver blocks to make a good design. reason for this is flow is a major factor...same thing with heatisnks, fan speed is critical.

 

[Cheesy Peas]

As I remember it Gold is one of the best heat reflectors that's why the McLaren F1 road car's engine bay is lined with it.

you get what you payed for!

Love that car

Isnt Diamond the best conductor... coz of its composition.. each carbon linked to another 4 carbons, so its moves heat throught it easily

 

[dxiw]

you are correct

I won't get into the details of thermodnamics but diamond is the best heat conductor, silvar is much better than gold, then copper and then aluminum.

 

[AntmanMike]

If you are wondering why they don't make "Arctic Diamond", there are several reasons. The first reason is that Diamond, being so hard, destroys what they use to mix the stuff. Another reason is that Diamond molecules are jagged, making it so they can easily transfer heat thru the particle, but not from particle to particle, making it useless. Nevin House has made several batches, and they all sucked.

 

[EBFoxbat]

I could be wrong, my assumption is that a compounds ability to move electrons is what gives it it's ability to move heat... so why is it that a heatsink with an electric current can't move heat faster, its electrons are already moving....and yes, i fully realize the implications of running electricity through a heatsink, i'm just speaking from a thermodynamic stand point....can anyone with more thermodynamic knowledge shead some light?

Also, if diamonds are so great because of their carbon-carbon bonds, why don't they make a copper heatsink with a base of synthetic diamonds? Lab brewed diamonds are far far less expensive than natural ones... put a little artic silver under it and you'd be golden...or should i say diamond

 

[R.Rabbit]

yep thats a good a thougt, but the actual heat transfer is taking place with the copper which would be limiting you, but if this were totally true that would be like saying aluminum heatsinks with copper bases dont get any better performance, so what im trying to say is there's always a limiting factor, it would be easier and more productive to make a direct-die water block

 

[EBFoxbat]

Wel..... make a water block with a very thin synthetic diamond base and copper walls. The water will transfer heat. and the copper will just hold the water in. I mean I assume a thin sheet of diamonds can caontain water. Or replace the synthetic diamond with another carbon-based synthetic material. I'm sure u can make one cheap that donducts heat as if it weren't even there

 

[safemode]

what's the use of super conducting heat for a tenth of a millimeter? thin sheets of anything wont help you lower temperatures. Gold plating something is useful for protecting against corrosion, it's thermal insulating properties aren't noticable at that thickness. Silver plating something for heat transfer is a waste of time. Diamond plating it is also a waste of time. Creating a heatsink of carbon nanotubes from fullerenes would be nice however as these structures make diamond it's biotch. even a thin coating of this would be ideal since you could move the heat laterally easily because the heat transfer is unidirectional i believe due to the tube shape.

Also, heat conductivity has nothing to do exactly with conducting electricity. Heat does not depend on electron transfer. Rather simply the physical contact of one structure to another. molecular/atomic motion == heat. Transfer simply depends on how quickly the molecules start moving and how fast they can bump into the next one.

The difference a material is going to make in thermal transfer is going to end up being some proportion of the amount of that substance the heat has to pass through. At the plating level, it's not going to make any noticable difference since we're talking a less than a tenth of a millimeter.

I really doubt you'll find a substance better than copper until some synthetic is created either from fullerenes or some as of yet unknown compound. Your best bet is to investigate heatsink designs. Finding a way to spread heat out laterally is a difficult problem and this is what we need to do as die size shrinks and heat output increases.

 

[ziah]

there is a diamond heat sink compoud out there! its a hell of a process and the results arent very impressive though.

see this thread!!

~z

 

[EBFoxbat]

The trick is to get the heat of the die. The suggestion of spreading the heat out laterally is good because it adds surface area. A thin sheet of synthetic diamond (say 3 mm thick) would spread the heat out laterally. If this sheet od synthetic diamond were the base to a water block, the water could carry the heat off the diamond. This is probably the second best solution aside from making a waterblock that uses the die itself as the bottom. Synthetic diamonds are not hard or expensive to make. A 5.5cm x 5.5cm x 0.5cm thick sheet of synthetic diamond would be the best bottom for a water block.

 

[Tipycol]

The trick is to get the heat of the die. The suggestion of spreading the heat out laterally is good because it adds surface area. A thin sheet of synthetic diamond (say 3 mm thick) would spread the heat out laterally. If this sheet od synthetic diamond were the base to a water block, the water could carry the heat off the diamond. This is probably the second best solution aside from making a waterblock that uses the die itself as the bottom. Synthetic diamonds are not hard or expensive to make. A 5.5cm x 5.5cm x 0.5cm thick sheet of synthetic diamond would be the best bottom for a water block.

Now I would like to see that waterblock

 

[safemode]

direct die is not better than spreading the heat laterally and cooling the cpu that way. You realize that the temperature and pressure to create synthetic diamonds would destroy our blocks.

 

[Sonny]

I could be wrong, my assumption is that a compounds ability to move electrons is what gives it it's ability to move heat... so why is it that a heatsink with an electric current can't move heat faster, its electrons are already moving....and yes, i fully realize the implications of running electricity through a heatsink, i'm just speaking from a thermodynamic stand point....can anyone with more thermodynamic knowledge shead some light?

Thermo Electric Cooling products, T.E.C./Peltier, are available in the market & have been used with varying success in either air or water cooling.