Gold leaf the bottom of a heatsink?

[Twigsoffury]

well i've played around with some gold leaf in the past and i've always wondered if you leafed the bottom of a heatsink really thick, then sat down for a few days and rubbed that leaf down with rouge to a jewlers finish if it would make a difference?


http://i53.tinypic.com/ighvdv.jpg

[slaveofconvention]

I'm no expert but I would think that whatever gains you might make from the higher heat conductivity of the gold would be pretty much cancelled out (or utterly defeated) by the small amounts of adhesive...

It'd look cool though :P

[OvRiDe]

I have to agree with SoC. It may sound good in theory, but it still has to get from the gold leaf to the block, so it seems like it would perform about the same. Also gold's thermal conductance is better then aluminum but is still lower then both copper and silver.

Still might be interesting to see.

[Aldersan]

Id say just put it on the outside and make it look rich without affecting performance so much :p

[Twigsoffury]

I'm no expert but I would think that whatever gains you might make from the higher heat conductivity of the gold would be pretty much cancelled out (or utterly defeated) by the small amounts of adhesive...

It'd look cool though :P

oh they have adhesive-less leaf, the sheer amount of rubbing it onto the surface and contact area keeps it on.

I've got lets say... "tobacco" pipes that i've gold plated that don't even have a scratch on them and they are some years old now at this point.

[DeviceUnknown]

Do it man, put a piece in there and start-r-up. before you do though record temps, then put gold in, and record temps again. Let us know. If this works i will definitely do it myself.

[Twigsoffury]

Do it man, put a piece in there and start-r-up. before you do though record temps, then put gold in, and record temps again. Let us know. If this works i will definitely do it myself.

oh well it'll take a some hours to polish it out before you can attach it or it'll just fall back apart.

[Snowman]

i say try it you have my interest

[Kayin]

It would lower the thermal conductivity, so I wouldn't do it. Now leafing a waterblock to make it noncorroding would be interesting...

[Twigsoffury]

Well looking here

http://physics.info/conduction/


It says that silver is significantly better then gold at conducting heat.


I know for a fact they have genuine silver leaf at the hobby store.


Hmm.... and its like 1/3rd the cost of gold lead.


hmmm...


[edit]

Also why the hell is carbon graphite 1950 Wm/k?

Should i be making me a silver heatsink with pencil heatpipes?

also looking at doing this.

-d9ugkIFgdU

[Kayin]

When I sort out if I'm headed to the hospital tonight, I'll send you some info on that. Carbon is directional, it's not unidirectional like silver.

You're showing the right kind of ideas to make your own heatsinks or water blocks. Research some homemade blocks, you might be surprised.

[Konrad]

Agreed, gold is an inferior thermal conductor compared against silver or copper. Applying it would diminish efficiency and present very little swag since it's a completely hidden part.

But, if you insist on putting the golden bling on your heatsink I would personally recommend using an electroplating or chemical plating method (same thing, really, since electricity isn't strictly required and only speeds up the process). Layers of adhesives and trapped air, no matter how impossibly thin, will act as insulators and greatly impede heat transfer.

My past obsessive craze with improving thermal interface material wasn't really worth the effort. Carbon (that is, aerospace Highly Oriented Pyrolytic Graphite, HOPG) was fairly effective but not significantly more than a pure copper-to-copper interface using a quality silver (or phase-change alloy) based TIM. It was also tremendously difficult to apply without damaging the chip, fuses the processor IHS to the heatsink in a permanent fashion, and it's basically impossible to ever remove. Molecule-thin plasma-deposited diamond coat layers would be much easier to implement (in a fabrication facility), but they would again require the use of a (semi)liquid TIM application and so wouldn't present any overall gains in heat transfer efficiency.

[x88x]

It would lower the thermal conductivity, so I wouldn't do it. Now leafing a waterblock to make it noncorroding would be interesting...


Agreed, gold is an inferior thermal conductor compared against silver or copper. Applying it would diminish efficiency and present very little swag since it's a completely hidden part.

/\Those.

In terms of naturally occurring materials, the top end goes:
Diamond
Silver
Copper
Gold
http://en.wikipedia.org/wiki/List_of_thermal_conductivities

Several synthetic carbon structures (nanotubes, graphene, etc) beat out diamonds, but really all you really need to worry about (unless you want to go overboard for the sake of going overboard...Konrad.. ;) ) is the Silver, Copper, Gold. The reason gold is used in electronics a lot, as Kayin referenced, is not because it is has a higher electrical or thermal conductance, but because it does not corrode. This makes it very appealing for I/O connectors...not so much for thermal interfaces..

[Twigsoffury]

Well i'm not a nasa laboratory, but i've made some hilarious things in the past that were pretty complicated and talking to a buddy who has access to the good acid and wash tanks for cleaning off aluminum and copper before powder coating, so i'm sure hanging my lil old heatsink off of a rod for a few seconds wouldn't cause much trouble.


this is the spare heatsink i have.

http://www.txcesssurplus.com/catalog/10384-amd-heatsink-2.jpg

(mines a little bit more like this on the bottom though

http://4.bp.blogspot.com/__3bnHqr5jYE/TSIwTBxQrpI/AAAAAAAAAIo/ueqoij1nHyE/s1600/amd7.jpg


I've also got one of these

http://www.dansdata.com/images/quickshot011/v12base640.jpg

725 grams of pure copper ( 0.44°C/W...thanks thermaltake )


Now what i'm wondering is, if aluminum and copper electro-plate at different rates because of the differences in electrical conductivity ?


But if it does...does that mean i might not get a even coating on the aluminum and copper heatsink? I was looking at the aluminum/copper heatsink and i think i could just dip the lower part of the heatsink into the acid on the + side with the crumbled up silver leaf on the - side and wait for about an hour or so depending on the amperage used. ( i'm thinking a low a@# 200mA 3.3v)

[AmEv]

Diamond?!?


Hey, now I want to have a huge diamond and carve it out for my graphics card!

[Twigsoffury]

Diamond?!?


Hey, now I want to have a huge diamond and carve it out for my graphics card!

wheres a super ballin rich computer guy when you need him.


That facebook guy should have a diamond heatsink...lets ask him on facebook.

[AmEv]

Bill Gates??????????? lol

[Kayin]

Aluminum won't silver plate.

[Twigsoffury]

Aluminum won't silver plate.

So i would either need to do my large thermal take heatsink, or just the base of the copper/aluminum?


Theres also something i looked at called "strike plating" which lets you plate two different metals with a thin coat first. But it'll involve some dank amperage.

[TheGreatSatan]

In terms of naturally occurring materials, the top end goes:
Diamond
Silver
Copper
Gold
http://en.wikipedia.org/wiki/List_of_thermal_conductivities

That's why I prefer IC Diamond paste

[Konrad]

There are ways to plate any metal onto any other, even silver on aluminum. Although they tend to involve rather extreme temperatures and pressures.

All these thermal conductivity figures assume a passive block. If you have a dynamic mass, say circulating fluid, then heat transfer goes up dramatically. Pimping out your heatsink with gold and diamonds is good for boutique bling, but in terms of thermal efficiency and cost effectiveness it's a distant last place after far more realistic engineering solutions.

It's the old adage about the chain being as strong as the weakest link. No matter how awesome the gold or diamond or NASA unobtainium materials are, they're still gonna be limited in absorbing and dissipating a quantity of thermal energy dictated by the materials they interface with. Even perfect diamond, in direct contact with the chip die, somehow applied in a manner which is 100% efficient in terms of theoretical maximum, would still be limited by the heat transfer capabilities of the chip silicon.

Adding more layers impedes the function of the heatsink. Adding materials which are inferior in this application impedes the function of the heatsink. Spending a million dollars to get half the efficiency of a $100 cooling loop is kinda pointless.

[dr.walrus]

I'm not convinced a tenth of a micrometer of gold would make any difference at all to 700g of copper!

[Twigsoffury]

Spending a million dollars to get half the efficiency of a $100 cooling loop is kinda pointless.

http://s3.amazonaws.com/kym-assets/entries/icons/original/000/004/457/challenge.jpg?1291495361


I'm still silver plating my heatsink, just gotta wait for pay day to roll around. figure ill try it on the spare heatsink and see how it works out. and if it kicks ass.

I'll dip that 100% copper heatsink in and plate that top to bottom in pure silver.

[Twigsoffury]

hehhehehhe wait 4 pix i tried some sh@# with a dime, a old 18k earring , some citric acid , bucket, new 22g wire and a old variable DC transformer.