Copper Foam heatsinks?

Speed addicts anonymous.

Moderator: Starfalcon

Copper Foam heatsinks?

Postposted on Wed Jul 30, 2014 1:26 pm

so, after seeing this:
http://hothardware.com/News/Silent-Powe ... f-Awesome/

it got me thinking.

1. it's going to attract a **** ton of dust.

2. why does it have to be external?

3. No heatpipes? it's all conductive.

But what if we took the standard tower-style CPU heatsink, with heatpipes, forced airflow with a fan, and hopefully inside a case that has dust filters and positive pressure?

would it be better than today's designs? Smaller heatsinks with better dissipation? One problem i've seen is, basically, smooth flowing air is actually bad for dissipation. you want turbulence and high surface area. Foam does that well, smooth parallel fins, not as much.
willyolio
Gerbil XP
 
Posts: 301
Joined: Thu Oct 13, 2005 12:15 am

Re: Copper Foam heatsinks?

Postposted on Wed Jul 30, 2014 1:37 pm

It looks great, but all that mish mosh of copper foam blocks airflow as you get towards the center, meaning the center gets much hotter than the outside. I doubt it's a much better solution than a standard heatsink.
Sheep Rustlers in the sky! <S> Slapt | <S> FUI | Air Warrior II/III
FireGryphon
Darth Gerbil
Gold subscriber
 
 
Posts: 7358
Joined: Sat Apr 24, 2004 7:53 pm
Location: the abyss into which you gaze

Re: Copper Foam heatsinks?

Postposted on Wed Jul 30, 2014 1:42 pm

willyolio wrote:would it be better than today's designs? Smaller heatsinks with better dissipation? One problem i've seen is, basically, smooth flowing air is actually bad for dissipation. you want turbulence and high surface area. Foam does that well, smooth parallel fins, not as much.

Heatsink performance with forced air scales pretty linearly with the velocity of the impinging air. A fan blowing onto this foam would cool the outer 3-4 mm of the windward side, but the rest of the block probably not so much.
This problem was caused by Windows, which was created by Microsoft Corporation.
sluggo
Gerbil Jedi
Gold subscriber
 
 
Posts: 1546
Joined: Wed Feb 16, 2005 8:44 pm
Location: under the table and dreaming

Re: Copper Foam heatsinks?

Postposted on Wed Jul 30, 2014 1:44 pm

then just make the blocks thinner. some tower heatsinks already do this.

http://www.frostytech.com/articleimages ... w_det2.jpg

the most important 3-4mm is the part surrounding the heatpipes anyway. heat conduction can only take it so far away before the material:effectiveness ratio drops to nothing.
willyolio
Gerbil XP
 
Posts: 301
Joined: Thu Oct 13, 2005 12:15 am

Re: Copper Foam heatsinks?

Postposted on Wed Jul 30, 2014 1:51 pm

Elemental copper corrodes quickly in the presence of water vapor (ie, in normal rooms people live in), and copper oxide on the surface will impede the transfer of heat to the air. In order to keep it bright and shiny and non-oxidized like this mockup, it has to be something other than elemental Cu (ie an alloy) or be coated with something that protects it without significantly impacting its role as a heatsink/radiator. Which isn't impossible, but presents a challenge they don't seem to have addressed.

It seems like a better design would just be something that looks more like a conventional modern heatsink -- with heatpipes, etc -- just scaled up to cover the entire top of the mini PC. Or maybe even emerging out of it, like the tubes in some high-end amplifiers. Though "better" is obviously relative; I'm not convinced this kind of passive design is really a winner regardless. If it was really a good idea, I would have expected to have seen it already in other notorious heat-producing consumer electronics -- amplifiers again. The makers of those tend to put a shell around the heat sinks for good reasons: it keeps people from handling them when hot (and then suing when burned), and it allows for some airflow even in the worst case by ensuring nothing gets piled immediately on top of the heatsink in a crowded AV stack.

(Aside: back in the "mobile Pentium IV" days, when packing performance into a laptop without melting it was a real problem, I briefly worked with a guy who had a similar idea for laptops. The entire metal back of the screen acted as a heatsink -- it had a cool-looking fluted texture to increase surface area -- and a closed-loop heatpipe to transfer heat to it from the guts of the system. And yes, despite rigid connections the design allowed the hinge to function while it was running; that was the clever bit. But it was far too expensive to be put into production, and the Pentium M mostly solved the problem soon enough.)
UberGerbil
Gerbil Khan
 
Posts: 9998
Joined: Thu Jun 19, 2003 3:11 pm

Re: Copper Foam heatsinks?

Postposted on Wed Jul 30, 2014 1:55 pm

UberGerbil wrote:If it was really a good idea, I would have expected to have seen it already in other notorious heat-producing consumer electronics -- amplifiers again.


I still have yet to see an amplifier that utilizes heatpipes... does that mean they're not a good design?
willyolio
Gerbil XP
 
Posts: 301
Joined: Thu Oct 13, 2005 12:15 am

Re: Copper Foam heatsinks?

Postposted on Wed Jul 30, 2014 2:38 pm

willyolio wrote:
UberGerbil wrote:If it was really a good idea, I would have expected to have seen it already in other notorious heat-producing consumer electronics -- amplifiers again.
I still have yet to see an amplifier that utilizes heatpipes... does that mean they're not a good design?
http://youtu.be/Fh0O_nF45m4
UberGerbil
Gerbil Khan
 
Posts: 9998
Joined: Thu Jun 19, 2003 3:11 pm

Re: Copper Foam heatsinks?

Postposted on Wed Jul 30, 2014 2:55 pm

willyolio wrote:
UberGerbil wrote:If it was really a good idea, I would have expected to have seen it already in other notorious heat-producing consumer electronics -- amplifiers again.


I still have yet to see an amplifier that utilizes heatpipes... does that mean they're not a good design?

The heat-producing parts in amplifiers (the parts that require heatsinks, that is) have a much higher operating area than microprocessors. Power transistors and FETs can operate at up to 150 C die temperature. For this and other reasons they don't require expensive cooling solutions. Just big fins.

Edit: The Technics amp referenced above is some sort of Class-A design from the late 70's. This is not your "normal" consumer amplifier. It's power dissipation at idle is quite high.
This problem was caused by Windows, which was created by Microsoft Corporation.
sluggo
Gerbil Jedi
Gold subscriber
 
 
Posts: 1546
Joined: Wed Feb 16, 2005 8:44 pm
Location: under the table and dreaming

Re: Copper Foam heatsinks?

Postposted on Thu Jul 31, 2014 10:20 am

My eyeballing guess is that the total volume of copper in a mesh like that would be too small for the distance between the outer surface of the sponge and the heat source; Large heatsinks require heatpipes to tranport heat throughout the distance of the whole heatsink. Years of heatsink trial & error in the development of PC cooling has proved that making heatsinks larger is pointless without heatpipes beyond a certain point. Likewise, the shape of the heatsink isn't anywhere near as relevant as the surface area so whether the increased surface area is via mesh, sponge, foam, grid, serrated fins or just densely-packed flat vanes, it doesn't matter as long as total surface area is equal and the airflow is turbulent.

This mesh looks to have all the hallmark problems of large heatsinks without heatpipes - it's too large to work beyond a certain wattage with a certain temperature delta, because the fixed value for copper's thermal conductivity (~400W/mK) uses a unit that includes a dimension; the resulting temperature differential over great distances would mean that the business end in contact with the processor would need to be at throttling temperatures before the heatsinksponge's outer edges became relevant to the heat transfer process. This is why you don't see giant one-piece heatsinks; You see smaller heatsinks that are connected to a heat source by a much faster heat transport device - commonly a vapor-change mechanism (heatpipe). In the instance of giant fins, they have 8 or 12 heatpipes meaning that the actual heatsink per heatpipe source is 1/8th or 1/12 the size of the whole unit.
<insert large, flashing, epileptic-fit-inducing signature (based on the latest internet-meme) here>
Chrispy_
Minister of Gerbil Affairs
Gold subscriber
 
 
Posts: 2094
Joined: Fri Apr 09, 2004 3:49 pm


Return to Overclocking, Tweaking, & Cooling

Who is online

Users browsing this forum: No registered users and 4 guests