I always thought ceramics were great at insulating, not conducting heat? (think like the space shuttle tiles)

There are companies out there professing that ceramic heat sinks are better at dissipating heat than aluminum ones are. Supposedly, you can replace a 30mm x 30mm x 6mm thick aluminum heat sink with a 30mm x 30mm x 1mm thick plate of this ceramic and get better heat dissipation. (!!?)



The only thing I can figure is maybe it's better at radiating heat vs conducting/convection... or maybe it has super-crazy surface area? Any thoughts?

I was reading wrong >_<

Brad

[This message has been edited by twofatguys (edited 11-04-2010).]

Weird - I never thought ceramics were very good at that - take two ice cubes, put one on some aluminum, put the other on a ceramic tile (non-glazed, I suppose)

Maybe the ceramic loses heat quicker? (less heat retention?)

Weird.

This type of ceramic pulls heat away faster then aluminum & slightly slower then copper.




It's advantage over copper is less cost, less weight, no degredation, and the big advantage that curcuts can be printed directly on the ceramic.

[This message has been edited by Boondawg (edited 11-04-2010).]

As a long time shade-tree ceramicist, none of this surprises me. Ceramic tiles will absorb heat of whatever is set on them.. I put some in my oven to even out that cantankerous heat, it really helps even your baking temperatures.

Ceramic tiles are used on top if electric in floor heat often. This says something too.

quote Originally posted by 2.5: Ceramic tiles are used on top if electric in floor heat often. This says something too.

And in walk in fireplaces to help radiate the heat into the room.

quote Originally posted by 2.5: Ceramic tiles are used on top if electric in floor heat often. This says something too.

They also use ceramic plates to radiate heat in space style heaters. One of the best heaters I ever used was about 4 foot long with a ceramic core to radiate the heat.

About a month and a half ago I sent my Father a 30,000 btu ceramic heater that uses gas to heat the ceramic plates. He was saying once the room warms up and it shuts the gas off, the pilot light will keep the ceramic plates hot enough to keep it from firing up again for a amaizingly long time.

quote Originally posted by ryan.hess: The only thing I can figure is maybe it's better at radiating heat vs conducting/convection... or maybe it has super-crazy surface area? Any thoughts?

I am puzzled, too. After looking at their web site, it appears that the primary advantage claimed is that ceramic allows better thermal coupling between a semiconductor (e.g. LED) die and the heat sink. Probably not very much to be gained there, but the prospect of applying a metallization layer directly on the ceramic seems to hold some promise.

Last night I had dinner with an engineer friend who said that they are evaluating a diamond heat spreader for a project, because of its low thermal impedance and high resistivity.

Maybe if people would think of ceramics as a very good heat sponge, it would help.
It soaks up heat quickly.

quote Originally posted by Boondawg: Maybe if people would think of ceramics as a very good heat sponge, it would help. It soaks up heat quickly.

Yup, that's basically it. In physics, it's called "thermal conductivity".

There are many different types of ceramics. While most ceramics don't conduct heat very well, there are a few that do.

quote Originally posted by Blacktree: Yup, that's basically it. In physics, it's called "thermal conductivity". There are many different types of ceramics. While most ceramics don't conduct heat very well, there are a few that do.

The problem with this ceramic in particular is it's thermal conductivity is an order of magnitude less than regular aluminum! Based on that, it shouldn't be able to take much heat at all without the electronic part bursting into flames.

Comments from a "knowledgeable" friend:

"The thermal conductivity is listed as 24 W/mK which is about 8 times the very low thermal conductivity of the low temperature cofired ceramic that is commonly used for hybrid microcircuits. Our diamond like coating process ( I have been working on this project with ... for a couple of years and now with some profs at ... Univ) is looking to have a conductivity of at least 10 times that, and we think we may eventually be able to get to 40 times that or even more.

"As to being "better than aluminum" they apparently are counting on some sort of better contact or larger area because aluminum has a thermal conductivity of 210 W/mK, almost 10 times what they claim and about at the minimum level that we think our ... may start to become a bit interesting. 250 W/mK is about our minimum threshold for 'success' and we are not there yet. We think that 1000 or more is possible with imperfect man made diamond-like carbon. Natural diamond has a thermal conductivity of 2500 W/m K."

[This message has been edited by Marvin McInnis (edited 11-08-2010).]

I didn't know it was cost effective... but I guess diamond is an option...

http://www.sp3inc.com/spreader.htm

I ordered some samples of the ceramic heatsink to put it to the test. Should be interesting...

To paraphrase Mark Twain, diamond is just coal that went to college.

Sometimes the unique properties of a material justify its cost. I once talked with a manager who described the procurement challenges of getting approval for a diamond window on a spacecraft intended to penetrate the atmosphere of Venus. In that case there was simply no other material that could perform adequately and survive in the Venusian atmospheric environment as long as diamond.

[This message has been edited by Marvin McInnis (edited 11-08-2010).]

quote Originally posted by ryan.hess: I didn't know it was cost effective... but I guess diamond is an option... http://www.sp3inc.com/spreader.htm I ordered some samples of the ceramic heatsink to put it to the test. Should be interesting...

Would that diamond heat spreader take place of the heatsink or a replacement for thermal paste, or as a supplement for thermal paste, or...???

quote Originally posted by Blacktree: Yup, that's basically it. In physics, it's called "thermal conductivity". There are many different types of ceramics. While most ceramics don't conduct heat very well, there are a few that do.

Blacktree about covered it. There are many types of ceramics. Just because its defined as a "ceramic" doesn't maen it HAS to have poor thermal conductivity. Just that most of the ones we run into every day, happen to be really good insulators.

quote Originally posted by Rallaster: Would that diamond heat spreader take place of the heatsink or a replacement for thermal paste, or as a supplement for thermal paste, or...???

Thermal paste provides a thermal interface (i.e. a bridge or coupling) between two rigid substrates.

The function of a diamond thermal spreader would be as an intermediate layer to transport heat from a very small point source into a heat sink with a much larger area. The whole point is to move heat away from a point source as quickly and efficiently as possible. High-power LEDs for general lighting have elevated this from being mainly a theoretical problem to being a significant practical problem.

[This message has been edited by Marvin McInnis (edited 11-09-2010).]