Re: OT: Marcury has a layer of diamond 10 miles thick, NASA spacecraft finds

On Sat, 17 Aug 2024 14:21:34 +0200, albert@spenarnc.xs4all.nl wrote:

In article <daunajltg82fsitpn6uje2p75agjhur5j5@4ax.com>,
John Larkin  <jjlarkin@highlandtechnology.com> wrote:

On Thu, 01 Aug 2024 13:33:11 -0700, boB <boB@K7IQ.com> wrote:
>

On Wed, 31 Jul 2024 15:00:49 -0700, John Larkin
<jjlarkin@highlandtechnology.com> wrote:
>

On Wed, 31 Jul 2024 13:04:54 -0700, boB <boB@K7IQ.com> wrote:
>

On Wed, 31 Jul 2024 12:38:15 +1000, Bill Sloman <bill.sloman@ieee.org>
wrote:
>

On 31/07/2024 3:46 am, john larkin wrote:

On Tue, 30 Jul 2024 10:46:45 GMT, Jan Panteltje <alien@comet.invalid>
wrote:
>

Mercury has a layer of diamond 10 miles thick, NASA spacecraft finds
https://www.space.com/mercury-diamond-layer-10-miles-thick-nasa-messenger#main
>
Now there is an incentive to go!

>
Diamond, like gold, is valuable because it's rare. Accessing cubic
miles of diamonds would trash its value.

>
Wrong. Diamond is light, hard, strong and has very high thermal
conductivity. You can buy synthetic diamond heat-sinks right now.
>
https://www.msesupplies.com/products/diamond-heat-sink-thermal-conductivity-1500w-m-k?variant=39601902846010
>
--
Bill Sloman, Sydney

>
>
Would love to have some diamond (man-made is fine)  heat sink
insulator material !
>
boB

>
>
If you mean an insulator between a power transistor and an aluminum or
copper heat sink, a thin aluminum nitride insulator would be almost as
good as diamond. Or hard anodize.
>
Really using the heat sink would require lateral heat spreading,
namely a big thick slab of diamond.

>
John, this would be used as a lower thermal resistance insulated
interface from case to heat sink.  I wonder if you still need to use
the white bird shit on that interface ?
>
boB
>

>
Silicone grease makes a huge difference in thermal conductivity,
especially if the mating parts are not optically flat.
>
A typical TO-220 mosfet tab can be bowed by a couple of mils, and heat
sinks are usually extruded, not very flat. Air gaps are rotten heat
conductors.
>
Given some grease, keep the insulator thin.
>
Polycrystalline diamond is crazy expensive.
>
One trick is to bolt a transistor directly to a copper block, to
spread the heat footprint, and insulate that from the main heat sink.
>
Or just parallel a few fets to spread the heat around.

>
Putting my machinist cap one.
It is not particularly hard to lap the two surfaces, that virtually
no air gap result. This is used to construct parallel gauges, they
are so flat that the cohesion holds it together.
>
Whether it holds up with thermal stress,I don't know.
>
Groetjes Albert

Lapping a fet and its heat sink will seriously reduce theta, but it's
a lot of work. Grease is easier. But the thermal conductivity of
aluminum is still a bottleneck.

Heat sinks and machining are both expensive, and mosfets are cheap. It
often makes sense to use several fets, distributed about a heat sink,
to avoid hot spots.

I assume that heat sink vendors test theta by applying a big, uniform
heat source. Real parts tend to be small spots, and local heating
makes theta go way up.

We have several new products that use copper CPU coolers on the board,
to cool mosfets. They are magic.

https://highlandtechnology.com/Product/P945