aluminum heatsink anodizing
in [Design]
|
Prev: Automotive Protection Circuit
Next: uav matlab model
From: Rich Grise on 22 Dec 2008 14:27 On Fri, 19 Dec 2008 20:04:23 -0800, John Larkin wrote: > > I drives NMR and MRI gradients,... You does? ;-) Cheers! Rich
From: Rich Grise on 22 Dec 2008 14:29 On Sat, 20 Dec 2008 08:58:27 +0000, Eeyore wrote: > John Larkin wrote: >> Eeyore <rabbitsfriendsandrelations(a)hotmail.com> wrote: >> >John Larkin wrote: .... >> >> About 300 watts peak each. >> > >> >Cheat ! Average figure please ? Which you damn well know is the >> >important one unless your pulses are several seconds long. >> >> I drives NMR and MRI gradients, as it was designed to do. > > So you won't answer the question because it wouls show you up as a > charlatan. > Hey, John might be a lot of things, but a charlatan he's not. And anyway, why are you making a religious war out of this? Who cares what he uses on his heatsinks? It's like the Pascal crowd and the PL-1 crowd digging in for the duration. Cheers! Rich
From: John Larkin on 22 Dec 2008 15:48 On Mon, 22 Dec 2008 19:22:51 GMT, Rich Grise <rich(a)example.net> wrote: >On Fri, 19 Dec 2008 16:38:27 -0800, John Larkin wrote: >> >> AlN is in the ballpark of 150 w/m-K, close to aluminum itself and not a >> lot worse than BeO. In modest quantity, custom AlN insulators aren't >> terribly expensive, close to the high-end sil-pads as I recall, but with >> 40x or so the thermal conductivity. > >Is it possible to "nitridize" the aluminum, laying down a ~1 mil film of >AlN, a la anodizing? > >Thanks, >Rich That would sure be interesting... http://en.wikipedia.org/wiki/Aluminum_nitride#Manufacture Another trick is to epoxy the semi directly to a bare aluminum heat sink, using a thermally conductive epoxy lightly seeded with glass spacer beads, like the Cataphote things. John
From: Archimedes' Lever on 22 Dec 2008 19:57 On Mon, 22 Dec 2008 12:48:40 -0800, John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: >On Mon, 22 Dec 2008 19:22:51 GMT, Rich Grise <rich(a)example.net> wrote: > >>On Fri, 19 Dec 2008 16:38:27 -0800, John Larkin wrote: >>> >>> AlN is in the ballpark of 150 w/m-K, close to aluminum itself and not a >>> lot worse than BeO. In modest quantity, custom AlN insulators aren't >>> terribly expensive, close to the high-end sil-pads as I recall, but with >>> 40x or so the thermal conductivity. >> >>Is it possible to "nitridize" the aluminum, laying down a ~1 mil film of >>AlN, a la anodizing? >> >>Thanks, >>Rich > >That would sure be interesting... > >http://en.wikipedia.org/wiki/Aluminum_nitride#Manufacture > > >Another trick is to epoxy the semi directly to a bare aluminum heat >sink, using a thermally conductive epoxy lightly seeded with glass >spacer beads, like the Cataphote things. > >John Or use the HARD surfaced Al and use CONDUCTIVE, silver filled epoxy (VERY HIGH THERMAL CONDUCTIVITY) bond it directly to the sink. Even closer, and the gap filler media is high conductivity metal/epoxy matrix. It sounds like super thin, hard anodized on at least on side, Al 'pads' could be used and would perform better than a sil pad.
From: John Larkin on 22 Dec 2008 20:25
On Mon, 22 Dec 2008 16:57:05 -0800, Archimedes' Lever <OneBigLever(a)InfiniteSeries.Org> wrote: > It sounds like super thin, hard anodized on at least on side, Al 'pads' >could be used and would perform better than a sil pad. Those are commercially available, thin and anodized both sides. I'd use grease, too. Nothing I know of is worse than a sil-pad, except the Thermalloy clear amber things, which I think are kapton. John |