On Tue, 2007-03-06 at 19:50 -0500, Gary Hoffman wrote:
> It is interesting that most of us don't know how important the heat sink
> goop is, and how much of the heat transfer takes place through it.
>
> The two metal surfaces (heat sink and transistor) even if machined quite
> flat and smooth actually touch microscopically at a very few tiny points.
> That is the limit of the metal to metal transfer. Even if you work really
> hard at it, that is the best you can do.
>
> The rest of the space is filled with air, and air is an excellent insulator.
>
> To fix this problem, one approach is to use conventional heat sink compound.
> These common compounds use a passivated aluminum oxide extremely fine powder
> suspended in a petrochemical. This compound fills the air gaps and the
> aluminum oxide conducts the heat far better than the air would have.
>
> In extreme applications, a better heat sink compound uses extremely fine
> silver in the place of the aluminum oxide, for many times better results,
> but at the expense of having the compound be electrically conductive which
> can cause many issues.
>
> So, generally we use the aluminum compound and it works really well.
>
> To go beyond that, you have to turn to the heat sink itself. Make it
> bigger. Make it Copper. Give it more and larger fins. Blow air on it.
> Or - if you really want to do it right, use a TEC for active cooling and
> make that heat sink really really cold. Just short of the dew point of
> course. And much more practical than liquid cooling.
So today in Iowa that heat sink could be cooled to 15 F but on a warm
summer day in Houston, the dew point might be 95 F so you would have to
include a dew point sensor to pick the cooling point, or a moisture
sensor on the heat sink to detect the dew...
>
> TECs really rock when used properly in this manner.
>
> 73 de Gary, AA2IZ
>
--
73, Jerry, K0CQ,
All content copyright Dr. Gerald N. Johnson, electrical engineer
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|