Many years ago I visited the Philips (as it was then) factory in
Nijmegen. Their production test fixtures for parts like BLF278 and
BLF378 had cutouts under the transistor flange so that cooling
water was fed upwards in at one side, ran across the flange, and
exited down and out at the other side. There was an O ring seal to
prevent leakage and a valve opened when power was removed so the
water drained away a bit when the transistors were swapped. A
pressure clamp held the transistor in place and pushed the leads
onto contact pads.
If you're using water, I reckon that's the way to go. Anything
that avoids a thermal interface between the flange and other metal
has to be good. Most of the 1kW transistors I've handled have
needed attention to get the flanges flat enough for proper heat
transfer. I like the 300W pressure clamp parts.
Steve
It sounds great in theory and maybe in practice too. How do you mount the
transistors to the chassis without a through hole into the water jacket?
If you look at the NXP BFL578XR demo board, they are using a 3/8" solid copper
heat spreader which then has a water inlet and outlet for cooling the heat
spreader. Similar idea except the transistor mounting does not penetrate the
water jacket and there is no chance it will leak due to thermal expansion and
contraction.
A second note you might find interesting is that many of the parts are supplied
as solder mount and are mounted directly to the copper heat spreader via a
solder mask. That may be one way to mount to the chassis bottom, but I'm not
so sure it's easy to solder to Aluminum or what two very dissimilar metals will
do when bonded.
More food for thought.
Paul, kg7hf
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|