Paul Whatton wrote:
>I remember a conversation with Ian GM3SEK about heatsinks when I was
>building my GI7b & GS31 2m amplifiers (apologies if I misquote you Ian,
>you'll probably pop up and say something). I also suggested heat sink
>compound but Ian pointed out that the tube-heatsink interface
>temperature was too high.
>
That may not be true of modern high-performance heat sink compounds...
but the second point below is the more important one:
>He did suggest that the machining on the tube anode & heatsink were not
>great and that it might be a good idea to stick them in a lathe and
>skim them to improve heat transfer. I didn't actually do this but it
>seems a good idea to me.
>
The poor surface finish of the GS35/GS31/GI7 metalwork is sure to be the
greatest barrier against efficient heat transfer between the tube and
its anode cooler. As we know from high-power RF transistors, the aim is
to make the contacting surfaces so smooth that they *almost* don't need
heatsink paste at all. Then apply the paste as thinly as possible (a CPU
cooler is a very good example of how *little* heat sink paste is
actually needed).
The sand-cast anode coolers on the GS31 and GI7 can't be very efficient
either, because of the small number of fins. One option might be to
replace these with US-style coolers salvaged from dead 8877s and
3CX800s. It should be possible to shrink fit a solid copper plug inside
of the cooler, with a fixing hole through the centre and a turned flange
to mate onto the Russian tube. (Just an idea; I haven't tried it.)
--
73 from Ian GM3SEK
http://www.ifwtech.co.uk/g3sek
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|