[Elecraft] Auxiliary Fan

[Ron D'Eau Claire]

I decided to try Eric's suggestion to add a muffin fan on top of the K2
to reduce the heating and subsequent frequency drift.

My junque box turned up a 3- in (8 cm) Radio Shack "muffin fan" with a
d-c brushless motor rated for 12 volts at 0.16 amps (Catalogue No.
273-243B). I mounted four small rubber feet to the corners and placed it
over the right side of the heat sink, about 1/3 of the way "back" from
the front panel as Eric suggested. That put it directly over where the
power transistors are mounted. 

Not wanting to add another whining, noisy thing to the shack (in
addition to the little fan on the back and, of course,  the operator
hisself) I put a 100 ohm resistor in series with the fan that cut its
speed down enough that it is only a very soft "whisper" even though it
is sitting directly in front of me on the rig. Putting my hand above it,
I can just feel a soft breath of air coming up from the fan (Eric
explicitly recommended mounting it to blow upwards, away from the heat
sink). 

I re-ran my 15 meter test sending CW at 100 watts (into a dummy load, of
course). I have a message buffer loaded with V's that recycles with a
4-second break every 30 seconds or so. 

Without the fan I had measured approximately 240 Hz of drift after a
cycle of 10 minutes of CW, then a five minute listening period, followed
by another 10 minutes of CW. 

With the fan, the drift is cut to about HALF... down close to a total of
100 Hz. 

Keep in mind that is in the worst case situation on a higher band where
the drift characteristics of the BFO and PLL Reference Oscillator add
instead of cancel each other out. On the lower bands the drift is much,
much smaller. Still, 100 Hz drift for CW is not bad at all. I can
understand that many ops - especially those running digital modes a lot
- would want it much less. 

In any case I can say that an auxiliary fan does a lot to help. I might
try running mine a little faster to find the best tradeoff between the
fan noise and the cooling efficiency. 

It makes me wonder what a larger heat sink would do. Who said that the
K2 couldn't be a little taller anyway? 

And would the little fan still be needed?

And I keep thinking about a true temperature-compensating circuit for
those oscillators. Something that senses the temperature and applies a
correction voltage to keep them on frequency over a truly wide
temperature range. 

Ron AC7AC
K2 # 1289