I wonder about some other practical considerations for cooling a
preamp. What I am thinking about is the effect of cooling on the other
parts of the amplifier besides the active device. I would think that
tuned circuits, resistor, and capacitor values would all change. This
could have the effect of offsetting or maybe more than offsetting any
benefits of the theoretical noise figure improvements.
It would be interesting if someone who had access to one of those
expensive HP noise figure meters could set up a typical preamp (for 1420
MHz, for example) with a proper temperature sensor and then continuously
measure its gain and noise figure as it was cooled with dry ice or
liquid nitrogen. A plot of the amplifier's temperature versus noise
figure (without any compensating adjustments) would be very
interesting. Also, after the amplifier was cooled down as low as
possible, what improvements, if any, could be seen by making whatever
adjustments were possible.