[time-nuts] Rb cooling

[Charles P. Steinmetz]

Poul-Henning wrote:

>Cooling Rb's should happen only through the base-plate.

>For frequency stability, you should strive to have a constant
>baseplate temperature.

I mounted an FRS-C in an unvented cast aluminum box by its baseplate 
(after cleaning up the mating surface in the box on a mill to get 
good contact), then mounted the cast box on shock mounts in an 
enclosure along with a power supply, batteries, dist amp, dividers, 
monitoring, etc.  The outer box has a small, variable-speed fan that 
is controlled by the temperature of the inner box in the middle of 
the FRS-C baseplate contact area.  The control loop keeps that 
temperature at 45C +/- about 0.1C.  It seems to work very nicely.  I 
can't detect any effect from the fan's vibration.

Best regards,

Charles