[time-nuts] Muffin Fan and Rb Oscillator

Magnus Danielson magnus at rubidium.dyndns.org
Sun Jul 17 20:54:09 UTC 2011


On 17/07/11 22:07, Mark Spencer wrote:
> I've done something similar to this.  In my case the Rb out performs the 100 second adev and re trace specs from the manufacturer with the fan running.  I can't measure phase noise and I only use the rb to occasionally use as a reference for checking the frequency of an oxco over a time span of an hour or so.
>
> I did notice that the current drain of the rb is slightly higher with the fan running  and probably more stable than before.  I suspect the heater in the rb  is working a bit harder than before and I may be over cooling the unit.  This might be something to look at if you cool your rb with a fan.

The thing is... if you have a fan under control, you can allow the 
cooling plate become hotter as you have higher capability of cooling 
than just a passive radiator.

The heaters of a rubidium is there to achieve stable temperatures of 
about 70 and 130 C but then need cooling to maintain stability.. but if 
you modulate the cooling with a time-constant longer than the control 
loop of the heater, then you can get good stability with less power 
consumed.

Cheers,
Magnus




More information about the Time-nuts_lists.febo.com mailing list