[time-nuts] Antique Rb Standard - Thanks, Pictures, Parts Request, Question

[Magnus Danielson]

On 04/30/2012 10:32 AM, Attila Kinali wrote:
> On Sun, 29 Apr 2012 13:19:08 -0700
> ed breya<eb at telight.com>  wrote:
>
>> If the Rb cell "drifted" enough (maybe if it was filled with water)
>> to de-tune the mechanical cavity resonator, it wouldn't work at all.
>> Only a small amount of the RF power in the cavity is coupled into the
>> Rb gas. The bandwidth of the Rb resonance is a fraction of a Hz (Q in
>> the millions), while the bandwidth of the cavity is probably 20-50
>> MHz, (Q may be up to 100).
>
> Not quite true. While the theoretical line width of Rb is very small,
> the line width of a Rb gas cell is in the range of 100Hz to 100kHz,
> depending on the size of the cell, wall coatings, the buffer gas type,
> the pressure and temperature.

If there would be no buffer gas, line-width of 100 kHz would be normal, 
relating to the time it takes for the atoms to go across the chamber in 
ballistic dispersion. When an atom hits the wall, it's state is re-set, 
but when it bumps into some gases, it's hyperfine state isn't shifted 
all that much, but it's ballistic path is interrupted and we get to see 
it much longer and hence the Q value is decreased to around 100 Hz.

Temperature comes in on many subtle effects. Align temperatures right 
and you can approximate TC balance.

Q-width is very clearly tied to observation time, and has been a driving 
factor for gas cells, active masers, beam devices, fountains and ion traps.

The pressure in the absorption cell also makes it sensitive to 
barometric pressure.

Cheers,
Magnus