[time-nuts] Exact Rubidium frequency

Richard (Rick) Karlquist richard at karlquist.com
Sat Jun 30 23:11:31 EDT 2007


In a cesium beam standard, the only systematic
offset from the unperturbed state is the C-field,
and its offset can be determined and removed very
accurately by Zeeman techniques.  Many other possible
offsets are servoed to virtually zero (less than
10^-14).  The servoing is all done using cesium
physics, without reference to any other frequency
source.

Rubidium and hydrogen use gas cells.  These have
frequencies that depend on many things, especially
pressure and temperature of the gas.  In Rb standards,
there are also optical offsets.  It is also possible
to make a cesium gas cell standard.  It will behave
like a Rb standard in terms of being a secondary
standard.

When we used to build 5071A's, we simply built them
and they came up on frequency.  There is no "setting"
procedure.  We could have built them in a "clean room"
containing no clocks, GPS, or other cesium standards,
etc. without difficulty.  The very first one we ever
built was taken to NIST to be checked and was within
a couple of parts in 10^13.  

Rb frequency standards will come up on different frequencies
depending on differences in the gas mixture and pressure
in the various cells, etc.  You always have to have a 
cesium standard to set them to frequency.

One possible point of confusion is that the 5071A allows
the user to introduce a controllable offset as a convenience.
This is done with a synthesizer that was available anyway.
This has nothing to do with "setting" the 5071A to the
correct frequency.

I hope this clears up any confusion.

Rick Karlquist N6RK
Member 5071A design team
and 10816A Rubidium design team



More information about the time-nuts mailing list