[time-freq] Rubidium standards

[John Ackermann N8UR]

Randy said the following on 02/08/2008 12:15 PM:
> I see advertised once in a while Rubidium frequency standards that include
> the following statement:
> 
> "The Rubidium standard is carefully calibrated to a GPS Disciplined
> Oscillator."
> 
> 
> I thought those standards were based on the atomic physics of Rubidium and
> that was what determined the frequency.  Why do they need to be "calibrated"
> at all?  This also suggests that a GPS-disciplined oscillator is better than
> the Rubidium standard.  It that true?  Are the HP Z3801 oscillators, for
> example, ordinarily more accurate than a good Rubidium standard?

Hi Randy --

An Rb standard is based on atomic physics, but it's not intrinsically
accurate because there are various effects that can move its frequency.
 So, while a Rube is much more stable than a crystal, it still needs to
be set to the correct frequency and checked every now and then.

Even a Cesium standard, which is by definition "correct" has a certain
amount of uncertainty and will be in error by some amount; if you have a
good source (like a GPS or GPSDO), you can fine tune it to more closely
match the national standard.

A good GPSDO is a very good all purpose standard, with the short term
stability of a crystal and the long term accuracy of GPS.  Its weakness
is that if it loses its GPS signal, it will quickly revert to normal
crystal performance.  An Rb or Cs doesn't need an external reference to
retain its accuracy, and that self-contained nature can have some
advantages.

73,
John