[time-nuts] Is 5061A by itself a primary reference? Was: Modern Rb atomic reference vs classic Cs

[Anders Wallin]

>> At one time, only cesium standards were considered truly primary because
> of
> >> the definition of the second.  However, the quantum mechanical
> constants of
> >> other atoms such as Rb have been measured to much more accuracy than
> the 5071
> >> so that Rb standards can be considered traceable to Cs if they are
> otherwise
> >> of primary architecture.  The key idea is that all Rb atoms are
> absolutely
> >> identical.  Rb gas cells are of course never primary.
>

FWIW BIPM just recently started publishing a graph like this that shows
monthly TAI-contributions from what BIPM considers primary frequency
standards (Cs beam and fountains) as well as 'SRS'-clocks (secondary
representation of the second):
https://webtai.bipm.org/database/show_psfs.html
for these contributions they require a published uncertainty-budget where
you calculate 'on paper' all the known shifts and show how you have
controlled them to give a total uncertainty of the clock.


> >
> > Can the physics-nuts calculate the Rb frequency relative to Cs?
>
> I used to wonder about why there isn't a formula for calculating
> the transition frequencies for an atom, at least relative to
> another variety, but I know almost nothing about quantum mechanics.
> Evidently, if it were possible, we would have heard about it by now.
>

I think there might be around 12 digits (?) or so of agreement between
theory and experiment for Hydrogen (and anti-Hydrogen!?) - but for the
complex clock atoms I think there's no theory that predicts the transition
frequencies very well.


>