[time-nuts] Exact Rubidium frequency

[Magnus Danielson]

From: Peter Vince <pvince at theiet.org>
Subject: [time-nuts] Exact Rubidium frequency
Date: Sun, 1 Jul 2007 00:11:34 +0100
Message-ID: <v04210100c2ac925eb18d@[10.0.0.3]>

> Everyone on the 'net agrees on the frequency of the Caesium 
> oscillation, but the precise frequency quoted for Rubidium varies by 
> a few Hz :-(  Now I realise that you can make it anything you like 
> (within reason) by altering the magnetic field, and perhaps different 
> sources quote different figures which are more or less easy to 
> synthesise?
> 
> On the (USA) NIST web-page 
> (http://tf.nist.gov/general/enc-re.htm#rubidiumoscillator) they quote 
> 6,834,682,608 Hz, but the (German) PTB (on 
> http://www.ptb.de/en/org/4/44/441/info2_e.htm) give 6,384,682,612.8 
> Hz, and the (British) NPL quotes 6,834,682,610.904 324 Hz in a 
> powerpoint presentation 
> http://www.npl.co.uk/time/club/meeting1/secondaryrepsec.pdf 
> (admittedly for a Rubidium fountain).
> 
> I have also seen the frequency of a Hydrogen Maser given as both 
> 1,420,405,751 Hz and ...752 Hz.  I would be very surprised if the 
> frequency was an exact whole number of Hertz different from Caesium, 
> so perhaps this is just rounded for convenience as again it can be 
> steered to anywhere you like?
> 
> Would anyone care to comment on this differences please?

Please check this page out:
http://rubidium.dyndns.org/~magnus/physics/hyperfine/

6.834682610904324(4)(7) GHz is the best number I've got for Rubidium and
1.4204057517662(30) GHz is the best number I've got for Hydrogen.

I have provided references for those numbers.

The MathML stuff look like crap for me as I checked it but it used to be
alright.

Any normal Rubidium and Hydrogen clock is being pulled from actual numbers not
only by magnetic field but also wallshift and cavity shift for hydrogens and
in the Rubidium case by additional buffer gas pulling. The Hydrogen is being
examined by various sizes of bulbes in the same cavity as the wall shift
changes in a predictable form and this way its effect can be cancelled
numerically in a fashion which doesn't lend itself to a normal clock.

The very good Rubidium number comes of a Rubidium fontain in Paris. Rubidium is
a better atom for fontain clocks than Caesium. Far better actually. Cools down
better, so there's why.

Cheers,
Magnus