[time-nuts] 5061A and Zeeman freq question

Tom Van Baak tvb at LeapSecond.com
Sat Jul 19 11:09:02 EDT 2008

> My understanding is that the precise cesium hyperfine transition energy, 
> and hence microwave frequency, is dependent upon the applied C-field, 
> and the field setting to select a particular transition is the same for 
> all clocks (which is why these are primary standards). The C-field is 
> set by applying an LF signal at the correct frequency to cause Zeeman 
> splitting, and optimizing the C-field to peak the splitting.

It's the same for all cesium tubes of the same design but over
the decades a number of different tubes have been designed.

42.82 kHz (older model 5061)
70.40 kHz (model 5062c)
53.53 kHz (newer 5061A/B and 5071A)

See this 2005 time-nuts posting:

The Zeeman shift would also be different among the one of a
kind cesium standards that national laboratories build. But the
key is that based on the beam design the zeeman shift can be
calculated and thus compensated for.

What makes a standard primary is if all the frequency shifts
can be calculated (not measured) as well as the magnitude
of all error sources. For a great example of this read:
including table 3 and table 4.

> So, shouldn't the Zeeman frequency always be 42.82 Khz, which is what 
> I've seen in every reference? But, my 5061A with a high-performance tube 
> has a sticker that says 'Zeeman freq 53.53 Khz'. I can indeed get Zeeman 
> splitting at that frequency with a C-field setting of 3.45, but is this 
> correct? If so, why? (BTW - I don't seem to be able to get any splitting 
> at 42.82 Khz)
> Confused,
> Bill

There's also a nice description here:

Homemade plots of the 7 cesium hyperfine peaks here:


More information about the time-nuts mailing list