[time-nuts] Question about SA.33 Rb clock
magnus at rubidium.dyndns.org
Thu Oct 12 15:02:22 EDT 2017
On 10/12/2017 06:06 PM, Tom Van Baak wrote:
> Your use of the phrase "real cesium" may be the source of your confusion. The SA.3x uses rubidium and the SA.4x uses cesium. They are all real atoms. These modern MAC / CSAC atomic standards compete with high-end DOCXO quartz oscillators with respect to factors like temperature, stability, and drift. They do not compete with traditional laboratory rubidium or cesium standards.
> You may be thinking that because some CPT clocks use cesium instead of rubidium that they are special or more accurate, but this is not the case. None of these compact low-power laser / VCSEL / CPT -based frequency standards are primary standards.
To follow up on that, even "rubidium" and "cesium" is a bit of
misnomers, since you should really say "rubidium optically pumped gas
cell clock" and "cesium atomic beam clock", you can build gas cells with
cesiums and you can build atomic beams with rubidium. Traditionally
rubidium have been very easy to build optically pumped clocks with, but
today you can also do that with cesium, which is what the VCSEL based
CPT clocks really is.
All gas-cells will have the issues of frequency pulling, regardless of
rubidium, cesium or other atom being used.
The different technologies have different benefits and for some
implementation aspects have made different atoms being more preferred
As we moved to fountains, rubidium turned out to be a better choice.
More information about the time-nuts