[time-nuts] CS reservoir depletion

Magnus Danielson magnus at rubidium.dyndns.org
Thu Jan 13 20:33:14 UTC 2011

On 13/01/11 17:24, J. Forster wrote:
> As I said last go-round, I think the chances of rebuilding an HP Cs tube
> are slim to none.
> BUT, if you were really dedicated, I think you could build up the physics
> package from pretty much standard stuff, like the Kimball Physics gun
> parts and Conflat SS vacuum hardware, etc.  Check Duniway.
> I'd start by reviewing the old, academic papers. Much of the electronics
> can be bought on eBay as instruments now.
> Yes, it'd be refrigerator size, but it is do-able, IMO.

Many of the pieces of a physical package can either be found off the 
shelf (ionizer/masspectrometer) or at least be hammered up fairly easily 
(microwave link) even if performance is far optimum. The cesium oven 
should also be fairly buildable. The C-field coil could certainly be 
built with some care. My-metal shields may take some effort. Outer 
shield is a challenge for vacuum performance and there is such a thing 
as high-vacuum nuts. Similarly the ionpump is off the shelf. The piece 
that I would not expect to be easy to buy or build is the A and B 
magnets which deflect wrong state atoms. Also, handle Cesium and 
breaking the Cesium ampule may be a bit of a challenge to get right.

There is certainly be a bit of challenges to get something which works. 
Being able to get good performance would take many many years of 
experimentation and learning. The HP/Symmetricom efforts spans from the 
60thies. Many steps of improvement, of which only some is found in 
papers and patents. But they have left such traces too. There are many 
practical solutions which can be learned from the archives.

The electronics could either be to reuse a HP5061 or whateer is 
available, but most of it should not be all that hard depending on 

It would be a hell of a project. Just like building a H-maser.
In comparision a Rubidium is much easier.


More information about the time-nuts mailing list