[time-nuts] HP5065B !!!

Magnus Danielson magnus at rubidium.dyndns.org
Sun Apr 28 12:36:16 UTC 2013


On 04/28/2013 02:01 PM, Attila Kinali wrote:
> On Sun, 28 Apr 2013 17:50:06 +1200
> Bruce Griffiths<bruce.griffiths at xtra.co.nz>  wrote:
>
>> Has anyone considered a laser pumped variant like:
>> http://tf.nist.gov/general/pdf/1009.pdf
>>
>> Apart from the ECDL laser (can be assembled using readily availalble
>> parts) it looks fairly straightforward.
>
> Considered, yes, tried, no. From what i've read sofar, this system
> has the problem of locking the laser wavelength onto the right
> absorbtion line. IIRC the linewidth of Rb in a vapor cell is a couple
> of 10kHz to a few 100kHz. Using an ECDL you get a laser linewidth of
> less than 1MHz easily, usally in the range of a few 100kHz and less.
> Ie. the laser would need to be kept on the absorption line with a stability
> of a couple of 100kHz at most. Using just a laser diode (without the
> external cavity) with its linewidth of>100MHz makes it actually a
> little bit easier to handle.
>
> But getting the laser to the right Rb absorption line and detecting whether
> it's off is still not solved. Most of the papers that i've read that do
> something similar use an additional vapor cell to determine the correct
> position of the laser.
>
> I'm quite sure that it could be done with a single vapor cell using
> some sophisticated control loop that steers both the 6.9Ghz signal
> and the laser wavelength, but i doubt it's easy. But then, i didn't
> have an in depth look at this.

When being at NIST last summer, they had us tour the facility. One of 
them was not as much a show but a physical lab. We got to trim up and 
test laser-cooling of... rubidium vapour. It was about 100000 of 
hardware on the table if I recall correctly, but nothing really advanced 
really. A secondary rubidium cell was used as a detector for the right 
rubidium features. Using that and tune two lasers with lock in 
amplifiers also having sweep features, it became fairly simple to do. 
More than half the table of equipment does not apply to a rubidium clock 
anyway, so the set up could be rationalized down considerably.

We succeeded to laser-cool both the Rb-87 and Rb-85 :)

Turned out that once you got about the right wavelengths and fairly 
balanced polarization, much other things like the magnetic 
field-strength was not very critical at all to achieve and maintain the 
effect.

It was a fun lab to do, and I wish to learn more about steering laser 
wavelengths etc. I do far to little optical stuff.

It is also interesting that our optically pumped rubidiums is in fact 
more closely related to modern atomic standards than traditional caesium 
beams. The fountains is really a cross of them both to some degree.

Cheers,
Magnus



More information about the Time-nuts_lists.febo.com mailing list