[time-nuts] HP 5065A questions
magnus at rubidium.dyndns.org
Mon Sep 26 01:51:53 EDT 2016
On 09/26/2016 04:04 AM, Richard (Rick) Karlquist wrote:
> On 9/25/2016 6:42 PM, Skip Withrow wrote:
>> 1. What the heck is rubidium cell flooding, and how does the TEC in the
>> 5065A fix this problem? None of the -many- rubidium oscillators that I
>> have been inside before has a TEC, and I have never seen the subject
>> The manual suggests that it may take several weeks at the 1A current
>> to fix
>> the problem (normal operation of the 5065A applies .7V across the TEC).
> Cell flooding is one of the dirty little secrets of Rb standards.
> I worked on the 10816 mini rubidium. There is supposed to be liquid
> rubidium in a reservoir somewhere. We used the tip off as the
> reservoir. The oven design (optimistically IMHO) attempted to have
> a heat leak on the tip off that kept it cooler than the rest of
> the cell, yet the oven was still supposed to isolate the cell
> temperature from the environment. Capillary action was supposed
> to keep the liquid rubidium in the tip off. However, it might
> come out of the unit was jiggled or turned over or stored, powered
> down, in a hot place. Then, Rb might get on the optical window,
> and then you have a serious error in frequency. The (again optimistic)
> concept was that you would just run the unit for a long time until
> the rubidium on the window hopefully evaporated and ended up in the
> tip off. Yeah, right.
> The 5065 (a VASTLY superior design compared to the 10816) was much
> less likely to be jiggled or turned over, and at least had a
> TEC to cool the Rb reservoir in case it got flooded.
You might also develop a thin film of rubidum on the window, which
filters out the D-lines and less optical pumping is achieved. Heating
and collecting it back at the tip-off does help. The 5065A is the only
one I know that has a TEC to support this action.
More information about the time-nuts