[time-nuts] Frequency processing scheme of HP5065 vapour rubidium standard

Ulrich Bangert df6jb at ulrich-bangert.de
Mon Nov 6 08:13:33 EST 2006


Hi folks,

me and my friend Frank (who has got his hands on two of these HP5065)
have a problem in understanding the frequency processing scheme of these
beasts. 

At a first glance everything looks pretty straightforward: A 60 MHz
carrier derived from the OCXO is multiplied by 114 to get a microwave
frequency approx. 5.315 MHz above the rubidium's resonance frequency.
Then the signal of a 5.315 MHz oscillator is subtracted from this
frequency to get a signal ON the rubidium's resonance.

One of the tricky things that HP built into the HP5065 is a synthesizer
board that has two jobs:

1) make the 5.315 MHz signal phase locked to the 5 MHz OCXO

2) make the 5.315 MHz signal tuneable in fine steps of 1/10000 of 5.315
MHz thereby staying phase locked to the OCXO.

The tune is set by means of 4 thumbwheel switches and a high/low switch.
The overall effect of setting a certain tune value is to fine-set the
OCXO's output frequency in small steps. The manual says the sense of it
all is to generate different 'time scales'. We are well aware of the
existence of different time scales and our understanding is that by fine
tuning the OCXO one can indeed 'generate' these different time scales.

But what surprises us completely is the fact that different physics
packages need DIFFERENT thumbwheel settings to generate the SAME time
scale as seen with the two devices available. We would have expected to
see different C-field-settings between different physics packages but
how is it possible that the atomic resonance beetween different physics
packages can be THAT different? We are not talking about rubidium drift
in the 10E-12 to 10E-11 region but about frequency differences some
orders of magnitude greater.

Anyone who can give us a clue? We must be missing something really
fundamantal.

Cheers

Ulrich Bangert
df6jb at ulrich-bangert.de
Ortholzer Weg 1
27243 Gross Ippener 




More information about the time-nuts mailing list