[time-freq] Re: Atomic Oscillators & Rubidium Phase Noise

[K&J Aggett]

Hi,

Surely what is being talked about here is essentially the quality of the 
design (and how much money you are prepared to spend)?  Eric wrote:

 .....  I have some experience using Z3801 GPS-disciplined oscillators, 
and I can state that they are quite stable and accurate. However, they 
use an oven-stabilized crystal oscillator ........

And Ron wrote:
... Yep, but what about phase noise ..... ??  .... I think your GPSDO 
would not even come close .....

Don't all atomic standards incorporate oven stabilized crystal 
oscillators and isn't phase noise largely dependent upon that crystal 
oscillator?  My understanding is that the frequency control system 
can/does contribute to /total/ noise on the output.  I also appreciate  
that at 1PPS it is a long time between drinks for a GPS control system,  
and that if GPSDO's were better than rubidiums then there would be no 
new rubidiums on the market, but what are the numbers for phase noise 
and accuracy for /good/ GPSDO's?  Are they being made to be a cheap 
reference rather than as a realistic rubidium replacement?

Now I am supposing that Brooks Shera is not the technical equivalent of 
a HP design team nor has he the resources such a team would have had, 
but if the crystal oscillator in a cesium or rubidium standard was 
controlled in situ by his controller, does anyone know what degradation, 
if any, in phase noise output would be achieved?

Regards,

Keith.