[time-nuts] Got 60HZ?

[Michael Poulos]

Javier Herrero wrote:
> It seems that we're all trying to find the most obstuse way to obtain 
> 60Hz from 10MHz since the division is not integer... why not simply 
> multiply the 10MHz input signal by 3 and feed the resulting 30MHz 
> signal to any suitable divider by 500000 chain? the result will be 
> nice, spectrally pure and even 50.000% duty cycle :).
To multiply the frequency means weird RF circuitry to start off. If you 
start off with the 10 million HZ and divide, you only get better as any 
phase jitter gets lost in the division. As you divide far enough, the 
phase jitter gets to those "leap pulse" jitters. To design around the 
problem of Rb (my bad about Rb versus Ru) movements you have to have a 
second one ready to start once the original goes bad. This will occur 
with a cesium movement, like a good Agilent 5071A cesium movement every 
bit as well A fun question is why Rb movements are so common but Cs 
movements are rare and expensive.