[time-nuts] Stepping up the output of an OCXO

Dr Bruce Griffiths bruce.griffiths at xtra.co.nz
Sat Feb 3 00:43:42 EST 2007

Don Collie wrote:
> Why not use an optocoupler as an isolation amplifier? - to the best of my 
> knowledge it would provide infinite isolation.
> Cheers,.......................................Don Collie
And lots of noise.
You will need a cleanup PLL on the output side.
Also optocoupler isolation isn't infinite, there is a finite capacitance 
between the closely spaced emitter and detector.
Not all optoisolators are fast enough.
The most effective way of using optical isolation is in fact to modulate 
a stabilised laser and then couple the output into a fibre.
This technique is employed to distribute stable frequencies to Radio 
telescope antennas especially when a large number are employed in an array.
A phase locked loo is used at the other end to cleanup the signal after 
the photodetector.
Fibre has the advantage in this and other applications where long runs 
are involved that its propagation delay tempco is significantly lower 
than that of coax.
It is also possible to use fibre stretchers and other techniques to 
actively compensate variations in the fibre propagation delay.

Isolation becomes particularly important when one mixes a pair of 
signals a few Hz apart and then analyses the zero crossing times of the 
mixer output beat signal.
It is easier to achieve a high isolation between RF frequencies than at 
frequencies of a few Hz, JPL found fibre optic isolation can be 
invaluable in this case.


More information about the time-nuts mailing list