[time-nuts] Synchronisation of crystal oscillators

[Hal Murray]

richard at karlquist.com said:
> Injection locking requires that the oscillators be within each others 3 dB
> bandwidth, or at least close to that.  Oscillators on different nominal
> frequencies are no problem (EG 3 MHz and 10 MHz). 

Will oscillators on integral multiples of some frequency lock?  For example 5 
MHz and 10 MHz.

Is there a significant difference between odd and even multiples?  If I have 
10 MHz, is that more likely to lock to 5 MHz as compared to 3.33333... MHz.


> OTOH, if you carefully adjust a couple of HP10811's to zero beat, you will
> have to go to extraordinary measures to keep them from injection locking. A
> lot more than just running them on individual voltage

How extraordinary?  How likely are 2 GPSDOs sitting side by side to ock?



-- 
These are my opinions.  I hate spam.