[time-nuts] locking oscillators - an increase in power and/or stability ?

[Dr. David Kirkby (Kirkby Microwave Ltd)]

On 8 Oct 2014 23:52, "Bob Camp" <kb8tq at n1k.org> wrote:
>
> HI
>
> In the case of a magic Tee or a normal power splitter (both passive
devices), the current will not be limited by the combiner or the source.
With a proper combiner, the source will always be running into 50 ohms. You
will indeed get 6 db in the in phase sum case.
>
> Bob

Waveguide doesn't have a 50 Ohm impedance.  Impedance in waveguide is not a
useful concept.

In theory in a magic T the two inputs are isolated. If that was the case in
theory there would be no coupling between them and no injection locking.
In practice the isolation will not be perfect and so I can understand how
the frequency of the two will lock.

I suspect that the locking also gives rise to increased frequency
stability.

Has anyone ever investigated if pendulum clocks improve in accuracy if
coupled? I am not surprised by that. Neither would I be if one done it with
any other oscillator.

I agree that in principle two voltage sources can be combined to give twice
the voltage in an infinite impedance.  But for any liad resistance this
would require extra current too. and so extract more power from the
source.  Given the individual unit would be tuned to get maximum power out,
I don't understand how that would happen.

My best guesses are

1) Instrument errors.
2) Better matching.

Probably the former.

I am tempted to buy a magic T myself to try this. I have the Gunn modules.

Dave