[time-nuts] Example of clocks interlocking with each other

[Magnus Danielson]

On 09/14/2013 04:37 AM, David McGaw wrote:
> Not true.  It depends on the strength and reciprocity of the
> coupling.  If essentially equal as with the metronomes which are
> coupled pendulums (the math exists for this), the systems will come to
> some combination frequency, though as has been mentioned in a
> many-oscillator system there may be multiple solutions.
>
> The TV oscillator example is different as the driving source is not
> influenced by the locked oscillator.
>
> Now, can we lock multiple atomic oscillators together into a
> super-accurate ensemble?
No. You do not accuracy that way.

In order to achieve accuracy you need to show that systematic errors
distributes systematically around the accurate value. However, several
shifts offsets away in ways which doesn't help. You can achieve
stability, but by keeping clocks sufficiently isolated, you can alter
their weight as they perform and those performing badly will be given
less weight and the ensemble performs better.

Cheers,
Magnus