[time-nuts] Synchronisation of crystal oscillators

[Magnus Danielson]

Hi,

On 2020-03-02 03:53, Richard (Rick) Karlquist wrote:
>
>
> On 3/1/2020 5:58 PM, Alex Pummer wrote:
>> don't forget the  oscillator is one amplifier with infinite gain on
>> his own frequency
>> 73
>> KJ6UHN
>> Alex
>>
>
> Actually, the closed loop gain is not infinite, rather it is whatever is
> needed to boost thermal noise enough to account for the
> RF output level.  I will give you that the gain is very high though.
> The "oscillator" is a Q multiplier of noise.
> The poles are NOT quite on the j omega axis.
> This was first explained (AFAIK) by Edson in "Vacumm Tube
> Oscillators".

I am thankful for you having me pointed that way. Reading him is however
a bit difficult, as his notation is not the same we use today. His work
also seems to be missed by many, as the published articles does not
cover this work. The model he builds cover many points, and provides a
concrete derivation 13 years prior to the Leeson article, which actually
is openly weak about why the integration mechanism is there, just builds
a model for it.

Trying to get history correct and uncover the line in the rubble can be
difficult.

Anyway, I actually intend to go back to Edson and re-read that part
carefully.

Cheers,
Magnus