[time-nuts] Basic regenerative-divider questions
bruce.griffiths at xtra.co.nz
Fri Sep 28 22:20:34 EDT 2007
Magnus Danielson wrote:
> From: "John Miles" <jmiles at pop.net>
> Subject: Re: [time-nuts] Basic regenerative-divider questions
> Date: Fri, 28 Sep 2007 18:51:58 -0700
> Message-ID: <PKEGJHPHLLBACEOICCBJEEKKBJAB.jmiles at pop.net>
>> ); SAEximRunCond expanded to false
>> Errors-To: time-nuts-bounces+magnus=rubidium.dyndns.org at febo.com RETRY
>>> You can do better than that, a single regenerative divider can be
>>> configured to divide by 4.
>>> A pair of parallel feedback paths (with amplifiers), one tuned to F/4
>>> and the other to 3F/4 are best.
>>> NIST did some work (together with Indian collaborators) on this type of
>>> generalised regenerative divider recently.
>>> Papers are stored on my Windows machine, will boot it up and locate them.
>> Thanks much, Bruce. I suspected either you or Enrico R. would have some
>> knowledge of that.
>> Note that I need to end up with 40 *and* 20 MHz, hence the plan to cascade
>> two /2 dividers. If there is a better topology for obtaining both of these
>> outputs, it would be good to know. I'd imagine that a /4 divider running
>> alongside a /2 divider would be better from the additive-noise perspective.
>> I will probably end up wanting a 10-MHz output as well. The obvious
>> question would be, should that be a separate F/8 + 7F/8 path, or a /2
>> divider following the /4 divider? I haven't seen many references to /8
>> regenerative dividers but I suppose they'd be workable. Availability of
>> 8.75 MHz crystals might be what decides that question.
> The article in question is...
> but also
> See for yourself.
To which you can add Enrico's paper:
More information about the time-nuts