[time-nuts] Digital Pots, wiper noise and fine tweaking oscillators

[Dana Whitlow]

Perrier,

In answer to your 2nd question, I am unaware of any oscillator technology
for which
tuning the actual oscillator frequency has either (much less both) of
instant full
response or absolutely zero effect on its frequency drift trend
afterwards.

Instantaneous full response is generally not expected nor even desirable,
as long
as it's fast enough as not to interfere with closed loop bandwidth &
desired loop
stability goals.  This is true even if the EFC input is being adjusted
manually- it is
frustrating to tune a slowly-responding oscillator by hand, for the
operator then has
to wait around for some time before it is possible to meaningfully assess
the effect
or each trial frequency adjustment.

On the other hand, a truly instantaneous full response to EFC input is
possibly
incompatible with use of a high-Q resonator in the oscillator.  Further, it
would
imply infinite tuning bandwidth, which would require extra care, possibly
*considerable*
extra care, to assure that high frequency noise in the tuning circuitry is
not passed
along to the VCO.  If proper attention is not paid to this detail, one is
likely to find that
the phase noise sidebands of the oscillator's output are intolerably strong.

As to trend effects there are numerous mechanisms by which resetting the
frequency
of an oscillator will also alter its drift trend.  The severity of this
effect varies widely with
circumstances, but in many cases leads to intolerable problems.  Hence in
critical
applications is it often preferred to leave the primary oscillator alone
and deal with
drift in some other way.  Celestial navigators of yore were well aware of
this issue and
adjusting the rate of a watch used for this application was a cardinal sin
(especially if
attempted at sea).  Instead, they started a voyage armed with a record of
past drift rate,
updated this as circumstances permitted, and probably kept a healthy degree
of
skepticism regarding each new update.

In the more modern context, the primary oscillator is left alone and the
desired final
output frequency is achieved by use of a "microstepper", which is a
effectively a special
purpose synthesizer.  A good example of this is the well-regarded MHM-2010
active
hydrogen maser, which comes equipped with a good integrated microstepper.

Obviously an exception can (and must) be made in the case of a secondary
VCO which
is being phase locked to the primary oscillator, as a part of such a
microstepper or other
synthesizer, for example.  Here, long term drift of the VCO is of less
immediate concern
until it grows so large that the phase locking loop is no longer able to
maintain lock.

Regarding "digital pots", I'll write separately, probably as a private note.

Dana



On Mon, Jan 27, 2020 at 12:54 AM Perry Sandeen via time-nuts <
time-nuts at lists.febo.com> wrote:

> Yo Bubba Dudes!,
> Previous posts mentioned wiper noise and stability of a mechanical pots
> after tweaking.
> My questions are:
> Do digital pots after setting have wiper noise?
> When making fine tuning tweaks to the EFC of an OCXO, can one move it to
> its *dead on* setting right away or is there some lag that must be
> considered requiring to do it in steps?
> Regards,
> Perrier
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to
> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.
>