[time-nuts] Grandfather clock sync'd to 1PPS (from time-nuts Digest, Vol 200, Issue 4)

[Gerald Swann]

Hi Philip,

   I have read about the approach you described with a magnet on the
pendulum and a stepper adjusting a steel plate up and down.  The one I saw
(on Hackaday) seemed to work very well.

   I don't really see what the difference is between that method and the
method I used.   My electromagnet generates a magnetic field the same as
the method you described.  My pendulum is not being pushed/pulled into
correct timekeeping.  For example, over the past thirty minutes,  the
magnet is energized for 16 - 20 beats per minute (which for this clock is
29 - 36 seconds) depending on the error during that minute..

   I wasn't clear in my first email on how the error (phase difference) is
measured.  Because of the strange period of the pendulum (actually 33 beats
per minute), and the period of the 1PPS signal, the two line up only three
times per minute.  That is 11 beats, at 20 seconds, 22 beats, 40 seconds,
and 33 beats, at 60 seconds.   I hope I said that right.  So for all
practical purposes, the error can be measured once or twice per minute.
The magnetic field period is calculated and applied for whatever portion of
a minute is necessary.

   I believe that you are thinking that you can adjust your plate up and
down to apply a variable magnetic field to speed up the pendulum
as required.  Maybe on a brand new clock, or a well maintained one, you
might be able to set the distance and let the clock run for a long time
with a given field..

   The mechanism on my clock has serious wear.  Over a two hour period, it
speeds up and slows down periodically as the gears bind.  So to maintain a
constant phase error, the pulse generated compensates for that problem.
Seems to me, it works the same as the method you described.  For my clock,
you would have to adjust your plate up and down over a 2 hour period to
keep the time error the same.

   Regardless, good luck with your project - I'd love to see the results.
Keep us posted!

Gerald
AI4DE


-----------------------------
>
> Message: 7
> Date: Wed, 10 Mar 2021 10:43:19 -0500
> From: Philip Gladstone <pjsg-timenuts at nospam.gladstonefamily.net>
> To: Discussion of precise time and frequency measurement
>         <time-nuts at lists.febo.com>
> Subject: Re: [time-nuts] Grandfather clock sync'd to 1PPS (from
>         time-nuts Digest, Vol 200, Issue 4)
> Message-ID:
>         <
> CAMd_wWrfz5Pm2Zeb6ASga6b-15vYOxC4v2S8QQFsfWvquRQRLg at mail.gmail.com>
> Content-Type: text/plain; charset="UTF-8"
>
> I've been wanting to do this as well, but I feel that it is "cheating" to
> actively drive the pendulum. I want to have a passive approach to
> disciplining the clock so that the clock itself is keeping time.
>
> One approach that I have seen was to put a magnet at the bottom of the
> pendulum and then have a metal plate beneath it -- and the correction is
> performed by adjusting the vertical position of the plate. This effectively
> changes the local gravity and hence the tick period. However, I don't think
> that this will look "nice". I'm hoping that I can do the same magnet trick,
> but put it much higher on the pendulum so that it is hidden when you open
> the clock case door. I envisage having the metal plate position being
> controlled by a small stepper motor. I'd like to have a range of maybe
> +/-50ppm with a resolution of around 1ppm. This seems doable....
>
> Thoughts?
>
> Philip
>
>