[time-nuts] Regulating a pendulum clock (Jim Palfreyman)

[Jim Palfreyman]

Well last night I did a quick and dirty.

I got my (GPS locked) 3325B to generate square waves (0-5V) at 1 Hz. 500ms
on. 500ms off. I ran this through a relay that delivered 10mA at 25V to the
coil.

By adjusting the phase of the 3325B I got the ON to occur as the magnets
approached. But I had no idea whether the magnets were attracting or
repelling at that point.

Watching the timing of the pendulum it started to radically deviate. I
concluded it was repelling and so reverse the polarity. It started heading
back the other way. Cool. As it was late - I went to bed.

Next morning the pendulum had locked in nicely. It was only deviating a few
ms either side of a fixed point. I will let this run for today and make sure
it has settled in and doesn't vary.

The pendulum clock is fast by 200ms and so I'll adjust the phase to correct
it tonight.

Next step will be to drive it properly from my rubidium driven
microcontroller.

While I organise that I'll turn off the electromagnet and see if I can pick
up the tides.

Regards,

Jim Palfreyman