My feeling was that a fairly small battery could run the weight adjuster for a year or two. It has to power some type of RF receiver to pick up a locally broadcast time signal (probably on 2.4GHz). Given that it can characterize the local crystal, the actual 'on time' of the receiver would be small -- certainly < 0.01%. Power budget for radio is therefore maybe 2uA average. CPU etc maybe brings it up to 5uA. If we run the motor drive say for a second per day -- maybe in 10 0.1 second bursts, then that power budget is a few uA. A couple of AAA batteries could (naively) run this 10uA load for 100k hours assuming no other losses. However, the actual losses will be much higher, and I'd be happy to get a couple of years of runtime.
But maybe changing batteries every couple of years is ugly too..... On Wed, Mar 10, 2021 at 9:17 PM Hal Murray <[email protected]> wrote: > > [email protected] said: > > You guys have me thinking about another "non cheating" technique. I > am now > > imagining a small gear motor/screw arrangement that raises or lowers a > mass > > on the pendulum to trim out small variations in swing frequency. > > How are you planning to get power to the motor? > > How much would a thin wire running down the pendulum shaft screw up the > timing? How long would it last? I'm picturing thin/flexible wire with a > generous loop from the fixed clock frame to the pendulum. > > I assume people make wire for use cases like that. Is there a buzzword to > search for? > > Plan B would be to use a photo cell. You also need control as well as > power. > Maybe 2 photo cells, one for power and the other for data with a filter to > block the wavelength used for power. > > > > -- > These are my opinions. I hate spam. > > > > > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there. > _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
