OT Re: Is this true?? (TI & watches)

Jim Beacon jim at g1jbg.co.uk
Thu Mar 8 11:25:24 CST 2007

From: "Tony Duell" <ard at p850ug1.demon.co.uk>

> > >
> > > Does anyone remember that the oldest of said wall clocks required the
> > > owner to start the motor manually by spinning a little knurled shaft
> > > located on the back?
> I've enver actually used one, but I've seen them, and read about them.
> Another method that was common over here was to have a spring-loaded
> lever that was operarted when you pusehd or pulled the set-hands knob to
> set the time. Releasing said knob caused the lever to give a 'kick' to
> the motor's rotor to get it going in the right direction. These little
> synchonous motors were symmetircal and could run either way, so you might
> end up with a clock that runs backwards :-).
> Mroe recent clocks have a mechanism that detects if the motor is turning
> backwards and if so, stop it and kick it back again, thus getting it
> going the right way. On a clock fitted with a second hand, you will see
> it run backwards for a few seconds sometimes when you connect it to the
> mains until said device operates and getis going the right way.

I have one of the early manual start synchronous clocks running in the
lounge at this moment. It's a Ferranti clock from around 1934, and gives
very good service. It's quite happy running in either direction.

I also have an electrically re-wound clock from about 1920, which was used
as a time switch. The clock is regulated by a pendulum, but uses a large
rotary solenoid to rewind the clock about every five minutes - by using a
small proportion of the spring, the accuracy is improved, as the drive power
varies as a spring runs down, leading to inaccuracies in the timekeeping of
a simple spring driven clock.

A few other interesting timepieces about the place are a Junghan mantel
clock which uses an electrically impulsed pendulum (impulsed every swing of
the pendulum), but is fairly inaccurate, due to a short pendulum made of
steel. A GPO Clock No36, which was used in British telephone exchanges for
generating metering pulses - it is a Hipp type clock, with a 1 Second
pendulum, driven by an electromagnet, and gives out 1, 6, and 30 second
impulses. The rated accuracy is 8 seconds per week, but in practice, they
can be regulated to a better accuracy than that. Finally, there is a
Telephone Rentals master clock, which has a 1 second pendulum, and a gravity
impulse mechanism, similar to the Synchronome type (think cheap copy.....),
which is attached to a uniselector chronogram - basically a seven day time
switch made of uniselectors, allowing  on and off periods to be set to an
accuracy of 1 minute anywhere in a week.

Bringing the subject back to vintage computers, my current project is to use
one of my old Vaxstations as a NTP server on my home network, coupled to a
radio clock of similar vintage.


More information about the cctalk mailing list