[time-nuts] A different timenuts interest

[jimlux]

J. Forster wrote:
> The problem is straight forward, except for sensing the position of the
> pendulum so the impulse is applied at the correct phase.
> 
> There must be a bunch of published designs, but if I were to try it, I'd
> use something optical or capacitive.
> 
> For optical, I'd put a annular ring of IR LED/Phototransistor assemblies
> around the center, wired OR the outputs, and use the signal to trigger the
> impulse. The bottom of the pendulum should be polished or mirrored.



The usual scheme is a couple of optical paths at the top of the pendulum 
with lenses to focus the beam down to a very small diameter (smaller 
than the wire diameter).  The two optical paths are at 90 degrees to 
each other, and are logically anded either by using a pair of 
detectors/paths, or by using a mirror to fold the path.

It's aligned with the pendulum perfectly stationary (i.e. it's been 
sitting still for a day or two) so the beam trigger occurs precisely at 
the center point.