[time-nuts] nubie querie

[Bill Hawkins]

It's all very well to talk about zero-crossing detection in an hour glass,
but you haven't got an oscillator until you have some uniform means to
flip it. Then you can talk about measuring the period (two flips) over
some useful interval.

Seems like the flip should take a second or so. Too fast and the glass
breaks, too slow and you get sliding effects on the pile as it rotates.
Tom says the deltas are around ten seconds, so whatever happens in one
second doesn't much matter, as long as it is repeatable.

Same thing with the crossing detector. You don't want much sand left in
the upper glass, but you want it to be repeatable. Across the throat is
probably OK, but you could put a led in the center of each end, and
detect light scatter from the top on the bottom pile when enough sand
runs out.

Hmmm . . . This might be more interesting than the pendulum experiment
that I never got to. Won't need expensive counters, either.

Bill Hawkins