On 06/24/2013 11:56 PM, Jim Lux wrote:
On 6/24/13 5:21 AM, Brian Alsop wrote:
The time issue was effectively eliminated by the Michaelson-Morley
interferometer. One used a monochromatic light and an array of mirrors
which split the light in opposite directions around the track. The two
beams were recombined and an interference pattern resulted. One counted
the number interference fringes passing by while moving one mirror in
one path.
Knowing the number of fringes, wavelength of light and the mirror
movement, one could compute c.
There are easier ways to measure wavelength of an EM wave (Young's
double slit experiment, for instance), or measuring the voltage along a
transmission line carrying a reflected wave.
BUT, then, an interferometer (or melted marshmallows in the microwave
oven) tells you the wavelength of the EM radiation. But you need to know
the frequency of that signal in order to calculate c from that measurement.
and we're back to the "what can be done simply to measure time/frequency"
(I guess.. it's all really arbitrary.. a meter is so many wavelengths of
the light from a Krypton lamp)
The merit of the proposal I had was that once the actual optics and
amplifier stuff have been hacked up, it is relatively benign signals to
deal with, such that the counter he has will be able to handle it.
1 m optical ping length, 2 m of optical flight, almost 7 ns of delay, an
inverting amplifier stage and the period will be 14 ns plus the
re-generator delay.
To calibrate it, use two distance tubes and an end-plug with mirror.
Measure each of the distance tubes and then measure it with both in
series. Three measures, three unknown (delay of each tube and that of
the regeneration)
Additional measures can be taken with the tubes turned around as well as
all four combinations of turns and order between the tubes in series.
It does not make proper compensation for group delay in the electronics,
but it's a start.
After all, for a demo of use of time measurement it isn't bad. From a
physics and engineering side there is much to be improved.
Cheers,
Magnus
_______________________________________________
time-nuts mailing list -- [email protected]
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
