Re: [Phys-l] question on averaging.

[Brian Whatcott <betwys1@sbcglobal.net>]

curtis osterhoudt wrote:
> With a Doppler system  :)
>
> --co
>
> ========Original message follows:============
> How does one observe the speed of a bob?
>   
Interesting proposal. If one substitutes measuring delta frequency
in place of delta position, one has several modalities from which to choose:
dopplar sonar
doppler radar and
doppler lidar.

Sonar? Supposing the bob is moving at a top speed of 0.3 m/s
then the delta f would be 0.3/342  - a  0.1% difference

But c varies appreciably with temperature: 331 to 342 m/s
over 20 degC for example
so the need for temperature control would be quite stringent.
To be confident of ones reading to 0.1 percent accuracy might
call for holding ambient temp within 
2* (342 - 331)*100 / (20 * (342 + 331)
say 0.2 percent of its nominal reading in degC -  say 30 millideg C

Perhaps doppler radar would work better?
Looking for 0.1% accuracy with a direct counting method,
might call for counting 1000  cycles at the pendulum speed of interest
during some small distance for which the bob's speed is reasonably constant,
say  0.01 m   I imagine this would be a millimeter wave doppler,
using a pendulum mounted reflector in vacuum perhaps...

Hmmm....recalling an encounter long ago with some researchers who
excited a brass instrument with a miniature spark plug in order to 
capture the
acoustic resonances, perhaps we could do as much with a fast plug discharge
in order to capture an electromagnetic echo pulse. Two shots like that
in a closely controlled time interval would certainly provide a nearly
instantaneous speed reading, one supposes.

Moving on to the supposition that measuring the positions of
multiple nodes in a resonance pipe
and averaging the adjacent measured differences of positions somehow loses
data, I have to suppose that I am misconstruing the objection.
This acoustic method is however subject to the stringent temperature
constraints already mentioned.

Written in haste

Brian W