Re: motion detector

[Brian Whatcott <betwys1@SBCGLOBAL.NET>]

At 07:54 PM 10/1/2003 -0400, Ludwik, you wrote:
> This morning I
> conducted the following experiment. The Vernier
> detector was consecutively placed at several
> distances (between 0.73 m and 0.98 m) from a
> smooth board. The sampling rate was 10 per
> second for 10 seconds (meaning 100 separate
> measurements of the same distance).
>
> I wanted to see if all the 100 results were always
> identical. It turned out that for some distances they
> were and for others they were not. For example,
> d=0.933 m was the same 100 times but 0.952 was
> mixed with distances equal to 0.950, 0.951 and
> 0.953.  These differences are very small. I attribute
> such behavior to the step-wise nature of the ADC
> (analog to digital convertor). What else can it be?///
> Ludwik Kowalski


I am unfamiliar with this device. I understand it is
an ultrasonic distance measure. If this is the case, it
is not obvious that there needs to be an AtoD as such.
Supposing c = 330 m/s and that there is considerable
difference in detecting intervals which are a quarter
or half cycle different, so that half or full waves
 are preferred lengths,  one might ask, given a distance
 spread of 3 mm between 0.950 and 0.953m, if the
 operating frequency of the device
  represents a wavelength of 6 or 12 mm
 i.e. for 12 mm lambda,  3.3E2/12E-3 = 2.8E4 Hz
= 28 kHz, or for 6 mm lambda, 56 kHz.

The latter seems like a lower frequency limit of
 usable transducers,  so I speculate that the sensor
 is a unipolar detector of the electrical waveform,
like a regular AM radio in fact, and is adapted to
 sense the amplitude of one polarity of the input.

A digital counter started with the transmitted pulse,
driven at an appropriate frequency, and stopped
by the return pulse would provide the distance
measure directly.


Brian Whatcott    Altus OK    Eureka!