I think this will be easier than using RF. However, even though the IF is
25 kHz, you are still measuring small time differences. I just think this
hardware will be easier than RF. Mostly because you can control where it
goes better. This, of course, assumes you can do that in the target system
The OP has to decide.

That being said.... oneof those 10 GHz door opener or motion detector units
(does the GunnPlexer have a DC output from its detector?) would be another
interesting method. You can easily see small motions in the reflected
signal phase difference and this is at the 10GHz freq. so the resolution is
high. You do need to insure that the desired reflection is the only
one...similar to the LED type of system.

An aiming problem.

Steve N.

wrote in message
...

We did a unit at my last job called 'velocity of light'.

It was pretty simple and you cud accurately measure distance down to 1cm.

Infrared led emitter switched at 50MHz, focused using a lens to a 2 inch
beam,
reflected back by a mirror to an infrared detector (5 inches away from the
emitter).

The tx 50MHz was generated using a 50MHz xtal which drove the tx led.

Their was a 50.025MHz second xtal osc which was used to mix down both the
tx
signal and the rx signal to 25KHz IF (we now have 2 25KHz waveforms),
these two
25KHz carriers were then phase compared - so easy to see the smallest of
movements in the mirror on a basic scope.

So all the hard work is done at 25KHz (phase measuring) - one of todays
little
mcpu's will do this easily (ATmega16 for example).

You could just as easily use laser or maybe rf in place of the IR led's,
though
directing RF at such low freq's would be somewhat difficult.

Obviously at 50MHz, the phase difference would cycle every 6 meters (total
reflected path), but if your a bit cleverer (though not hard to do) you
could
get the freq to sweep from a low freq (say 5MHz) upto wot ever you like -
using
a pair of single xtal referenced PLL's to generate the two oscillator
freq's
(whilst maintaining the 25KHz difference) and then calculate an exact
distance
in the cpu.

This method is simple and relatively cheap to judge distances very
accurately
with no need for very short pulses (high bandwidths) and very fast logic.

Clive


Forgot to say.

A 90deg phase change in the reflected beam (at 50MHz) also results in a
90deg
phase change in the rx'ed 25KHz IF) - this is why no fast logic is required.