> I've been following this with some interest also having had a bit to do
> with TDRs (&OTDRs) in a past employment..

I've been following it too, but I've kept my mouth shut so far because I
don't have any particular expertise in this area.

However, I just got this idea that doesn't require a fast micro nor any
comparator at all.  There may be some problems with this, but here goes:

The other end of the cable must be open (cable DC resistance is infinite),
and that the cable impedence is known.  Pull one lead of the cable to as
high a voltage as the cable and your analog electronics can stand via a
high value resistor, like 1Mohm.  The other cable lead is tied to ground.
A high speed opamp is set up as an integrator of the cable voltage.  Hold
the integrator in reset (FET clamp probably) and wait for everything to
stabalize.  Then as simultaneously as possible release the integrator and
switch in a resistor accross the cable that matches its characteristic
impedence.  Read the integrator value when you get around to it.  Its
voltage will be directly proportional to cable length.

THEORY OF OPERATION: Assuming a perfect impedence match with the resistor
that is switched in, the cable voltage will drop half way to ground before
the reflection from the end comes back.  Once the reflection comes back,
the voltage should go to 0 and stay there indefinitely because this is the
steady state condition.  The length of time at the 1/2 voltage is the
cable propagation time to the end and back.  The value on the integrator
will be proportional to this time, since the voltage is fixed.  Once
steady state is reached, the input to the integrator is 0 and it will
therefore hold its voltage.  Of course there will be leakage and offset
errors, so the integrator voltage will drift over time, but this is much
slower response than the cable.  You therefore should read the integrator
as quickly as possible after steady state, but a few microseconds more
shouldn't make any difference.  It should be fine to initiate the pulse
from a PIC output, then wait the A/D acquisition time plus a few extra
microseconds, then do a conversion.

Of course you won't get the nice perfect 1/2 voltage pulse followed by
steady 0, but integrating whatever you do get will probably still work as
long as the resistor matches the cable impedence reasonably well.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu