[lug] ethernet cable tester

[J. Wayde Allen]

On Mon, 18 Dec 2000, Alan Robertson wrote:

> The 46060 Sure Signal tester can be found in this catalog:
> 	http://www.greenlee.textron.com/download/archive/ticat.pdf
> on page 61.  It is definitely yellow, too ;-)
> 
> I don't quite understand how you can test a cable just from one end.  How do
> you use this gadget?  [or how does it work?]

I don't know how this particular gizmo works exactly, but the spec.'s in
the catalog says it checks primarily for short circuits and
miswires.  The short circuit is easy, just check that there is no
connectivity between any two wires in the cable bundle.  

The miswiring is a bit unclear.  I'd think that would depend on the
specifics of how a properly wired cable would be terminated.  That is info
I don't have.  However, some obvious miswiring would create short circuits
between the wires in the cable bundle or complete opens where there should
be a nominal resistance due to the terminating electronics.  Another
possibility would be to test for any supply or bias voltages that should
exist on a properly wired and connected line.  For example, a properly
wired phone line will have nominally 50 volts DC between the two wires.
This doesn't look like a particularly sophisticated gizmo, so I'd be a bit
surprised if it does much more than this.  

A more sophisticated machine such as a time domain reflectometer could
send an RF pulse down the line and see how the pulse propagates.  Changes
in line impedance would cause reflections in the line.  Measuring the time
delay of these reflections from the initiation of the pulse allows you to
compute the distance down the cable to the impedance change as well as
allowing you to identify to some extent the cause of the reflection.  For
example, an open circuit has a reflection coefficient of 1 while a short
circuit's reflection coefficient is -1.

Just some possible ways that you could test a cable from just one end. 

- Wayde
  (wallen at lug.boulder.co.us)