Time Domain Reflectometer

Originated on July 24, 2001

 

Originated from Chris Riffe on July 24, 2001

 

 Hello:

 

 I trust you are all very busy; and so will appreciate your reponses all the  more.

 

 I would like to purchase a Time Domain Reflectometer to check our EM cables (the standard UNOLS assortment). Has anyone suggestions, preferences, warnings, or pertinent anecdotal information.

 

 Thank you all.

 

 Chris Riffe

 Technical Supervisor

 Louisiana Universities Marine Center

 8124 Hwy 56

 Chauvin, LA  70344

 

Reply from Marc Willis on July 25, 2001

 

 Chris,

       The only TDR which will work for long lengths (up to 30,000 ft.) of metallic cable (that I'm aware of) is the Tektronix 1503. Unfortunately, Tek no longer makes this TDR, all of their TDRs (and most everyone else's) are short haul (<5000 ft).  On the up side, there is a large secondary market for these, and there are several reputable refurbishers who have them available - cost is in the $5K range.  Tek sometimes has refurbished units available, but the cost is somewhat higher.  A search on the web will bring up the refurbishers.

 

       Long-haul TDR's seem to be a thing of the past - they were used mostly in telecom applications, which are now all fiber optic. TDR's seem to be used today mostly for in-building network troubleshooting, and other hi-rez units for circuit board QC.  We have a Tek 1503, and we have been pretty happy with it.  Try to get one *without* batteries, get the AC-only version.  It's unlikely that you'll be using it in the absence of power, and the batteries are a 3-Tylenol headache.

 

       More info if you need it.

 

 Marc

 

Marc Willis             Oceanic and Atmospheric Sciences

Oregon State University

104 Ocean Admin. Building

Corvallis, OR  97331-5503   USA

 

Reply from Dale Chayes on August 12, 2001

 

 Chris,

 

 I've used a number of TDRs over the years but never owned one. Nearly every time I've used a borrowed one, the battery was dead. The older Tektronix TDRs won't run without either a charged battery or external DC power.

 

 It took me a while to understand the message: no one in our business has enough use for one to keep it charged!  I subsequently discussed this with Tony Bogeman (MPL/Deep Tow engineer, retired) and he said ".... don't need one...." and showed me how to make a very acceptable measurement with a scope and a decent pulse generator. You get an estimate of characteristic impedance as a by product.

 

 We could put this on the agenda for RVTEC if anyone is interested.

 

 The magic part is not really in the TDR, it's in the (usually) non-existent docs for the cable: what is the propagation velocity (coefficient) for this cable (when it was working right.)

 

 A slightly different issue is that the commercial units that I know about either do very long cables (kilometers- for telecom work) or very short ones (meters- for complex wiring harnesses) but not both. You can't easily resolve which end is the problem and then figure out which connector/slipring/deck/umbilical cable is at fault with a single instrument.

 

 Finally, they won't tell you anything useful (that I know about) for cables that have suffered a bit of leakage or shield deterioration and thus have compromised high frequency attenuation but still have good continuity and no big impedance mis-matches. You need to sweep the cable to figure this one out if it matters in your case.

 

  -Dale

 

 Senior Staff Associate

 Lamont-Doherty Earth Observatory of Columbia University

 

 

Reply from Rich Findley on August 12, 2001

 

 Some years ago I wanted to buy a TDR for testing conductor cables, so I had the Tektronix sales engineer come out to demo several of the TDRs they had.  I used an old piece of .322 that was severely rusted with broken armor strands.  I don't remember how long the cable was but it was representative of what we wanted to test.  I picked a spot in the middle of the cable, and tried physically damaging while the sales rep tried to identify a problem with the cable.  I crushed the cable in a vice but he couldn't detect the fault.  His conclusion was the cable so "lossy" that it would be impossible to detect the problem using a TDR.  I don't remember the model number but it was somewhere around $8,000 around 10 years ago.

 

 Richard Findley

 Scientific Liaison

 University of Miami, RSMAS/

 Harbor Branch Oceanographic Institution

 5600 US One North Fort Pierce FL 34946