The time domain reflectometer, or TDR, is a device which sends a pulse down a transmission line. An oscilloscope is often used to view any reflections of the pulse from the end of the line. The TDR can be used to check that a transmission line is functioning and that it is terminated correctly. The TDR can be much simpler than its name might suggest.
The first TDR that I have built is based on an article from the "Eurotek" column in Radcom January 1998 (p56).
This diagram shows how the TDR is connected. A short patch lead (shown in the photographs below) is used to connect the device to the oscilloscope. Note that the power leads have been omitted from this diagram. The patch lead connects to a BNC T adapter, the socket part of which connects directly to the input of the oscilloscope. The other side of the T adapter is then connected to the transmission line under test. As pulses travel out from the TDR, they are displayed on the oscilloscope. The horizontal timebase of my 20MHz bandwidth oscilloscope is set to its fastest setting, and is synchronised with the TDR pulses. For a perfect transmission line that is terminated with the correct load (of the same impedance as that of the transmission line), the pulses will travel down the transmission line and be completely absorbed by the termination. There will be no reflected pulse.
For loads that are of too high an impedance, or for an open circuit, the pulse will be reflected back towards the TDR and will be displayed on the oscilloscope. This reflected pulse or "echo" will be of the same polarity as the outgoing pulse, but of a lower voltage due to losses in the transmission line (and in the load resistance if used).
For loads that are of too low an impedance, or for a short circuit, the pulse will be reflected back towards the TDR and will be displayed on the oscilloscope. This reflected pulse or "echo" will be of an inverted (negative) polarity compared to the outgoing pulse, and of a lower voltage due to losses in the transmission line (and in the load resistance if used).
The synchronisation of the timebase to the outgoing pulses from the TDR can be difficult, because the pulses are so narrow that the oscilloscope can struggle to get a lock on them. I intend to add a second output from my TDR just for the purpose of providing sync pulses to the 'scope. These will be of 50% duty cycle (mark : space ratio), and therefore will be easier for the 'scope to deal with.
(links to http://www.mw0llo.com/images/tdrpatchcomps.jpg)
A BNC T adaptor is connected to the input socket on the oscilloscope, and the TDR is connected to one side of this with a short RG-58 patch lead terminated with 50 ohm BNC type connectors. The other side of the T connector is connected to the transmission line under test. I find the BNC to SO-239 adapter useful for this.
(links to http://www.mw0llo.com/images/tdrassembledpatch.jpg)
This is the patch lead assembled and ready to go.
(links to http://www.mw0llo.com/images/tdrandscopeoper.jpg)
This picture shows the connected TDR. The pulse sent out by the TDR is very narrow, and repeats every 0.6µs. It is just possible to see the pulses on the oscilliscope screen in this picture, although the photos below have been optimised to make the waveforms easier to see.
(links to http://www.mw0llo.com/images/veryshortunterminated50ohm.jpg)
Here a short length of unterminated 50 ohm coax is connected. The pulse seems to have a "double-peak". In fact the second peak is the pulse reflecting back from the unterminated end of the coax. The cable in this case is a 3m long length of RG-58 50 ohm coax.
(links to http://www.mw0llo.com/images/shortunterminated50ohm.jpg)
When another 5m of coax is added to the 3m length, the pulse now takes longer to travel along the cable and reflect from the unterminated end.
(links to http://www.mw0llo.com/images/shortshorted50ohm.jpg)
Here the open end of the 8m coax (3m + 5m) has been shorted out. The pulse still reflects, buts its polarity is reversed.
(links to http://www.mw0llo.com/images/longunterminated50ohm.jpg)
With a much longer length of an old reel of coax (borrowed from John GW0OAJ), the pulse takes much longer to return. The tall pulse is the outgoing pulse from the TDR, and the reflected return pulse now comes back just before the next TDR pulse goes out. In this case, it takes nearly the full 0.6µs. Note also that the return pulse size is much smaller than with the shorter lengths of coax, showing that more than half of the initial power into the coax is lost as it travels first forward and then backward along the cable.
(links to http://www.mw0llo.com/images/longshorted50ohm.jpg)
This is the long coax shorted out, showing the polarity inversion.
After discussing my interest in the TDR at the Blackwood Amateur Radio Society, one of the members Mike Rackham GW4JKV gave a fascinating demonstration of the principles, and this encouraged me to finish my project. Mike showed how the device can be used to measure velocity factors in cable. I demonstrated my completed TDR at the Cwmbran Amateur Radio club on 19th November 2009 and it generated some interest at the club.
My next steps with this project are to experiment with balanced feeders (rather than just the unbalanced RG 58 demonstrations so far completed) and to see what happens when the pulses pass through an Antenna Matching Unit (AMU, also referred to as an ATU).
