Using dual trace for fault location

A time-domain reflectometer (TDR) is an invaluable tool for finding faults on metallic cabling. Other common tests, such as an insulation test or a ‘high-pot’, will show that the cable is faulted. However, they do not tell you where the fault is. It could be miles away, or it could be right under your feet.

The time-domain reflectometer sends an energy pulse down two metallic conductors and then analyses the reflected energy. The conductors can be twisted-pair, concentric, hot and neutral, conductor and shield, two phases, or any similar combination. It is important to note that TDR technology will not work on single-conductor cables. For that, you need different technology: a pinpointer. When the pulse encounters a change in impedance of the insulation, a portion of the energy is echoed back, while the remainder continues to travel down the cable. It will eventually echo off the termination.

The tester works essentially the same as a radar gun nailing poor saps on the turnpike, but modified to the application. It measures the time it takes the pulse to travel down the cable, echo and return to the source. It then takes the speed of the signal multiplied by the time it takes to return and divided by half (to the fault and then back), and that gives the distance to the fault. In order for the unit to make this calculation, it must know the speed of the impulse. On the turnpike, the air determines that. On cable, it depends on the type of insulating material.

Each different kind of insulation (EPR, XLPE, PILC, etc) allows the impulse to travel at a specific speed, called the velocity of propagation (VoP). It can be expressed as percentage of speed of light, or more familiar units such as metres per microsecond, and is usually selectable on the instrument.

The operator finds the VoP on a table, the cable specs or some other source, and enters it in the tester. Many high-quality testers include an on-board library of cable types and VoPs. So far, so good, but, like just about all electrical tests, it rests fundamentally on a laboratory model under ideal conditions. In the case of TDR, ideality means a perfect cable, with two conductors parallel down the length and uniform insulating material of precise thickness between the conductors. Cables are manufactured with varying degrees of quality. The thickness of the insulating layer may not be uniform. The conductors may not be strictly parallel. The material may contain voids and other imperfections. As it serves time in installation, of course it deteriorates under the elements and electrical stresses. This is why we have a fault to begin with. What does this mean in terms of testing it? Only that as the cable deviates from ideality, the test results become less precise. This is where operator experience takes over: the better trained and experienced the operator, the better the outcome.

A ‘perfect’ cable, tested with a TDR, would show a reflection at the end of the cable (upward for an ‘open’, downward for a ‘short’) and a perfectly flat line between. If there is a fault, of course another reflection will appear somewhere along the line, commensurate with its location. Different types of faults give different shapes to their reflections and the operator can determine the nature of the problem. A splice will give a distinctive shape and, if there is a known splice, that can then be disregarded on the trace and attention focused on another reflection. If the location of the splice is known, it can be used as a landmark to better position the fault. In addition, all the aforementioned imperfections affect the shape of the trace. Many are nothing more than the normal features of the cable and can be disregarded. Determining which reflection is caused by the fault that is affecting operation of the cable can be difficult, and that’s where enhanced features of the tester are invaluable.

Most testers offer a single trace, but the most advanced have a dual-trace feature. While single-trace TDRs are fine for most applications, do not ever let anyone try to claim that they will find every fault. No TDR will. For the most demanding applications in the power industry, it may be necessary to go beyond TDR and all the way to a TDR-thumper combination. However, a critical advanced technology that will identify many additional faults is the full-featured dual-trace, or dual-channel, TDR. Dual-trace models have two parallel channels that can be engaged separately or in tandem. This capability provides an enormous advantage in difficult fault location.

As with many technological breakthroughs, dual trace sprang from a simpler concept. Many older TDRs, often still in service, were equipped with plastic overlays so that a second trace could be added and superimposed on the screen during a live test. Modern units now do this electronically, possibly at no more than the push of a button. A second trace can be added either from a live test or from a previously stored result. Furthermore, on sophisticated instruments, the two traces can be at different ranges and have their own setup and zoom capabilities. On multicore cables, a known good pair can be traced simultaneously with the faulted conductor and the difference between the reflections noted. With twisted pair, a suspected faulted pair can be connected to one channel and a known good pair to the other. Even a separate but similar cable can provide the needed comparison, allowing for the natural differences between the two.

A good practice is to provide a historical record. Cables can be tested at the time of installation, during maintenance or during repair, and the trace stored for later comparison. A handy feature is the ability to add and subtract traces. Rather than displaying the entire trace, which may include many inconsequential imperfections, the TDR can subtract one channel from the other and display only the section(s) that differ. This can be a critical time saver in getting to the fault. Also important is to colour code the traces. Advanced testers offer numerous colour choices that can be applied to specific lines. Since dual traces criss-cross each other, combine and then separate, often in complex patterns, the ability to assign distinct colours can be critically effective in interpreting complicated traces.

High-resistance faults (arbitrarily over 100 Ω) are more difficult to identify than bolted faults or open conductors, and sometimes a conditioning step will make them more pronounced and easier to spot. ‘Burning’ the fault by injecting energy through a high-pot or insulation tester opens the fault and makes it more recognisable. After all, it will be cut out and spliced anyway. A comparison of traces before and after the conditioning step will call attention to the area most affected by the conditioning: the fault. And finally, two separate traces can be run simultaneously. 

Once the suspected fault reflection has been identified, additional refinements of the features list will help in finding it precisely. A cursor is moved to the point of reflection on the instrument’s display. The position should be at the start of the reflection, not the peak. Sophisticated models may have a feature that finds and positions it automatically. Another useful feature is that of dual cursors. With this feature, one cursor can be set at an identified landmark and the other at the suspected fault and, between them, the delta (Δ) is calculated and displayed; that is, the distance between the landmark and fault. This can be an enormous time saver in tracing and walking the route. Another use of dual cursors and delta calculation is that the first can be positioned at the end of the test cable connecting the TDR to the cable being tested, again making the distance more precise and the fault easier to find.

Faults of lesser magnitude can be more difficult to distinguish. This can be aided by use of the ‘gain’. This feature magnifies the trace and makes cursor positioning easier and more precise. However, be judicious with its use, lest meaningless anomalies be magnified around it and obscure the fault reflection. Pulse width is also usually adjustable. The longer the pulse width, the greater the energy and the farther the pulse sent. The reverse of this is that a second fault close to the first can be obscured, as well as faults that lie close to the point of contact between the tester and the cable. This is typically about 10 metres. Trimming back the pulse width can separate the reflection of the near-end fault from that of the initial impulse. Once these features have been effectively utilised and the fault found, be sure to identify and save the trace to a file. Next time there is a problem on this line, the action time will be significantly less