Chapter 19. Troubleshooting and Problem Isolation  

General

Is this a new installation problem or troubleshooting existing equipment? What are all of the basic symptoms? If you can't get the symptoms clear enough to write them down in black and white, you will have trouble explaining it to any other troubleshooter. Has the product practice or user manual been read and understood? Was this caused by a human error, such as accidentally bumping a working pair of wires? Wires can break. What was the last thing that happened before the symptoms were seen? Segment the problem and solve the segments before you solve the big picture.

Keep in mind that there is a standard terminology problem that always exists around the telephone industry. In some places T & R indicates Tip and Ring, which identifies one wire from the other wire in a two-wire pair (don't confuse the term Ring with the term Ringing signal). In other places T & R indicates Transmit and Receive. Frequently these are not single wires. In DS-1, these are Transmit pair and Receive pair. Of the Transmit pair, there is a Tip wire and a Ring wire. We could just as easily call them Transmit pair, wire A and wire B. But by some tradition, they were named Tip and Ring. Sometimes on a schematic they are named TT and TR (for Transmit Tip and Transmit Ring) and then RT and RR (for Receive Tip and Receive Ring). On still other equipment these are named T and R on the transmit pair and T1 and R1 on the receive pair. Again, remind yourself that the Transmit signal at the near end becomes the Receive signal at the far end. In some places Transmit is abbreviated as XMT or Tx and Receive is abbreviated as RCV or Rx. There is no consistency, so be aware.

Another set of terms is East and West. In the early days of long-distance circuits, AT&T Long Lines developed standard terminology to explain what long-haul equipment at one town was connected to what equipment at the next town, etc. So they arbitrarily called all transmission directions either East-West or West-East. So it is quite common to be tracking a signal from one Office Repeater Shelf marked "East," meaning this is facing the span to the east. It might then go to another Office Repeater Shelf marked "West," meaning this is facing the span to the west. It is often more foolproof to simply draw lots of arrows on a circuit diagram, indicating the direction of all signals.

Switch Options, Framing and Line Code

With DS-1 signals inside the Central Office, you have to verify where the framing format is SF (also erroneously called D4) or ESF. On a few products, it doesn't matter too much (they automatically detect and set SF/ESF), but it is good to know (i.e. line repeaters and office repeaters generally don't care). You have to verify whether the Line Coding is AMI or B8ZS. On most products, this matters. It is good to sketch in these attributes onto the overall facility sketch as you go along. Verify all option settings.

Problem Classification

Exact symptoms are extremely important. An intermittent problem is much harder to fix than a constant problem. If it is intermittent, then its pattern of occurrence must be made reproducible. In other words, if it fails once per day every day, then that means something. If it fails only on the hottest day of the year, then that means something different. If it is an intermittent noise problem on a DS-1 circuit, then this is a little tougher. In general with an intermittent signal, if you can keep testing it until you can get it to fail consistently, you are half-way to solving the problem. To a large extent on DS-1 signals: Either it works, or it doesn't work at all. Rarely is there a marginal noisy problem except with grounding problems. If we have a problem with a poor logic bit error rate, then that can be tracked down. "The test signal measures perfect at A, then again at B, but it is poor at C." We can then concentrate on the B to C connection. If there is a problem with Bipolar Violations (BPVs), then this means something.

If you see a "clean" BER with one legal DS-1 test pattern, then you see a "dirty" BER with another legal pattern, often this is a clue to a bridge tap on the copper pair.

As a general rule, BPVs are generated from two sources. One is a poorly performing T1 line repeater or bad copper T1 line pairs. The other source is the last ten feet of line coming into the failing product. Very commonly, there is a problem with the last few feet of DS-1 jumper wires, like a single broken wire (not a broken pair) at a wire-wrap post. Or maybe a "shiner" (a wire where the plastic insulation has been sliced off to reveal the shiny tinned copper conductor). Shiners can easily make an intermittent short to a grounded shelf or anything else. Thermal intermittents are hard to reproduce without a lab temperature chamber. A can of freeze spray can help once in a while, but this is unusual. Fortunately, thermals are not that common anymore with modern low-power components. Another easy test is to tap an intermittent shelf with the handle of a screwdriver to see if some intermittent component will produce a bit error. If you can get an intermittent problem to the point where you can consistently make it come or go, then you are halfway to solving the problem.