We post our phone numbers at the end of every column, and we train hundreds of technicians. Therefore, we also get hundreds of calls each year asking for our help in resolving a tough case of trouble.
While we often, even usually, help find the cause of the trouble, we aren’t right all the time. Following are some cases where we got it wrong.
Hi Don,
It has been a long time since I have" bothered" you.
My 72-year-old brain is letting me down.
A technical issue has come up where when we cross-connect a POTS line to a cable with no end point plugged into the jack we get a hum on the pair. What do you think is happening?
Frank
My response was:
That vacant cable pair is open on one side on a lateral (bridged tap), open on one side beyond your terminal, or crossed with a non-working conductor. Keep me in the loop.
Donald
Frank’s reply was:
Don,
Digging deeper we found that the hum on the FAX was from the FAX machine itself. Or more correctly the electrical outlet it is plugged into had 40 volts on the neutral.
We were consulting with an independent Telco where customers were complaining of snapping on the line. In every instance the vacant POTS line tested OK. We finally proved the faults were due to corrosion at the cross-connect box on 25-pair modules that tested OK until high voltage caused the gas filled modules at the customer’s network interface causing snapping on the line.
Shortly after, another reader called describing snapping on the line when the POTS circuit tested OK. I thought, "I got this" and told the reader that corrosion on modules at the cross-connect box was causing the snapping. The reader called back and stated that the root cause was a bad DSL/POTS combo card.
Another reader wrote:
Donald,
This is the weirdest circuit I have ever worked on and I could use your advice on how to proceed. I am working on a Subscriber Loop Carrier circuit that runs about 70 miles between two states. It has 5 t-1 circuits to power up an old SLC 96 system. The system crosses two states. The power from one end comes from the Central Office (CO) in one state and the power for the other end comes from the SLC 96 remote in another state. The circuit has 15 239a repeaters and runs approximately 68,000 ft.
When I got to the trouble, all 4 shelves and the protect were down but the system had commercial power. Only the A shelf had been up for some time as this is a remote part of the state with only a few residential customers and some lines for the National Park Service which are seasonal in nature.
I am responsible for one half of the system given this crosses two states. When I got to the dividing line, I wasn’t getting power from the other side so I passed it off to their crew. Their splicers told me they had a bad power board and heat coil in the CO and tested from their last repeater with a TBERD to prove the circuit good on their side. That didn’t bring up my A shelf and the circuit is still down.
I know I can go between repeaters with a test card and look at the copper to verify clean and balanced pairs but there are 9 repeaters with approximately 10 miles of cable between the first repeater and the remote terminal. Also, that method won’t test the repeater card. Is the proper way to test these T1’s to use a TBERD?
Also, what are the metallic thresholds on these circuits? I’ve read 4 million ohms or more tip to ground, ring to ground and tip to ring is acceptable.
Jack
Jack,
The TBERD is the proper tool for testing T1 circuits, but I don’t know if the TBERD will test through 15 repeaters. Also, my metallic threshold for resistance is greater than 20 Meg Ohms tip to ring, tip to ground, and ring to ground. Let me know your results.
Donald
Donald,
I just wanted to let you know we finally got to the bottom of this circuit. The proper way to test is as follows:
There is a cord that plugs into the LIU card. The other side of the cord is a 15-pin connector. (I found the cord in the CO. The toll man had no idea what it was for!) The cord takes the signal from the LIU and leaves the power so the span is still powered.
The 15-pin connector plugs into a jack panel located in the power plant of the remote terminal. There are 6 bantam jacks labeled term, line and monitor for both side 1 and side 2. If you take a bantam cord and plug it into term side 1 and 2 you have now looped your T-1 circuit.
You can then go to a repeater and pop the card, put in a test card and with bantam cords access your circuit and look for the loop.
Who knew?
Most of the old SLC96 systems around me are either fed by fiber now or if copper-fed, abandoned. They will never fiber up this one so someone has to know how to work on it!
Anyway, even though you couldn’t help me with this one, your columns are a big help to me at work, so thanks and keep up the good work.
More "Oops" From Vernon May
If you read the OSP Expert column in the March issue on "Returning to Earth", you saw just one example of where we were completely wrong for years. We saw no way that AC power could impact DSL frequencies with the harmonics of 60 Hz. Only when we saw the noise created by the electrical appliances returning to ground on the copper cable sheath did we realize what was happening.
Bits Per Bin (or Tone) analysis has no value: Wrong. Nothing is more accurate in depicting how the circuit is reacting to field related stress. We just had to learn how to read the graphs properly.
• On individual trouble cases, we can often miss the problem or the right solution. We see a lot of impulse noise and higher resistance series opens during our remote analysis. This is not because they are common root causes, but because those are ones missed most by local technicians. Still, we often try to prove or disprove those faults first when the problem is often something other than what was originally diagnosed.
• Sometimes during the process of discussing a trouble case with us, the local techs end up providing the root cause solutions with local knowledge; for which there is no substitute. Recently, a local technician had heard about a rancher bragging about his electric fence that was so good that you don’t need insulators. The tech also knew that this farmer was on the same cable path that had a problematic chronic trouble report. 10,000 volts and blackened fence posts told a lot more than the tests that we were recommending.
• In another case, a 25-pair fixed count terminal was in a sewage processing plant. Three residential customers in the same cable count had chronic issues. The plant was fed by single phase power and it has massive water pumps. We saw impulse noise close to the plant and the terminal was not bonded to the power ground, with signs of arcing at the existing ground connection. Amazingly, we found no correlation between the plant-caused noise and the customer symptoms. These customers now look like they have individual root causes.
You Are Wrong Until You Aren’t!
If you think about it, with any form of troubleshooting, you are wrong until you are right. "Nope, not that, not that, yep, there it is." When we understand that, trying something new or standing up for what we believe in is not as hard. I heard a manager in the 1980s quote a much older industry axiom: The only way to avoid being wrong sometimes is to never do anything.
Signing off
Thank you, loyal readers. I hope you find even our mistakes to be helpful as you face the toughest problem at your company: keeping customers happy and preventing them from jumping to the competition. Please share your concerns, questions, and challenges with us so we can focus on issues that affect you. We won’t use your name, company, or any identifying information, without proper written approval. Contact Don at 831.818.3930 or [email protected]. Contact Vernon at 254.979.4749, [email protected] or visit www.vmaysolutions.com.
