Vernon May and I wrote this column, and we want to let you know the topic will be discussed further by us at ISE EXPO 2016 in our seminar. So please join us in San Antonio, September 20-22.
When you ASSUME…. We all know the rest of that saying. However, assumptions are a necessary part of troubleshooting and of life in general. When we leave for an appointment, we generally time it based on the assumption that our car will start and the freeway is not locked up again.
Assumptions have a vital role in communications circuit trouble shooting. They literally make our jobs possible because they prevent us from having to physically check every circuit element on every trouble.
If we see packets flowing over a circuit, we can assume that the fiber is connected to light, the fiber to copper or W-Fi electronics are good, and the CPE at least has partial Internet access. Any of these things could be untrue, but the data will point to it, as long as we don’t preclude it mentally. In other words, remember what is a fact and what is an assumption during your decision-making process. When things stop making sense, go back to your most recent assumption and reanalyze it.
The downside of assumptions is that they often cause confusion and, sometime, service outages.
A recent short example:
• The DSLAM showed a subscriber port out of sync going back to the end of the stored history.
• The subscriber was up and surfing.
• The modem showed good bit rates and good reliability.
• We had pulled up the wrong port, but assumed we had analyzed the right one.
This mistake cost us about 10 minutes.
Another recent assumption was more complicated and cost dozens of dispatch hours and months of subpar service for 2 customers.
Subscriber 1 and Subscriber 2 are next-door neighbors
• They work in the same binder group.
• Their drops connect at the same serving terminal.
• Subscriber 2 reports “modem dropping”.
• About a week later, Subscriber 1 reports “modem dropping”.
• Nothing is found during either dispatch.
At this point, the universal assumption was made that the subscribers shared a common problem. And naturally, the “replacement game” was played for both customers.
• Modems changed out.
• DSLAM Ports changed out.
• Subscriber 1 was switched to a different cable pair.
• Both subscribers were still complaining.
• The Copper tests like rural copper: 3vdc, 85 MOhms of ground, and Capacitive and Longitudinal balance are good.
• Bit Rates and SNRM look good.
• Attenuation US and DS are reasonable for the distance and wire gauge.
• The circuit was completely dropping 40 to 50 times a day; the hits were spread out, but there were “quiet” times.
• The Copper tests like rural copper: 1 vdc, > 85 MOhms of ground/short, and Capacitive and Longitudinal balance are OK, within spec, but not great.
• Bit Rates and SNRM look good.
• Attenuation US and DS are reasonable for the distance and wire gauge, but on the high side of reasonable.
• The circuit was completely dropping 8 to 23 times a day; the hits showed no discernable pattern.
If the assumption was right that these subscribers had a common problem, the numbers would indicate that Subscriber 1 is being impacted more. The numbers were so close during comparison that there were no red flags. However, the attenuation US and DS for Subscriber 2 was approximately 4 dB higher than for Subscriber 1. The 2 subscribers shared the same cable run path, the same serving terminal and the drops were very similar in length, and 4 dB is a significant difference.
The cause of the extra attenuation and the modem dropping was a service drop that had been nicked during insulation stripping (tug your connections when done). A series (or high resistance) open condition caused the modem to go up and down. The longitudinal balance test did not detect this problem because there was not a 1,000 feet of copper beyond the fault. Subscriber 2 is now happy.
Subscriber 1 was being hit by impulse noise from 1 TV. The technician ran a new Cat 5 house wire avoiding the area of the “bad” TV, and the modem stayed up.
The assumption that the 2 troubles were related cost a lot of time and money. Assuming that Subscriber 2 would pass the longitudinal balance test when testing from the other end of the circuit prevented the high open from being detected early in the process.
So be careful about assumptions! We hope you will join us at ISE EXPO 2016 not only to hear more about what we have to say but to hear the best line-up of speakers and topics since the tradeshow began. If you want to discuss this topic or any other topic related to copper, call or text 831.818.3930 or email firstname.lastname@example.org.
For more information about your DSL and Vernon May Solutions LLC, call 254.979.4749, email email@example.com or visit http://www.vmaysolutions.com.