Last month, I mentioned the poor quality of some fiber optic patchcords I tested. This month, I examine the subject of cable quality, both fiber and copper.
The old saying goes, “a chain is only as stong as its weakest link,” and that is especially true for cabling. A well-installed cable plant that passes all performance and/or certification tests can be undone by faulty patchcords. For both multimode and single-mode fiber optic patchcords, the most important specification is loss (although reflectance can also be a big problem for single-mode fiber). Copper patchcords can suffer more problems since they are subject to bandwidth, crosstalk and other dynamic specifications that affect unshielded twisted-pair (UTP) cable performance.
Some of these issues arise from abuse. Tight bends or kinking, especially around connectors, can affect a patchcord’s performance, but most patchcord problems are probably traceable directly to the manufacturer, either in poor components used or poor manufacturing processes.
Many buyers seem to neglect patchcords, treating them as a commodity and deciding to purchase on price alone. Overseas, where most patchcords are manufactured, the goal is to produce them at the lowest cost. This is not an environment for quality products, and it fosters the cheapest components and corner cutting.
When I was in the manufacturing business, my company bought thousands of fiber optic patchcords to sell as reference cables for test equipment. We bought from local suppliers and reviewed processes, including how to test the patchcords and document the results, with them. We checked their suppliers of cable, connectors, adhesives and even polishing film. In some cases, we gave them test equipment for final inspection. As a result, we received patchcords with test results for each individual connector and with typical connector losses under 0.3 dB.
By 2001, production of high volume items, such as patchcords, was moving offshore, mostly to China, Southeast Asia and Mexico. The justification was, of course, cost, and over the last decade, prices have dropped by as much as 80 percent.
Since then, I’ve gotten fiber optic patchcords from various sources and, with the exception of those from major brand-name manufacturers, seen a continual decline in performance. Most of the cables I received from no-name manufacturers had losses above the TIA’s recommended 0.75 dB.
The best (or really worst) example is the patchcord pictured at left, which was sent to me as a sample from a supposedly trusted supplier. I tried to test it but could not get one end to mate with a power meter. The ST connector was missing one of the two locking slots on the connector nut, so it could not be coupled to a connector-mating adapter or power meter for testing or, obviously, a patch panel or transceiver on communications equipment. But the package had a label complete with red “QC” stamp showing it was tested on both ends and had losses of 0.17 dB in one direction and 0.19 dB in the other. Of course, the label looked like it was just a copy, probably the same one used in all these cables. Let me repeat that: the cable had one connector that could not be coupled for testing or use, and the test results were from a copy machine. The cable from which the patchcord was made had absolutely no markings.
So what would happen if you reached in a box to grab a patchcord for testing (or to hook up equipment) and picked out this one? If you tested with it, every test would have erroneously high loss, no matter how good the tested cable was. And if it was used as a patchcord for a 10 gigabit per second network with a limited power budget, it might not work at all. Then I’ll bet you’d have a few choice words for the buyer, distributor and manufacturer.
Is there a solution? Well, you could test every one of them yourself, but by the time you tested them all, the additional cost of testing plus the number of bad cables you’d throw away would probably make all of them very expensive patchcords. You are probably better off buying patchcords from companies who know their reputation depends on providing a quality product for a fair price.
‘Copper’ cable problems
Copper cable and patchcords present a different problem. Copper cabling systems are relatively straightforward to test if you have a cable certifier, but testing cable or patchcords can be more difficult.
Back in early 2003, I moved to California and built a large home office with a small telecom closet and Cat 5e jacks everywhere. I went to a large tech retailer to buy patchcords to hook up the computers, printers, wireless access points and switches. The retailer had a long row of them in various colors, and all were marked as having been tested to “TIA Category 6,” the new standard for UTP cabling ratified just a few months prior. Under closer observation, every cable on every patchcord I picked up had exactly the same writing on it, including an Underwriters’ Laboratories (UL) registration number, but no package identified the manufacturer or the place of manufacture. Shortly after that trip to the store, a cable and a test equipment manufacturer released information that it had tested patchcords similar to these and found that five out of six failed performance testing.
In December 2010, the Communications Cable and Connectivity Association (CCCA) announced that it had completed large-scale, electrical performance testing of Cat 6 copper patchcords. Test results show an 85 percent failure rate in patchcords produced offshore by largely unknown companies. A second, large sample set of Cat 6 copper patchcords produced by multiple, well-recognized manufacturers was also tested and showed 0 percent failure rate.
Warning against buying cheap patchcords is only the start for copper cabling. For example, I recently received a call from an installer with problems with some copper cabling. All the cable was hard to punch down; it was either breaking or not holding in the punchdown connector. He wanted to know if I had heard of this. When I inquired who made it, he said it came in a plain white box with a UL symbol on the side.
He reported that he thought he had found the problem. The wire in the cable was not copper but copper-plated aluminum! He sent me some samples of the cable and part of a box showing the UL symbol. The cable clearly listed a UL number (see photo), but the box did not identify the manufacturer or country of manufacture.
You clearly can see the difference between the usual copper conductors and the copper-plated aluminum. Now, we all know copper is very expensive right now, so much so that stealing cable for the scrap value of the copper has become a problem worldwide. But consider the effort this manufacturer has made, first to have someone pull 24-gauge aluminum wire then have the aluminum plated so it looked like copper. That took some effort.
Several companies are trying to get enough of this wire to test it and see how it performs, but there is little expectation it meets Cat 6 specifications. It certainly will be interesting to see what happens.
This is just another example of what happens when you buy components by price. Granted, standards tend to reduce all components to commodities and make it harder for manufacturers to differentiate their products from the competition, but brand name products are more likely to be higher quality, unless, of course, the brand has been counterfeited. And that is yet another serious problem.
HAYES is a VDV writer and educator and the president of The Fiber Optic Association. Find him at www.jimhayes.com.