Testing cabling components for the 21st century

Most manufacturers of copper-based network cabling and cable components have products that meet the current testing requirements for Category 5 unshielded twisted-pair systems as specified in the commercial-building cabling standard, tia/eia-568a, of the Telecommunications Industry Association (TIA--Arlington, VA). Many are also actively involved in TIA standards committees that are developing new testing requirements to ensure that an installed network channel will fulfill future needs.

Mar 1st, 1997

Ken Brownell

Superior Modular Products

Most manufacturers of copper-based network cabling and cable components have products that meet the current testing requirements for Category 5 unshielded twisted-pair systems as specified in the commercial-building cabling standard, tia/eia-568a, of the Telecommunications Industry Association (TIA--Arlington, VA). Many are also actively involved in TIA standards committees that are developing new testing requirements to ensure that an installed network channel will fulfill future needs.

This ongoing involvement is a good thing, because current testing requirements, although adequate for testing individual components, are not sufficient to measure the performance of a total network when they are used in real-world applications with bidirectional data flow. Total network performance will need to be measured and understood if we are to support existing--and future --high-bandwidth protocols.

Over the last year or so, a number of manufacturers of Category 5 cabling and components have published claims that they manufacture and test their products to a more rigorous standard than that included in tia/eia-568a, which was published last year. They can make such claims because the TIA`s TR-41.8.1 committee is already revising the tia/eia-568a wiring standard, and it is looking specifically at system performance requirements above the 100-megabit-per-second limit established for Category 5.

Network system performance beyond the Category 5 limit has yet to be fully defined, so it is too soon to ask what it really means. How, then, can you be assured that the network you are installing today will be "future-proofed"? How can you know that it will suffice for the bandwidth and protocol requirements of the twenty-first century?

The answers to these questions lie in the more rigorous test procedures currently being used to measure the performance of extended-performance Category 5 components. Products, for example, can be tested in both forward and reverse mode to simulate bidirectional data flow. You can also test with both common-mode and differential-mode terminations.

Common-mode terminations are necessary to simulate networking with active components, since most hub and switch manufacturers use common-mode, center-tapped transformers to shunt high-frequency noise to ground. The current standard only requires differential-mode, uncenter-tapped transformers for testing, but to test products in the atmosphere in which they will be used, it will be essential to add testing with common-mode terminations.

In addition, longitudinal unbalance of a patch panel can cause the interconnected station cable to exhibit unacceptable crosstalk levels by presenting a signal to the cable that is not precisely 180o out of phase. This longitudinal unbalance can be detected by measuring both differential- and common-mode crosstalk. It should not be surprising, then, that the TIA is seriously considering incorporating common-mode crosstalk testing into a future update of tia/eia-568a.

The current standard states that, for Category 4 and 5 connectors, the minimum return loss shall be 14 decibels or greater for all frequencies between 20 and 100 megahertz. This return-loss value was chosen to limit the peak reflected voltage to 20% or less above 20 MHz. There is interest by several manufacturers to tighten this requirement to 20 dB to reduce the deleterious effect of reflected energy on electronic equipment.

TIA documents now cover transmission systems at the component level, while the TR-41.8.1 committee is attempting to develop a link-performance standard, PN-2948. In developing this standard, it has been found that the test methodology currently presented in tia/eia-568a is not adequate to ensure link performance--primarily in the areas mentioned above.

We at Superior Modular Products have verified this assertion by randomly selecting the components of five manufacturers (including ourselves). We subjected these products to the proposed connecting-hardware test standards and found that three of the five failed at least one of the crosstalk criteria, and three of the five failed the test for structural return loss of 20 dB at 100 MHz.

This testing was not done so that we could criticize the products of other manufacturers. Most manufacturers, in fact, are building and testing their products to comply with existing standards and specifications. However, if a new testing standard is adopted, today`s products may not meet it. It is feasible today for manufacturers to test their products using bidirectional data flow and common-mode terminations. Since many manufacturers offer 15-year performance warranties on their components, it is incumbent on them to apply these more-rigorous testing standards.

Ken Brownell is the president of Superior Modular Products (Swannanoa, NC). He has a Ph.D. in electromagnetic field theory from the University of Tennessee.

More in Home
Sponsored
AV over Twisted-Pair