Frank Mara
CommTran Consulting
Extended-performance, high-performance, Level 5, Level 6, Level 7, Cat- egory 6, high attenuation-to-crosstalk ratio, 350-megahertz, 600-MHz--these are all terms currently being bandied about to describe the new world of high-speed cabling systems. Are these terms being used to bring enlightenment to the end-user or customer, or are they being used to further confuse the already confused?
About six years ago, the concept of the high-speed copper cabling system was first officially introduced to the cabling industry. The Electronic Industries Association and the Telecommunications Industry Association (eia/tia--Arlington, VA) issued telecommunications systems bulletins tsb-36 and tsb-40 (and the latter`s successor tsb-40a) to specify the electrical parameters and transmission performance characteristics of copper cable and connectors respectively. Three performance categories were established, with spectral bandwidth specified up to 100 MHz.
Before the publication of tsb-36, the only guide to cable selection was a vendor-sponsored cable-rating program. Unfortunately, this program contributed to the confusion surrounding cable selection by intermixing information bandwidth (specified in megabits) with spectral bandwidth (measured in megahertz). The two terms cannot be used interchangeably.
TSB-36 and TSB-40, now incorporated in the TIA/EIA-568a commercial building wiring standard, supplied the first benchmark that vendors could use to characterize their cable and connector products as standard-compliant. These standards, however, indicate only the minimum acceptable transmission performance. It did not take a heroic effort for vendors to meet the requirements for either cable or connectors. And, it did not take much time for this new cabling industry to grow into a sizable market.
In 1992, the marketing strategy of cabling and connector vendors appeared to rest on creating uncertainty about the transmission capability of currently installed systems and spurring a demand for new high-speed copper cabling networks. Category 5 unshielded twisted-pair (utp) cable became the mantra of the cabling industry. "Install my product and future-proof your network" was the message delivered over and over by manufacturers.
In a short time, a large proportion of the new cabling and components being installed consisted of Category 5 equipment--more than 60%, by some estimates. Testing procedures and test-set specifications were not yet available to permit verification of the performance of these newly installed cabling plants, and the electrical specifications needed to characterize high-speed networks were not completed at the time. Also, the standards that were available only specified a level of performance and provided guidance for network equipment designers, but these facts did not slow the marketing onslaught.
As the market grew, new players were attracted to the cabling industry, and Category 5 cable products soon became commodity items. There is, however, not enough profit in most commodity markets to support all the vendors selling in them. For such vendors, there are only two alternatives: Cut prices to gain market share, or introduce new product features and tout their benefits to differentiate one company`s product line from the competition. So far, most vendors in the cabling industry are adopting the latter strategy.
In 1997, it appears that vendors have once more turned to the marketing strategy that served them five years ago. Their aim seems to be to create uncertainty about the transmission performance of the currently installed base of Category 5 utp systems and to spur demand for new high-speed copper cabling networks labeled "extended-performance," "enhanced," or "Level 6."
Without doubt, most of the new extended-performance cable offers transmission capabilities that are superior to the minimum performance specifications of Category 5 cable measured at 100 MHz. And this is good for the end-user. The better the components of a cabling system perform, the better the overall system will perform.
Connectors--the weak link
What is insidious in these marketing claims, however, is the implication that such cable, when installed as part of a cable plant, will support signals in the 300 MHz-or-higher range. Most of the market hype to date has been about cable, but what about the limited performance of modular connectors at these transmission rates? It is well-known that connector performance has been a weak link in the performance of Category 5 systems, and a system will perform only to the level of its weakest link. Also, to what performance standard are the test values of these cables being compared? There is no standard beyond Category 5. And finally, what testing procedures are being followed, and which portable test sets are available to perform the tests? No testing procedures have been established for operation beyond 100 MHz, and no standard such as tsb-67 is in place that lets the installer or network manager evaluate the needed test equipment.
Look closely at the advertisements for extended-performance cable. How many of them specify link or channel performance? How many discuss the performance impairment to a link or channel caused by the limitations of the connectors? An increase in connector performance of from 8 to 16 decibels will be necessary to achieve the level of performance required for transmission at or near 200 MHz. How this performance increase will be accomplished is a technical challenge that remains unmet. What you will typically find in these advertisements is performance claims based on cable as tested in a laboratory and unencumbered by connectors.
Those who recall the introduction of Category 5 cable may remember the chaotic situation that arose only a few years ago and the resulting belief that the premature introduction of such products was best avoided. At that time, test sets were not available to validate performance claims, no test-set standard had been approved, and testing procedures had not yet been established. And yet, these facts did not deter marketeers from proclaiming that they had the products to future-proof the cabling plant, a situation that is beginning to seem all too familiar today.
Copper-cable vendors are doing a disservice to their industry by making claims of superior performance that are often misleading and of questionable value. Presently, there are no commonly used networks that require the bandwidth being touted. Category 5 systems still have a lot of life left in them when they are correctly specified, installed, and tested.
Before we in the cabling industry jump once more into a sea of confusion, wouldn`t we be better off cleaning up the remaining work to be done on Category 5? The standards groups must take the lead in this effort by providing the necessary specifications and test procedures for the high-speed networking world of the future. Expanded specifications need to be written, and testing procedures must be extended to incorporate these new specifications. If the standards bodies fail to move on these issues, then the industry will continue to be in a state of chaos--and marketeers flourish when confusion reigns.
ACR and distortion
Two major technical areas that have not yet been adequately addressed by vendors making extended-performance claims are attenuation-to-crosstalk ratio (acr) and signal distortion.
ACR is the signal-to-noise ratio of the wire pair, assuming that externally induced noise is minimal. For full-duplex operation, such as that required for 155-megabit-per-second Asynchronous Transfer Mode and 100Base-Tx, acr is a critical parameter.
This is because, as signal speed increases, attenuation becomes a proportionally bigger problem. Attenuation is a very predictable characteristic of copper cabling and is close to being proportional to the square root of the frequency. So, as frequency increases, attenuation also increases. As attenuation increases, near-end crosstalk performance must improve to maintain a positive acr.
There comes a point, however, where attenuation reduces the signal strength to a level at which it is almost impossible for receivers to recover the signal. A negative acr value means that the signal arriving at the receiver contains more noise than desired information. On the other hand, the higher the signal-to-noise ratio, the better the reception and the lower the bit-error rate. Connector performance and changes in signal balance due to connectors, as well as faulty installation practices, can all contribute to increasing near-end crosstalk and decreasing acr.
Distortion of the transmitted signal is primarily caused by channel attenuation because the higher-frequency components are attenuated more than low-frequency components. To compensate for these effects, electronic circuits must be employed to provide equal loss over the entire frequency band. The sophistication and cost of these compensating electronic circuits become significant with increasing bandwidth. The economics look much more favorable for optical fiber at bandwidths of more than 200 megahertz than they do for extended-performance copper.
Frank Mara is the principal of CommTran Consulting (Sandwich, MA), a telecommunications consultancy serving the cabling industry. He may be reached at (508) 888-3502 (phone and fax).