David Hess
Berk-Tek Inc.
The mainstream cabling medium for local area networks today is unquestionably Category 5 unshielded twisted-pair cable. Although it is less popular, shielded twisted-pair cable is equally standardized, available and suitable for LAN applications. Furthermore, when fully and properly implemented, STP technology provides more bandwidth and greater resistance to interference than UTP.
The current preference for UTP is, understandably, cost-driven. However, it seems worthwhile to take a closer look at these two technologies to evaluate their current state and explore future options.
Definitions needed
First, some definitions are needed. A shield is a cylindrical conducting tube surrounding twisted conductor pairs. The shield can be used in more than one configuration. For example, it can be placed over each twisted pair in a cable, or it can surround a group of twisted pairs that may or may not be individually shielded. A shield can consist of foil tape, braided wires or both.
The terms "screened" and "shielded" are not necessarily synonymous, nor are they mutually exclusive or descriptive of the appearance of the material used in the shielding. It may take an international standards committee to reconcile this difference in terminology; we will not attempt to do so here.
For our purposes, a general definition of a shield will suffice: It is the cylindrical conductor tube surrounding at least one twisted conductor pair. The shield is an additional, or third, conductor element added to the two conductors of a twisted-pair transmission line. Compared to that of a simple twisted pair, this transmission-line model is more complex. Consider, for instance, the two-conductor combinations and interactions that are possible in such an arrangement.
Effects of shield
Even so, the effect of the shield is simply described. The shield isolates the two electromagnetic fields adjacent to its two surfaces and isolates the signal carried on a twisted pair inside the shield from all outside signals. Copper tubing would make an ideal shield, as it would provide maximum isolation. However, considerations of flexibility and cost require the use of tapes and braids. This, in turn, limits the effectiveness of the shield.
Some isolation also results from the twist in the conductors of a twisted pair. This twisting balances the pair within the electromagnetic environment, because each wire of the pair is equidistant from any other given point. If the receiver only detects the difference between the individual signals on each wire, and any interference is imposed equally on both wires, then the interference is eliminated by cancellation from the detected signal.
This method of isolation is effective when the signal and interference wavelengths are large relative to the pitch of the helix formed by the twisted pair. There are practical limits to how tight the twist can be, however, and this limits the isolation effectiveness of the twisted pair. The isolation effectiveness, then, diminishes with increasing signal frequency.
Shielded twisted-pair cable has greater bandwidth than UTP, thanks to superior isolation from the electromagnetic environment, which makes possible a better signal-to-noise ratio. UTP alone is only partially isolated; using STP is the only way more isolation can be achieved through the use of twisted-pair cabling.
It is baffling, then, to hear about situations in which users require their networks be "upgraded" from STP to UTP. It is a common misconception, in fact, that UTP evolved from STP. In reality, the most popular implementation of STP technology is associated with IBM`s cabling systems, and UTP can be associated with AT&T`s cabling. UTP is an offshoot of telecommunications technology, while STP is an advance in data-communications cabling technology.
It is indeed possible, as some critics claim, to end up worse off with a botched STP system than an improperly installed UTP system, but we should expect more than this from UTP proponents as a reason for avoiding STP. And yet, this is the underlying principle behind all the arguments offered to prove the inferiority of STP. Users with requirements for shielded cable can be made to feel foolish, and fictitious rules are cited that cause fear, uncertainty and doubt about using shielded cable.
Myths abound
Listed are some of the myths concerning UTP and STP:
- "UTP avoids electromagnetic compatibility problems that are inherent to STP."
- "It is impossible to provide a proper termination for the shield, creating electromagnetic capability problems."
- "It is impossible to properly ground the shield, increasing electromagnetic capability problems."
- "Shielding has inherent grounding, and therefore, safety problems."
The fears and doubts created by such statements are the basis for the growing rejection of STP technology as a LAN medium; however, you should consider that shielded cabling systems used in community-antenna television networks are constructed from the simplest and least expensive components available, and are assembled and successfully operated under more demanding conditions.
Modular connectors
A limiting element in twisted-pair network technology is the modular connector. Modular connectors are neither designed to work with shielded cable nor designed for very high frequencies. Although manufacturers have produced many shielded versions of modular connectors and standards are pending, these solutions may be inadequate.
(The shields you see around modular jacks on equipment are mainly there to raise the cut-off frequency of the hole in the side of the box, so the equipment will pass Federal Communications Commission regulations.)
There are users who require a shield--and not necessarily for applications operating beyond current UTP frequency limitations. For some of these applications, STP solutions may be inadequate because of size and flexibility limitations. These users must look to emerging technology, then, for a fully integrated shielded interface.
Future development
It appears that development of STP in the future could shift the point of diminishing returns for the bandwidth capacity of twisted-pair cable to higher frequencies. Considering such issues as installation cost, UTP appears to be superior to STP as a medium today. However, nothing can improve UTP`s inherent bandwidth limitations, so STP is likely to become more popular.
This projection of a bright future for STP may evoke fear in optical-fiber proponents, but further enhancements to twisted-pair cable will not delay fiber to the desk. Fiber technology provides the ultimate level of isolation, as well as other benefits. Fiber to the desk will become a mainstream technology when it costs less than equivalent twisted-pair technology. STP technology that is slightly more expensive than UTP surely cannot prevent the eventual switch to fiber.
David Hess is vice president for technology at Berk-Tek Inc., New Holland, PA.