Communications wiring standards are defined by a number of national and international
agencies, including the Electrical Industries Association (EIA), the Telecommunications
Industry Association (TIA), the Institute of Electrical and Electronics Engineers (IEEE)
and the National Electrical Manufacturer's Association (NEMA). These committees recognize
and develop definitions for multiple electrical and communications definitions, i.e., how
things operate, how devices will inter-connect, how signals will be formatted, and how
manufacturers can design and construct equipment so as to ensure compatibility within the
marketplace.
The process of setting national and international standards is time-consuming as it
involves a committee and meeting structure, often on an international basis, and may take
several years to get draft standards distributed for comment. Such a draft standard is
voted on by the members of the various organizations and, upon concurrence, testing and
announcement follows before an end-user may be assured that their equipment or wiring
standards will be in conformance with a national or international standard. In the United
States, the final standard is issued and maintained by the American National Standards
Institute (ANSI).
Large manufacturers within the communications industry, such as IBM or DEC, often create a
de facto standard by virtue of their dominance of a particular field. For example, IBM's
dominance of the early years of token-ring development paralleled the introduction of the
IBM wiring standard and its reliance upon shielded twisted pair for data communications.
Seven Levels/Grades Media
Since Winter of 1990, electrical manufacturing associations have redefined existing
wiring product lines and introduced new cables according to a generally agreed-upon set of
"levels" or "grades" of media. For the purposes of this paper, these
levels are defined as follows:
Category 1 POTS (plain old telephone service) and Low-Speed Data (up to 9600 bits per
second)
Category 2 Integrated Services Digital Network (ISDN) Data (up to 4 megabits per second)
Category 3 Data Grade Media for Local Area Networks (up to 16 megabits per second)
Category 4 Extended Distance Local Area Networks (up to 20 megabits per second)
Category 5 Data Grade Media (up to 100 megabits per second)
Category 6 Coaxial Cable (up to 100 megabits per second)
Category 7 Fiber (in excess of 100 megabits per second)
It may be helpful to understand that Category 3 is the appropriate definition for the
recent 10BaseT (Ethernet over unshielded twisted pair) standard, while Category 5 is the
targeted range for the Fiber Distributed Data Interface (FDDI) over copper standard, or
CDDI (Copper Distributed Data Interface) now under development.
It is noted that, as a result of the adoption of national wiring standards, major
manufacturers and distributors of wiring projects (e.g. Anixter, Belden, etc.) have
developed marketing efforts for their products which define wiring performance
"Levels" or similar indexes. Much of these "Levels" parallel work by
the EIA/TIA defined as "Category," but in fact project standards which do not
yet exist. This results in considerable confusion as to what is actually a standard and
which materials are conforming. A vendor product specification is only relevant if a
standard has been specifically adopted and published by the EIA/TIA committee process.
Shielded versus Unshielded ("THE debate")
Throughout this document are references to shielded versus unshielded twisted pair cables.
By way of background, shielded cable was developed to provide protection for data signals
from electrical interference generated by power cables, fluorescent lighting, and other
data signals. At an earlier point in communications wiring history, it was a basic
requirement for maintaining data speeds as low as 9600 bits per second. Driven by the
enormous existing base of unshielded telephone cables present in buildings, manufacturers
developed devices to balance or re-balance the signal characteristics over twisted pair
cabling. In time, techniques and standards defining the size, purity, and twists in
twisted pair cabling were developed to take advantage of both existing unshielded cables,
and the relatively lower costs of placing new unshielded cabling versus new shielded or
coax cable.
The large-scale introduction of network hardware and software produced intelligent hubs
that provided sophisticated pre-balancing and re-balancing, and restoration of signals
carried over unshielded twisted pair cables. As a result, national standards now define a
ten megabits per second standard for Ethernet over unshielded twisted pair (UTP) and a new
16 megabits per second standard for token-ring over UTP. As indicated above, discussions
are underway on signal processing required to carry the 100 megabits per second speeds of
Fiber Distributed Data Interface (FDDI) over copper cables (CDDI).
Industry leaders speaking about future wiring talk about a combination of unshielded
twisted pairs and fiber strands. Because of the developments in signal
processing, the market for UTP products, and the projected decrease in the cost of fiber
optic cable, many manufacturers and industry leaders believe that shielded twisted pair
will become unnecessary. The factors that make an impact on an organization's decision to use UTP or
shielded twisted pairs (STP) lie in the marketplace, as vendors will weigh the higher cost
of advanced signal processing required for faster speeds over UTP versus the costs of
placing shielded twisted pair cable and fiber.
There have been three distinct stages in the integration of voice and data wiring at UCSB.
The first stage of wiring was distinctly telephone wiring, with the telephone companies providing, maintaining, and enforcing standards in phone wiring. The
telephone companies did provide limited data services over their
cable, but did so only as an adaptation of their voice wiring standard. In the early
development of data terminal attachments, computer shops ran their own
cables (coaxial, shielded, and unshielded), according to their own system requirements.
During the first campus wiring stage, Communications Services pulled a package of separate
data, voice, and video cables to work stations in all new and re-modeled campus areas.
These cables did terminate in a common wall plate, but followed individual distribution
routes within terminal facilities. Existing cable packages placed in the years 1982-1987 will not meet the specification for local area network wiring defined in the 10BaseT
standard.
The second stage of joint wiring was partially developed by "pirating" of
telephone cables by and for data users when it was available. Upon de-regulation of
telephone services, increased usage of telephone cabling for data took place. Under a
common utility, Communications Services, most ASCII, and later System Network Attachment
(SNA) attachments, took place over telephone cables, often sharing larger cables with
telephone services.
Early telephone multiple pair cables (campus pre-1975) were not twisted pair cables and
may, as a result, be unsatisfactory for data transmission as part of a local area network.
Multiple pair telephone cables placed after 1987 and prior to 1989 should be tested for
use in a 10BaseT local area network according to the EIA/TIA-568 specification (see
Appendix A-1).
The third stage of voice-data wiring integration occurred on this campus with the adoption
of the IBM Type 2 single sheath cable standard by Communications Services in late 1985.
The IBM-Type 2 standard specifically defines shielded and unshielded wire within a single
sheath for data and voice wiring. Composite cables containing both fiber and multiple
twisted pair could result in an economically viable "single cable" technology
carrying all communications.
In the period following the adoption of the initial Campus Wiring Standard, two patterns
of voice-data wiring integration have occurred specifically in the riser cabling:
- The first occurred in new buildings where multiple Category 5, twenty-five pair cables
were placed as risers with Category 5 terminations to be used by both voice and data
applications. While technically this shared resource worked as intended, the competition
for riser pairs was very strong, and the required labeling and dressing of jumpers to form
a reliable system was un-sustainable. Several building were augmented with additional
riser cables and in others, departments placed additional cables – frequently
multiple, individual Category 5 sheaths resulting in inefficient use of riser duct
capacity.
- The second pattern emerged as a response to the first pattern. Separate Category 5
cables for data and Category 3 cables for voice were placed into bothrenovated and new
buildings. Sufficient pairs were thus provided to alleviate or eliminate the competition
for spares and to divide the labeling and documentation responsibilities into two groups
of users.
The second pattern now reflects the campus standards practice. Both Category 3 and
Category 5 cables are placed in the risers.
DCC
