Categorical confusion

With the copper-cabling market flooded with a sea of new adjectives such as New Category 5, Enhanced Category 5, Power-sum Category 5, Category 6, Category 6a or b, and various forms of GigaTHIS and GigaTHAT, are you confused? You are not alone. Recently, in the course of polite conversation over lunch, I was asked these questions and thought you might want to knowellipse.

What's up with Category 6? Category 6 what? Link? Channel? Let's start with what is standard: There is not an approved industry standard for Category 6 cabling. Not for connecting hardware, links, or channels.

ISO/IEC JTC1/WG 3 is also developing ISO/IEC-11801 second edition, a document defining specifications for Category 6 and 7 components and Class E and F cabling systems.

The TIA's TR-42.7 UTP Systems Subcommittee is also developing the Category 6 specification, "Transmission Performance Specifications for 4-pair 100-ohm Category 6 Cabling," which is currently in Draft 5. Their goal is to keep the Category 6 channel limits harmonized with the second edition of ISO/IEC-11801.

ISO appears to be ahead of TIA with ratification of ISO/IEC-11801 second edition expected in the first quarter of 2001.

Who came up with the "Category" system, and why did they make so many different versions of Category 5 before moving to Category 6?

Several years ago, a group of far-thinking folks in the telecommunications cabling industry (TR-41.8.1) got together and agreed (after much quibbling) on a set of transmission parameters that would produce a standard for the most advanced unshielded twisted-pair (UTP) cabling system ever imagined. Then, they cleverly named it Category 5. This ultimate in UTP cabling was not an only child, it also had lesser-performing siblings-Category 3 and 4. I was in the room in 1992 when the discussion took place as to whether to make the "best cabling" the largest or the smallest number. Several marketing types decided, telecommunications systems bulletins TSB-36 and TSB-40 were published, and the rest has been a wild ride.

Did you ever ask yourself, with numbers now available from 6 to infinity, why we now see Category 5 with new delay and delay-skew parameters? Then there is the TSB-95 version of Category 5 concerning return loss, and the Enhanced Category 5 version (also concerning return loss, but with better numbers), which got a different name but still not a new number-all because Category 6 has been on the drawing board since ANSI/TIA/EIA-568A was published in October 1995. While little technical work has been done within the TIA TR-42 subcommittees since then, that's not true for the individual companies whose engineers participate in these committees. Pick up any industry publication and you will see that every cable and connector manufacturer has a Category 6 product line, and several distributors have programs to tell you if the cable and connector manufacturers' products meet the distributor's specifications for Category 6 performance.

Do the standards require testing or not?

I remember standing toe-to-toe in heated discussion with the former chair of TR-41.8 when he made the profound statement: "It doesn't matter if it works or not. If each of the parts is Cat 5, then the channel is Cat 5 according to the standard." Technically, he was correct. The scary thing is, he still is! There is still no requirement for link or channel testing within the TIA standards.

Terminating 4-pair cable to 8-pin connector

Q: We are in the process of designing a new campus to house an estimated 10,000 employees. In the cabling infrastructure design report that we received from the consulting firm, they recommend that we install three Level 7 cables to each workstation, which will terminate onto RJ-45 patch panels in the telecommunications rooms and then to RJ-45 jacks at the workstation. Since most private branch exchange vendors' telephones come with an RJ-11 plug, is this a standard solution or should we consider adding an RJ-11 socket to each workstation to accommodate the telephone? (In the past, I have plugged an RJ-11 plug into an RJ-45 jack, but it doesn't seem to be the best fit.)
John Freeman
Sabre

A: Yes, terminating a 4-pair cable to an 8-pin modular connector is standard in ANSI/TIA/EIA-568A. You are right: It does not always seem to be a snug fit. However, the interface will support your telephone connections today and whatever is on the horizon for that cable in the future. I recently had a similar discussion with a telephone installation foreman, who wanted the cable installers to split each 4-pair cable to be used for voice into two 2-pair terminations each on a separate RJ-11 jack. That was the day after I had received an e-mail from IEEE 802.3 asking for help in determining what percentage of the installed base of cabling did not have all four pairs terminated in a single 8-pin modular jack.

It seems that IEEE 802.3 is looking into supplying modest amounts of power for small devices such as an IP-telephone via 10/100 Ethernet. But do they have to do this on the signal pairs for 10/100 (1-2, 3-6) or can they reasonably assume that the spare pairs (4-5, 7-8) are actually spare?

In a campus environment such as you have described, the use of IP-telephone is very likely in the near future. Stick with the cabling standard, if for no other reason than it is what the applications engineers are designing, too.

For more information on delivering data-terminating-equipment power via a media-dependent interface, see grouper.ieee.org/groups/802/3/index.html.

Drain wire termination

Q: Your answer will settle an ongoing argument regarding a cabling-installation project on a ship scheduled for this summer in Cadiz, Spain.

When terminating shielded twisted-pair cable (Belden Datatwist 100 1730A) from a Krone 16-way RJ-K LN patch panel mounted on a wall rack to Krone RJ-K LN RJ-45 socket outlets, is it necessary to terminate the drain wire at both ends, or will terminating only on the patch-panel terminal post fulfill requirements?
John Reidy
Stena Line
Ashford, Kent, UK

A: According to TIA/EIA IS-729, Technical Specifications for 100-ohm Screened Twisted-pair Cabling, effective shielding requires that all cabling components, including the connectors, work-area cords, cable, crossconnect cords, equipment cords, and the equipment connection, be shielded; that all shields be properly bonded to each other; and that the shield be bonded to the telecommunications grounding busbar in the telecommunications room.

A few additional points to ponder: Belden 1730A is a CM (nonplenum)-rated, Category 5 screened twisted-pair cable, which Belden indicates is "not RJ-45 compatible." For more information, see www.belden.com/products/premcode.htm.

Krone indicates that the RJ-K LN Series, which is available only in Chile and Brazil, is designed around a high- performance, unshielded RJ-45 jack suitable for use with unshielded Category 5 systems. For more info, see www.krone.co.uk/.

I recommend that you seek professional assistance with the cabling design. For suggestions on whom to contact in your area, e-mail BICSI- [email protected].

RCDDs should be regulated

In your February 2000 column, you stated that Texas and New Jersey require that telecommunications design work be performed only by registered professional engineers (PEs). The State of Florida has the same requirement: Florida Statutes FS 471.008, 471.033(2) 471.033, and Chapter 61G15-33.005 apply.

Throughout the country, there has been an "industry exemption," whereby a person working in a company can design products and systems. How ever, when the public is involved, such as in school work, the PE is required by law. Local utility companies in Florida have been forced to remove the title "engineer" from all job positions not held by a registered PE. They are now given titles such as "technician" or "planner."

BICSI needs to petition the state of Florida to have the registered communications distribution designer (RCDD) status recognized and regulated, just as the state regulates and licenses barbers, hair stylists, doctors, and, yes, engineers.
Curt Wasko
BICSI member and PE
Long & Associates, Engineers/Architects Inc.
Tampa, FL

Click here to enlarge image

Donna Ballast is a communications analyst at the University of Texas at Austin and a BICSI reg-istered communications distribution designer (rcdd). Questions can be sent to her at Cabling Installation & Maintenance or at PO Drawer 7580, the University of Texas, Austin, TX 78713; tel: (512) 471-0112, fax: (512) 471-8883, e-mail: [email protected].