Learning the CAT 6A ROPES
As transmission speeds increase, there is far less margin for error. Any shoddy workmanship or oversights can have significant repercussions on network performance.
March 1, 2008
Print this page
As we witness increasing interest in 10 gigabit (10 GBASE-T)-speed networks, all eyes are upon Category 6 and Augmented Category 6 (known as 6A) as the cabling systems of choice.
The emergence of 10 gigabit is the latest challenge facing cabling and networking developers when it comes to installation and testing procedures. It only stands to reason that these higher speeds come with additional specification requirements.
The main purpose of Category 6A cabling is to support the IEEE 802.3an specifications for 10 gigabit per second transmission over twisted pair copper to channel lengths up to 100 metres.
As a totally new cabling, Category 6A comes with specific design and performance criteria, and does have significant differences from Category 6 cabling in certain areas.
However, the challenge for installers is that this new set of specifications have yet to be clearly defined.
The IEEE 802.3an was completed and approved by the IEEE standards board in June 2006, making 10GBASE-T a completed and released standard.
In addition, Telecommunications Industry Association specifications for Category 6A have been approved.
Those who have worked extensively with Category 6 and have become familiar with its properties already have a practical working knowledge of testing and installation requirements. When working with 6A, however, experience has shown that there are some essential points to consider.
Talking the crosstalk: Every cabling technician has had to deal with the issue of controlling signal coupling between cabling links. It is something that must be considered in everything from cable and connecting hardware selection, to design and installation practices.
The proper choices, will in turn, ensure that the transmission quality of the installed cabling system will meet certification requirements during testing.
A new requirement for 10 GBASE-T installations is testing for Alien Crosstalk, a phenomenon unique to these types of installations.
This refers to the signal coupling between cables within a bundle of twisted-pair cabling links that can result in a noise disturbance.
Crosstalk in its most common forms — Near End Crosstalk (NEXT) and Far End Crosstalk (FEXT) — is not an unfamiliar issue for experienced cabling technicians. Digital signaling processors used in 10 GBASE-T installations are designed to assist the Network Interface Card (NIC) in mitigating problems such as crosstalk.
However, this process can lead to extra latency as the NIC performs the job of “cleaning up” the signal from the transmitter on the other end of the cable.
An additional point to consider is that the10 GBASE-T interface card has no way of mitigating Alien Crosstalk. Alien Crosstalk is an additional concern that can further attenuate an intended signal and put extra work on the NIC when cleaning up the signal. Therefore, technicians must rely on testing the cable itself in order to understand how different transmission activities will have an impact on the NIC’s performance.
Compared to other types of cabling technologies, Category 6A requires a bandwidth of 500 MHz to support 10 GBASE-T. With a bandwidth frequency this high, signals are emitted both within the cable as well as to adjacent cables; hence the need for additional testing requirements.
In addition, while crosstalk can normally be mitigated by twisting wire pairs at different twist ratios this is not the case with Category 6A cabling. In most scenarios, the cabling bundles are installed from the same box and could contain pairs with the same twist ratios.
As with all cabling, in channel testing will help to first determine if transmission impairments (e. g. NEXT) meet Category 6A standards. However, Alien Crosstalk testing demands more.
Testing for Alien Crosstalk: The challenge for technicians is that current standards do not outline any fixed rules on Alien Crosstalk testing. However, the IEEE and TIA do recognize that Alien Crosstalk is an impair-ment that will affect transmission, and as such, describe field-testing procedures under Annex E Field measurements procedures.
In addition to the standards debate, there is also considerable discussion over testing procedures. In particular, when it comes to 10 GBASET, does one need to test an entire installation?
While the answer is not clearly defined, there are some well-established and proven guidelines one can follow. At this point in time, it is best to adopt the guidelines published in the IEC 61935-1 standard document. This advises that Alien Crosstalk performance be tested in the field for a limited sample of cabling links.
In sampling, technicians select the longest cables or the cables with the most insertion loss. The premise is that if the worst performing cables pass Alien Crosstalk testing, it can be assumed that the others will.
The guidelines suggest sampling a sampling rate of 1% or 5 links, whichever is greater. Therefore if a system has 500 links, then 5 can be selected as the disturbed or victim links. All links in the same bundle as the victim link should be included in the testing. In addition, all adjacent links to the victim link terminating in the patch panel should also be included in the test.
Installation considerations for 10 G BASE-T: As with testing, guidelines on proper installation procedures for 10 G BASET are in short supply. Adjustments must be made to address the specific properties of the cabling and its performance.
In some cases, following standard procedures used for other cabling can potentially have a negative impact on network performance. Even something as basic as organization and bundling must be reconsidered. Any kind of poor workmanship will have more impact on 10 Gig cabling than on cable designed for slower transmission rates.
Working with high frequencies and signaling systems means that installers must pay more attention to their cable handling and layout practices.
Whereas they may have typically bundled cables as tightly as possible to optimize space, this works against cabling for 10 GBASE-T installations since it increases the risk of Alien Crosstalk. More space is needed between cables, both in terms of panel density and the size and routing of the bundles.
In cases where installers are using existing patch panels, they must run fewer links in a bundle to allow for added space. New Category 6A patch panels have been designed to offer the appropriate separation between jacks.
Another option to reduce potential problems is to switch from a cross-connect wiring scheme (i. e. using two separate patch panels for incoming and outgoing connections) to an interconnect scheme. This eliminates one connection and reduces crowding in wiring racks.
Cables should also be well organized, documented and loosely bundled to allow for free space between them. Otherwise the installation will not meet testing requirements. Suggestions include tying bundles every 61 centimetres (versus every few millimetres) or laying them in trays without ties. Avoid using wraps, staples, tacks or fasteners where possible.
Some 6A cabling for 10 G BASE-T installations can be heavier and have a larger outside diameter than standard cabling.
This means taking care that there are sufficient routing pathways and ladders. In addition, cable lengths should be kept to a 60-metre maximum length; and should not be run in parallel with nor adjacent to electrical cabling. Installers should also avoid using J-hooks or other like devices for hanging cabling because of its added weight.
Learning the ropes: Even the most experienced installers will find that working with Category 6A will present some challenges at the outset. While basic practices will remain, there are a number of considerations from additional testing requirements to handling and installation that must be addressed in order to ensure optimum results.
As transmission speeds increase, there is far
less margin for error. Any shoddy workmanship or oversights can have significant repercussions on network performance.
Increased bandwidth, throughput and capacity demands have fuelled global interest in migration to 10 GBASE-T.
The publication of new standards is expected to accelerate this migration even more in the months to come. Installers would be well advised to take the time to get up to speed on guidelines and best practices for 6A testing and installation today.
Chris Pawelko is Canadian Channel Product Manager with Fluke Networks Canada. He has 18 years industry experience in data communications, networks and cabling.
Increased bandwidth, throughput and capacity demands have fuelled global interest in migration to 10 GBASE-T. The publication of new standards is expected to accelerate this migration even more in the months to come.