If my last column lambasting the standards process that is bringing the industry 40-Gbit Ethernet was “yin,” then this time around I’ll try a little “yang” by praising some of the work that has gone into the development of a different standard.
The Telecommunications Industry Association’s TR-42.7 Copper Cabling Systems Subcommittee has been working for quite some time on the specifications for Augmented Category 6 twisted-pair cabling systems. Cat 6A systems, as they’re called, are designed to support 10 Gigabit Ethernet in its twisted-pair flavor, 10GBase-T.
Structured cabling standards are funny because they are intended at least as much for cabling-system manufacturers as they are for those systems’ users. The Category 6A specs will be published as Addendum 10 to the TIA/EIA-568-B.2, and will include electrical requirements that each component-cable and connectivity-as well as the overall system must achieve in order to comply.
Valerie Rybinski, chair of TR-42.7 and a past contributor to Connector Specifier, reports that the standard is “in great shape. The transmission numbers have been firm for a long time; they haven’t changed in two years.” So, why wasn’t the standard published two years ago? It’s those pesky component requirements. 10GBase-T is going to transmit at frequencies ranging all the way up to 500 MHz. Never before had a “Base-T” protocol transmitted at frequencies higher than 100 MHz. While Category 6 systems (not Augmented 6 ... stay with me here) had to achieve 200-MHz useable bandwidth, never before had the proverbial rubber met the clichéd road to support real traffic at a frequency higher than the 100 MHz at which 100-Mbit/sec Fast Ethernet and 1-Gbit/sec Gigabit Ethernet topped out.
Now, in order to comply with Cat 6A, manufacturers have to test their components-and in particular their connectors-at 500 MHz.
“What has been challenging for us is how to make very tiny measurements,” Rybinski continued. Understand, she was talking to me. I had asked her if she could dumb it down so much that even I could understand it. We both laughed because we knew that was impossible, but she made it as elementary as it could be. At the risk of misquoting her badly, I’ll summarize my understanding of what she explained. The group has been working without the benefit of ASTM tests to guide them at frequencies as high as 500 MHz. With their existing technology and techniques, they could make measurements within a certain inaccuracy level-say 1 dB. But the group has been working for about 18 months to reduce that inaccuracy level as much as possible.
The trick has been to develop fixturing that provides impedance control, good repeatability, and the ability to isolate the test fixtures from the measurement. Historically, during the termination of test leads, the characteristics of those leads would change, due at least in part to movement. New fixturing was required to keep the test leads precisely located.
They believe they have that fixturing in place now, thanks in large part to the work of TR-42.7’s co-chair, Sterling Vaden, who very recently reported positive results from some early testing using the revised fixtures.
My fingers are crossed that, in a future issue, we’ll be able to provide you more significant detail about the fixture being used to help verify the compliance of Category 6A twisted-pair connectors. Until then, I’ll send kudos to the engineers who ventured into the previously uncharted territory of 500-MHz testing of twisted-pair components. It looks like their work will be complete by the end of this year.
Patrick McLaughlin
Editor-in-Chief
patrick@pennwell.com
