Powersum Alien Crosstalk:
The Truth is Out There
Introduction
This paper is intended to qualify the alien crosstalk performance of MediaTwistâ in a realistic installation environment. It briefly outlines the need for performance qualification, describes the test setup, and presents the results. It is our intent to dispel any false beliefs or inaccurate information concerning the performance of Belden's patented design. All performance comparisons are made to the July 1998 draft TIA/EIA Addendum 3 of the ANSI TIA/EIA-568-A document concerning bundled cable requirements. At this time, the requirement for bundled Category 5 cables is powersum NEXT, 3dB more stringent than pair to pair NEXT. Where appropriate, the specification was extrapolated to 350 MHz.
Background and Definitions
With the development of 1000BaseT, otherwise known as Gigabit Ethernet, the days of low frequency system demands are fading behind us. As an example, 10BaseT only utilized two pairs to transmit data. At a network interface card or port, one pair was a dedicated transmitter while the other was a dedicated receiver. Additionally, the transmitter was silent while the receiver was active and conversely, the receiver was not looking for incoming data while the transmitter was broadcasting. This helped isolate the data from noise and ensure data integrity. This is very similar to a phone call with two well mannered, slow speaking people conducting a conversation.
In order to meet the desired goal of 1GB of data throughput and minimize the frequency bandwidth, Gigabit Ethernet utilizes four pairs to transmit and receive data. Now each pair must transmit and receive data, at the same time. Now all pairs are contributing to the total NEXT noise in a cable. Additionally, adjacent cables can contribute noise. This has been coined "alien crosstalk" and, simply put, is data interference from adjacent cables. Currently, there is no set standard for this particular issue. However, there is a draft Addendum to the 568A document in process at the time of this publication.
Test Parameters
The goal of these experiments was to accurately determine the powersum alien crosstalk of realistic, standards compliant installations. Two different, 100 meter, cable orientations of MediaTwist were tested for alien crosstalk performance. The first, labeled "Typical Installation" was constructed in compliance to TIA/EIA 569A installation requirements. Six different MediaTwist cables were pulled and cinched with a tie wrap every 48 inches. The tension on the tie wrap kept the cables nested together but did not deform the jacket. The first tie wrap was two feet from the test station. The following figures represent two typical cross-sectional areas at the tie wraps.
 |
 |
Typical Cable Orientations
The second configuration, labeled "Meticulous Installation" was constructed in compliance to TIA/EIA 569A installation requirements with the exception of the last twenty feet. In order to simulate the meticulous cable management at the point of breakout in the Telecommunication Closet, 20 feet of the bundled cables were cinched every three inches. The first tie wrap was three inches from the test station.
Results
The first thing to note is that the specification for this particular parameter is in draft status. The current draft requirement is specified as 3 dB better than the specified pair-to-pair NEXT loss of the recognized cables (in this case Category 5).
A) Typical Installation
Powersum Alien NEXT performance was calculated and the cable bundle, which
represented the typical installation, exhibited an average worst case margin of 8.4 dB to
the bundled Category 5 draft specification. The absolute worst case margin was over 4.6
dB.
Typical Performance 6 cables (24 Pairs) Powersum
B) Meticulous Typical Install
Powersum Alien NEXT performance was calculated and the cable bundle, which
represented the extreme installation, exhibited an average margin 8.3 dB to the bundled
Category 5 draft specification. The worst case margin was over 3.9 dB.
Conclusion
When installation practices are compliant to the present standard ANSI TIA/EIA-569A 4.6.5.1, Belden's MediaTwist provides excellent alien crosstalk isolation. Six cable bundles exhibited almost 4 dB of margin in the worst-case industry standard compliant installation scenario. However, non-compliant installation practices, such as spiraling electrical tape to bundle cables, can produce marginal performance. We strongly recommend the installer follow endorsed industry installation practices to achieve optimum alien crosstalk performance.