The relationship between individual-component compliance with performance specifications and overall channel performance is worth a close examination.
When it comes to structured cabling performance, the objectives of a standard are both to ensure the performance of present, foreseen, and perhaps unforeseen applications, and to allow interoperability of components from multiple manufacturers that make up a structured cabling system.
Ensuring the performance of applications has rarely been an issue, because the requirements defined in TIA-568-B.2 and ISO-11801 are in excess of what typical applications (such as Ethernet) require, as developed in IEEE 802.3an. For example, the fact that some Category 6 cabling can support 10GBase-T is due in part to the extra performance margin the Telecommunications Industry Association (TIA; www.tiaonline.org) built into the Category 6 standard.
Manufacturers often design their structured cabling solutions to have the best margins when their components are combined as a system in an installation. This approach takes advantage of tradeoffs that can be made when the performance envelope of each component is known. The envelope is typically nar- rower than what the standard requires for each component, if the design is robust and manufacturing variables are well controlled. Connectors can be designed to work best with a given cable and/or patch cord so as to optimize channel performance.
When a structured cabling system has components from different manufacturers, the overall performance margin can drop due to the need to work with connectivity that may be anywhere in the performance envelope the standard defines for a permanent link or channel. To combat this, many manufacturers design extra margin into a component-rated connector, patch cord, or cable. The result is excellent performance margin, both with a manufacturer's own components and with a mix of brands if the end user chooses to go that route.
With the advent of 10GBase-T, alien crosstalk requirements are defined in TIA-568-B.2-10 at the channel, permanent-link, and component levels. In some instances, meeting component requirements for alien crosstalk for each part of the system—cable, connector, or cord—does not guarantee that the resulting channel or permanent link will meet alien crosstalk limits. The sum of the parts does not ensure the whole, as is normally true for internal parameters like near-end crosstalk and return loss.
The standard acknowledges this in Annex B, Section B.1 as well as Annex C, Section C.1, stating that in addition to component performance verification, cable and connectors should be verified for compliant channel alien-crosstalk performance. This is because alien crosstalk can take many forms, not only differential-to-differential, but also common-mode to differential-mode conversion. These different manifestations of alien crosstalk are affected by the balance of the component. The better the electrical balance of a cable, connector, or cord, the better it is able to reject these common-mode conversion forms of alien crosstalk, and the better alien-crosstalk margin a channel has when made up of these components.
Components that are designed and tested to exceed cur- rently published alien-crosstalk component requirements will not only work well in a mixed channel, but can attain maximum performance in a single-brand end-to-end structured cabling system.
Testing component compliance
Component alien-crosstalk testing for connecting hardware involves two setups: one for alien near-end crosstalk (ANEXT) and one for alien far-end crosstalk (AFEXT). For a given panel or wallplate, a victim port is selected, and the crosstalk from each disturbing port is measured. The results are summed to give a worst-case power-sum alien near-end crosstalk (PSANEXT), or power-sum far-end crosstalk (PSAFEXT).
To determine the disturbance level of a single port to a victim port, 16 pair-to-pair alien-crosstalk measurements are necessary. Results of one such suite of 16 measurements are shown here.
The TIA standard defines limits for connecting hardware alien crosstalk, as well as a limit for how many disturbers might be included in the summing of crosstalk affecting the victim port. These “significant disturbers” are those ports that exceed the 90-20log(f/100) limit for alien crosstalk when measured from the victim port. In a typical installation, a significant disturbing link can be nearly any port in a panel, depending on how the cabling is bundled, whereas significant ports in a panel are generally located near the victim port.
Shown here are component PSANEXT test results from a Leviton Category 6A patch panel, which provides performance margin beyond the TIA standard's limit line.
The opening figure illustrates how to measure the alien crosstalk seen by the victim jack from neighboring ports (disturbers). This crosstalk can come from the connectors, the traces on the printed wiring boards (PWBs), and the insulation-displacement connections (IDCs) associated with each port.
The same figure shows the results of measuring several ports surrounding the victim port for significant disturbers. Depending on the panel configuration, approximately 24 to 30 ports must be checked for significant ANEXT or AFEXT relative to a selected victim port. This process is time- consuming, because ports are tested moving outward from the victim port in all directions until non-significant disturbers are found. One victim/disturber port combination is tested at a time, using a vector network analyzer to apply a signal to each disturber wire pair, listening for any coupling of that signal to any of the four victim port wire pairs.
In conducting research for this paper, the authors assembled a 100-meter channel using their company's connecting hardware and approved Category 6A cable, as shown in the topology illustrated here.
The graph in “Pair-to-pair ANEXT” shows results from 16 pair-to-pair alien-crosstalk measurements that were taken to determine the significance of just one port to the victim port. This example shows the port to be significant, because the alien crosstalk measured exceeds the significance limit set by the TIA standard. As such, it will be included in the summing of all significant disturbers to the victim port. Only those pair combinations that exceed the significance limit will be selected for the calculation, and only at the frequencies at which they exceed the limit.
Notice that, for a disturber port located adjacent to the victim port, this disturber is remarkably quiet. In this case, careful attention to port isolation at the connector, PWB, and IDC levels ensures this patch panel will have margin to spare. Once significant disturbers are known, one can calculate the alien crosstalk the victim port sees, and compare this against the component requirements of the standard. In this test setup, only six disturbers were found to have significant ANEXT to include in the power-sum calculation, though 24 ports were checked.
The figure “Component PSANEXT” (page 26) shows results for the Category 6A patch panel recently verified by Intertek (ETL) to meet the component requirements of TIA-568-B.2-10. Even when surrounded by several disturbing ports, the typical victim port in this patch panel has plenty of alien- crosstalk margin relative to the standard.
This process is repeated for determining the amount of PSAFEXT seen by the victim port. Significant disturbers are located using AFEXT measurement techniques, and the results are power-summed.
Connectors like these that exceed TIA requirements can be installed into Category 6A permanent links and channels with exceptional results. When using cords and cable that have been designed and optimized to work together, the system's alien-crosstalk margin is maximized.
What is the installed performance of a Category 6A structured cabling system that is made up of component-compliant jacks, panels, cords, and cable? The components tested for this paper—Category 6A component-rated connecting hardware, including two patch panels, a 110 block, a connector, and patch cords from Leviton as well as approved Category 6A cable—were assembled to complete a full 100-meter channel. The installed channel was tested using a calibrated instrument with a very quiet noise floor, which is needed to accurately measure power-sum alien attenuation-to-crosstalk ratio at the far end (PSAACR-F). For both PSAACR-F as well as PSANEXT, the channel showed performance margin beyond the requirements of the TIA-568-B.2-10 limits.
The figure “4-connector, 6-around-1 topology” shows that excellent margins are possible when components and cable are designed to work with each other.
Alien crosstalk is one of the more challenging parameters to mitigate, with several coupling modes to deal with. Minimizing near- and far-end alien crosstalk in the channel involves addressing each of these modes of coupling in each component, as well as looking at how and where along the frequency of interest each component contributes to the measured channel performance.
Connection in tune
When combining standard-compliant connecting hardware that meets Category 6A component specifications with compliant horizontal cable, the result should be an optimized system. This system will typically meet the channel and permanent-link performance requirements; however, to assure the highest channel margin, both for internal parameters and the alien-crosstalk parameters explored here, it is critical that the cable and connecting hardware be tuned together.
When putting together a system of Category 6A connectivity and cable, ensure that applications such as 10GBase-T are supported well, with margin to spare. A component-rated panel, jack, cord, or cable is only as good as the channel performance the combination is able to achieve.
GARY BERNSTEIN, RCDD, is director of product management for copper systems, and JEFF SEEFRIED is development engineering manager with Leviton Network Solutions (www.leviton.com/networksolutions). This article is based on a white paper that can be found on the company's Web site.