With performance specs for structured wiring systems increasing, proper test procedures are more important than ever
Copper cabling failures in high-performance structured wiring systems are most commonly caused by one of a small list of problems. Inadequate component performance from something like a bad cable or cable segment, a connector that doesn’t meet specifications, or a poorly performing patch cord can bring data transfer to a halt. However, the problem oftentimes involves human error. Poor workmanship at installation, including something as simple as a mistake with the wiremap or failure to maintain the original twist in the wire pairs, can be just as harmful and make failure inevitable. In addition, incorrect test procedures at installation or during troubleshooting can cause cabling installation professionals to misidentify the problem, or worse, think there isn’t one.
Further complicating matters, the performance specifications for new structured cabling systems have dramatically increased in recent years in anticipation of higher bandwidth requirements. These new high-performance cabling systems will have to support faster data communication rates and more precise timing of voice and video delivery.
As a result, these systems must be tested and certified in the field with new test parameters, and new link definitions, more data points, higher bandwidth, new connector types, and revised patch cord requirements make this a challenging task. Due to their increased sophistication, determining the cause of failures and quickly restoring suitable performance in these systems is your ticket to success. But proper troubleshooting techniques also require the right test equipment. This article focuses on three key areas of the certification process:
- The link model – choosing the proper test setup and eliminating a common source of measurement error.
- The measurements – what’s tested and reported during certification.
- The diagnostics – when errors are found, how can they be fixed as rapidly as possible?
Link models. The term “link model” refers to configuration of the cabling link under test and the manner in which the test equipment is connected to the link. Most certification tests in the field should use the permanent link model, which consists of the cable and the permanent connectors in the wall outlet or patch panel.
The connection from the test instrument to the permanent link is via a permanent link adapter (Fig. 1). In a practical sense, this means field testers must be much more sophisticated because they must be able to measure the performance of the link without any of the effects of the test cords with which they’re connected to the link.
The reference point for the measurements isn’t at the test instrument but at the far end of the test cord. From an installer’s perspective, this means that if the effect of the adapter isn’t completely transparent, a good link can produce a failing test result. The likelihood of a false failure increases with performance requirements. In other words, Cat. 6 cables are more likely to produce false failures. Therefore, when performing permanent link tests, make sure you’re using adapters appropriate for the cabling under test (especially for Cat. 6/Class E links).
The channel link model defines the performance of the completed end-to-end link as the operating network uses it. The channel test involves everything in the permanent link test, plus the patch cords in the work area, and those used for interconnection and/or cross connection (Fig. 2). Therefore, the patch cords used to connect the PC or the network device must remain in the link and plugged into the channel adapter of the test instrument.
Channel measurements are typically made when restoring service or to verify cabling for application support. Channel tests aren’t commonly performed during initial installation, since the patch cords are rarely available at that time. Correct channel measurements must cancel the effects of the mated connection in the tester’s channel adapters. It’s important to use the correct link model standard with the appropriate tester interface hardware.
Certification measurements. Link certification requires a number of complex measurements. These measurements change in degree of precision depending on the category of the installed link. For Cat. 5e and Cat. 6 links, a certification tool needs to measure or calculate the following:
- Wire map
- Propagation delay (delay skew and link length)
- Insertion loss
- NEXT (near end crosstalk)
- Power Sum NEXT
- ELFXT (equal level far end crosstalk)
- Power Sum ELFEXT
- Return loss
The “accuracy level” of the certification tool determines its ability to make these measurements at an appropriate level for the type of link. For Cat. 5e installations, Level IIe defines the minimum required accuracy. For Cat. 6, a certification tool needs to meet Level III accuracy. Level IV defines a higher level of accuracy over a wider frequency range to certify the ISO Class F (Cat. 7) links. Cat. 7 refers to a system of cable, connecting hardware, and patch cords that are rated to a maximum frequency of 600 MHz. To achieve this performance, a shielded cable construction is necessary.
The Table shows an example of the allowable residual NEXT characteristic of the field tester as defined in Level III versus Level IV. Residual NEXT measures the amount of crosstalk inherent in the tester itself without any cable present on its input. For the baseline and permanent link specifications, a Level IV tester must exhibit a worst case value for residual NEXT at 100 MHz that’s 18 times smaller than the smallest signal to be measured by a Level III tester.
The certification tool must perform the measurements of all the test parameters listed above over the prescribed frequency range for the category of cabling and with the number of frequency points defined in the standards. On some testers, you can view the test results immediately or they can be stored in full detail to be inspected later on a personal computer.
Photos 1, 2 and 3 show a summary screen from a Cat. 6 test. Note the failed tests for length, propagation delay, insertion loss, and NEXT in Photo 1. The tester also shows detailed results from the NEXT tests across all pairs, and the values-by-frequency for a single pair.
Using advanced troubleshooting diagnostics. Knowing that a test failed is only the first step, since the link must be fixed so it will perform as intended. The reasons for failing certification tests fall into two distinct categories: connection problems and transmission performance problems.
Many tools can provide information regarding the connection problems like opens, breaks, and shorts, but you should select one that can properly locate a break or a short in the cabling as well as identify problems caused by improper wire pairings.
In addition, certification testers should include advanced troubleshooting diagnostics that identify and locate transmission defects. With this diagnostic information, you can dramatically improve troubleshooting productivity and help restore service quickly.
On some testers, you can view the diagnostic information related to the test that failed (Photos 4 and 5). The tester analyzes the performance of the crosstalk parameter NEXT over the length of the cable and identifies the location in feet or meters from the end at which a cabling or installation defect has caused a failing measurement.
You can also drill down one level to inspect a representation of the NEXT performance along the link.
With additional training, you can review the “raw” data and uncover more details about the premise cable link, which can reduce the time needed to make the correction.
Structured cabling is a growing opportunity for the professional installer. Testing and certifying is an essential part of the job.
Not only that, a certification report is your guarantee that you left the job in good working order. It lets the customer know that the installation will perform at levels necessary to support their needs, now and in the future.
Draye is marketing manager - certification for Fluke Networks, Infrastructure SuperVision group, Everett, Wash.