So, you think you know cable testing?
Many standards and industry practices have changed over the last few years. Unless you've been paying close attention, you might be surprised to discover that some things you took for granted are no longer true.
Here's a quiz to test your cabling testing savvy. Beware—some of the questions are tricky.
- Cabling contractors should certify new structured cabling to the latest TIA/EIA-568A requirements.
- To ensure best performance, all installations should be fully certified to meet TIA/EIA-568B and ISO 11801 requirements.
- When testing the basic link, always remember:
- To calibrate your test set adapters once a day.
- NEXT problems can be caused by incorrect twist ratios.
- The maximum allowable length is 94 meters.
- Marginal results should be reported with a star (asterisk).
- When testing to Category 5 requirements, always remember:
- Results should be stored in TIA/EIA-606A-compatible format.
- Saving summary results is sufficient (graphs are not required).
- ACR and PSACR are optional measurements.
- Category 5 certification is not sufficient to ensure Gigabit Ethernet operation.
- A channel has a maximum TIA/EIA-568B allowable length of 100 meters. If the four pair-lengths reported are 101, 99, 103, and 102 meters, the cable:
- Marginal test reports (those within the accuracy of the test instrument to the limit line) should be reported with a star (asterisk) to indicate the result is uncertain.
- I have a Category 6-rated tester. This means I can certify any Category 6 installation.
- I certify installations to meet Category 5, 5e or 6, depending upon the customer's requirements.
- Permanent links always include the:
- CP (consolidation point).
- TO (telecommunications outlet).
- User patch cords.
- Test equipment adapter.
- Patch panel.
- Category 6 has been approved.
- What is the easiest way to diagnose and fix NEXT failures?
- Replace connecting hardware.
- Re-terminate all connectors, checking twist ratios.
- Use test equipment that can locate failures in the time domain.
- Call the cable supplier tech support line.
- The most common cause for channel performance problems:
- Poor quality patch cords.
- Mismatched components.
- Installation workmanship issues.
- Adjacent noise sources.
- False. TIA/EIA-568A was superceded in April 2001 by TIA/EIA-568B, which includes a great deal of new and changed content, including new and significant field testing requirements. If you're testing or specifying TIA/EIA-568A, you're referring to an obsolete document.
- False. We see this error often. It is impossible to meet both TIA/EIA-568B and ISO 11801 requirements, because they have several conflicting specifications. You must choose one or the other. As an example, ISO 11801 has a "4 dB rule." This essentially states if measured insertion loss on the link is less than 4 dB, you are required to pass NEXT regardless of the measured NEXT values. So, for a short link with high NEXT, the TIA/EIA-568B test would fail, while the ISO 11801 test would pass. You could not possibly comply with both. While the documents are generally in harmony, there are several such incompatibilities that make it impossible to meet both at the same time.
- Trick question. With the ratification of TIA/EIA-568B, there is no longer a basic link. The question is meaningless because according to today's standards, you must test to the permanent link or channel. Don't test to basic link requirements—they no longer exist. And if you're an end user, don't accept basic link test results.
- Trick question again. Answers A through D were all true when Category 5 was a relevant standard, but it has been replaced by Category 5e. Category 5 became obsolete and replaced by Category 5e in April 2001 when TIA/EIA-568B was published. Testing to Category 5 is as meaningless as testing to Category 4. Don't test to Category 5 requirements—they no longer exist. And if you're an end user, don't accept Category 5 test results.
- It passed. TIA/EIA-568B dictates that the length of the cable is defined as the length of its shortest pair. So, if three of four pairs are over 100 meters, but one is 100 meters or less, then the length passes.
- False. Marginal results shall be reported. Not should. Standards require honest reporting oftest results, including when the results are too close to call with certainty. Unfortunately, some tester manufacturers include the facility to turn off marginal reporting (you'll never see an asterisk). Insist that any test results you use only come from testers with "permanently enabled" marginal reporting.
An installer tests a Category 6 permanent link.
- It depends. To get accurate, acceptable Category 6 results:
- The Category 6 supplier must have approved the tester (some "Cat 6" testers are not approved by all suppliers).
- You need to use the correct/compatible Category 6 adapter (approved by the supplier for permanent link tests).
- You need to ensure you have up-to-date software. Limit lines and test rules have changed over time. Testing with old software on an approved tester could still cause failures on good links or passes on bad links.
- False. By now, you should know that Category 5 is no longer approved for horizontal structured cabling tests under any circumstances.
- B. The TO is always included. The CP is included only if it is present. For more details on permanent links, see http://www.cabletesting.com/CableTesting/Standards/LinkDefinitions/Permanent+Link.htm
- True. It was approved as TIA/EIA-568B.2-1 in June.
- C. Today's DSP-based field testers can see into the time domain and show crosstalk vs. length, so it is very easy to see where in the cabling any sources of high NEXT may be present. Once you know the location of the NEXT, you can take appropriate steps to correct the fault. Automatically replacing or re-terminating connectors might work sometimes but is guesswork at best, and often wastes time and resources.
- A. Patch cords are often given little thought, and purchased later at lowest cost from overseas manufacturers who do nothing more than wiremap checks (despite the familiar-looking "certified" logos on the cable sheath). Channel performance and bit error rates can be severely impacted by poor quality cords. In fact, with the correct patch cord adapters (not channel adapters), you can now perform TIA-compliant patch cord tests yourself using some of today's field testers. Ask your cabling supplier to recommend cords compatible with your permanent link, and get them. Going cheap on patch cords is a false economy.
Well, how did you do? If you scored 10 or better, consider yourself an expert. If you got seven to nine correct, you're a bit rusty but not too bad. If you missed half or more, it's time to go back to school. An excellent source of generic information on testing practices, standards, current limit lines, link and measurement definitions, and much more, can be found at www.cabletesting.com.
Mark Johnston, RCDD, is the strategic alliance and planning manager for Fluke Networks. He can be reached at Mark.Johnston@flukenetworks.com