Cat 6A Channel Failing Alien Crosstalk and Length Margin
Your crew pulled 26 Cat 6A horizontal runs from a second-floor IDF to a new medical office suite. The longest cable measures 89 m with the tester's default NVP; the channel adds a 5 m workstation cord and a 4 m equipment cord at the patch panel. For most of its route the bundle shares J-hooks with four 20 A, 120 V branch circuits in EMT for a 30 ft section, and the hooks are supported every 5 ft. Permanent-link tests at the manufacturer's Cat 6A limits show the longest run failing PSANEXT margin by 1.3 dB at 500 MHz; two other runs fail insertion loss. The customer's IT director requires a manufacturer 25-year channel warranty and will not accept a permanent-link-only report. The GC asks whether the bundle can simply be re-tested with a different adapter.
1. Determine whether the length is compliant and what you must verify before accepting the tester's length reading.
Read the worked answer
Channel length = 89 m cable + 5 m workstation cord + 4 m equipment cord = 98 m, which is under the 100 m channel limit, and the 89 m permanent link is under the 90 m limit - but the margin is only 1–2 m, so the measurement should not be trusted blindly. Field testers derive length from NVP; if the tester's NVP is set for a different cable, the reported 89 m could be off by several meters. Before accepting the run I would (1) confirm the correct cable NVP and re-run the test, (2) physically measure the route including slack and service loops, and (3) verify the cords are the manufacturer's approved Cat 6A cords. If the physical channel is over 100 m, the run is noncompliant regardless of the tester's pass. The two insertion-loss failures also point to a physical problem - damaged cable, tight bend, or bad termination - rather than a tester setting.
2. Explain the likely cause of the PSANEXT failure and the remedies. Is re-testing with a different adapter a legitimate fix?
Read the worked answer
PSANEXT is alien crosstalk between adjacent cables - exactly what a 30 ft parallel run with four 20 A branch circuits in EMT will produce, especially in a tight bundle. Re-testing with a different adapter does not change the physics; if anything, changing adapters without recalibration invalidates the report.
Remedies, in order: separate the data bundle from the power raceway (follow TIA-569 and manufacturer separation guidance - commonly at least 2 in. from branch-circuit raceway, more for larger power or transformers); re-route the 30 ft section; loosen and re-dress the bundle so cables are not tightly bundled; and replace the runs that failed insertion loss. If separation is impossible, use shielded F/UTP with proper grounding or reduce bundle size. After remediation, re-test alien crosstalk (ANEXT/PSANEXT) on the affected runs, not just a quick re-test.
3. The IT director wants a manufacturer 25-year warranty. What testing and documentation do you require before submitting, and what do you tell the GC about the adapter suggestion?
Read the worked answer
The warranty is a channel warranty, so the IT director is right to demand channel testing. Before submitting: verify the field tester's calibration is current (typically within 12 months), use the manufacturer's latest test limit and approved cords, test 100% of runs, and submit native test files rather than screenshots. A permanent-link-only report may show the installation is acceptable but will not support the warranty. The GC's suggestion to re-test with a different adapter is not acceptable; if an adapter is changed, the tester must be reconfigured and the original failure still exists. I would tell the GC and IT director: two runs need replacement, the bundle needs re-routing, and the report will be re-issued after remediation - no "certified" letter until the channel tests pass.