A 250-Link Cat 6 Acceptance Report That Cannot Be Issued

None of them, as the report stands. Three independent reasons, any one of which invalidates the acceptance test:

1. Calibration is out of date. A tester 14 months past a 12-month calibration interval has no traceability. Its measurements cannot be defended to the customer, the manufacturer's warranty program, or a third party. The unit must be recalibrated (or replaced) and the crew should field-verify with a reference/calibration module at the start of each test session.
2. Wrong adapter for the object under test. Channel adapters measure the end-to-end channel including the patch cords. The deliverable here is the permanent link. Using channel adapters means the recorded result is not the thing the customer contracted for - and it also masks the length problem, because channel length limits differ from permanent link limits.
3. "Basic" is not a certification test. A Cat 6 acceptance run must include, at minimum: wire map, length, insertion loss, NEXT, PS NEXT, ACR-F (ELFEXT), PS ACR-F, return loss, propagation delay, and delay skew, evaluated against the Cat 6 limits across the full frequency range. A reduced autotest does not include the full suite and cannot certify anything.

What to say to the IT manager: the report needs to be re-run, not signed. Give her a realistic date and the reason in writing - handing over a report you know is unsupportable is worse than a one-week delay, because it becomes your liability after payment.

Check the NVP setting first.

Length in a copper tester is not measured directly - it is derived from the electrical time-of-flight multiplied by the nominal velocity of propagation. If the tester is set to a default (commonly ~69%) and the actual cable is, say, 72%, every reported length is inflated by roughly 4%, and a compliant 88 m link reads as 92 m. Set NVP from the cable manufacturer's published data or calibrate against a known measured length.

Then re-test and separate the cases:

1. Genuinely under 90 m (295 ft) after NVP correction - the "failure" was a tester configuration artifact. Re-test with correct settings and the right adapters; document the corrected NVP in the report.
2. Genuinely over 90 m - the permanent link cannot be certified to the standard. Options: re-route for a shorter path; relocate or add an intermediate cross-connect/consolidation point; or accept the link as a documented variance with the owner's written approval. What you must not do is record it as a pass.

Second check - the channel budget. Even a compliant 90 m permanent link plus patch cords must stay within the 100 m channel. Allow about 5 m at each end. A 96 m permanent link cannot work in any configuration.

Third check - return loss on the short links. Eight return loss failures on links that are well within length is a separate defect: suspect tight cable ties, a hard bend, a crushed section, or damaged cable from the pull. Locate with the tester's HDTDR/reflection function to find the distance to the discontinuity before you cut anything.

The 8 return loss failures:

• Treat them as real until proven otherwise - return loss is a reflection measurement, so a failure means a physical impedance discontinuity: a tight tie, a kinked or crushed section, a damaged jacket from the pull, or a bad termination.
• Use the tester's time-domain reflectometry function to get the distance to the event and go look at that spot rather than re-terminating blindly.
• Re-terminate or replace only the defective section, then re-certify the whole link with the correct adapters - do not re-test just the failing parameter.

The closeout package should contain:

1. Test reports for every link, in the tester's native format and a readable PDF, showing: tester make/model/serial, firmware version, calibration date, adapter type used, test limit selected (Cat 6 permanent link), date/time, and operator.
2. As-built drawings reflecting the actual routes, pathway types, firestopped penetrations, and locations of any intermediate points.
3. Label schedule matching cable IDs, panel positions, and outlet numbers to the test report.
4. Grounding/bonding records (TIA-607) for the TMGB/TGB and any shielded elements.
5. Firestop records - system used, location, installer, date.
6. Component/warranty package - manufacturer, part numbers, and the warranty registration.
7. Written approvals of any variances (over-length links, substitutions, accepted exceptions), signed by the owner's representative.

Judgement: the report and the as-builts must agree. A clean test report that cannot be tied to a labelled outlet on a drawing is not documentation, it is a stack of paper.