2.8 Certification: Permanent Link vs Channel
Key Takeaways
- TIA-568 defines two link models: the permanent link (≤90 m, including the outlet and patch-panel consolidation point) and the channel (≤100 m, including all patch and equipment cords).
- Permanent-link and channel-test adapters apply different compensation, so the tester must be configured to the correct model before testing — a permanent-link adapter on a channel test produces false failures.
- PASS/FAIL is reported per parameter against per-frequency limits; margin in dB is the worst-case difference between measured and limit. A channel with margin below ~3 dB is technically passing but borderline.
- All certification results must be stored with cable ID, link model, tester serial, and timestamp, and delivered as part of the project's as-built documentation.
Two Link Models
TIA-568 defines two distinct link models that the Technician certifies against. They are not interchangeable — the limits, the test adapters, and the cord set used for each are different.
Permanent Link (PL)
The permanent link is the fixed portion of the cabling: the horizontal cable plus up to one consolidation point and the work-area outlet and patch-panel terminations. Maximum length: 90 m. Patch cords and equipment cords are not part of the permanent link. The permanent link is what the installing contractor certifies as the installed infrastructure, because it is the part that will not be moved or replaced over the cable's life.
Channel
The channel includes the permanent link plus all patch cords and equipment cords at both ends: the work-area cord, the patch cord in the telecom room, and the equipment cord to the switch. Maximum length: 100 m. The channel is the end-to-end path a signal traverses. Channel certification is typically done by the end user or integrator after patch cords are in place.
Why the Distinction Matters
The two models have different limit curves. The channel's limit allows for the additional connectors and cords at each end, which contribute insertion loss and crosstalk. The permanent link's limit is tighter because it represents the fixed infrastructure, without cord contributions.
A common error is to test a permanent link with a channel adapter (or vice versa). The tester applies different compensation for the adapter's internal connections; using the wrong adapter produces systematically biased results. The Technician must:
- Select the correct link model on the tester (PL or CH).
- Use the correct adapter (PL adapter for permanent link; channel adapter for channel).
- Establish the reference per the link model.
Test Adapters
A permanent-link adapter has a short, fixed test cord with a controlled connector; the cord's performance is mathematically subtracted from the measurement. The adapter's connector mates to the work-area outlet or patch-panel port.
A channel adapter accepts the patch cord plug directly; the patch cord is part of the measurement. The channel adapter's internal connection is also compensated, but the patch cord contribution is included.
Mixing these up produces false results in both directions. A PL adapter on a channel test produces an apparent PASS where the channel would actually fail (the patch cord's crosstalk is not measured); a channel adapter on a PL test produces false failures.
PASS/FAIL Criteria
For each parameter, the tester compares the measured value against the limit at each frequency in the band. The parameter passes if the measured value is on the compliant side of the limit at every frequency, and fails if it crosses the limit at any frequency.
The test reports the worst-case margin — the smallest distance between the measured curve and the limit curve across the band. Positive margin means PASS; negative means FAIL.
| Parameter | Direction | "Good" Side |
|---|---|---|
| Insertion loss | Lower is better | Measured < limit |
| NEXT, PSNEXT | Higher dB is better | Measured > limit |
| FEXT, PSACRF | Higher dB is better | Measured > limit |
| Return loss | Higher dB is better | Measured > limit |
| Alien crosstalk (PSANEXT) | Higher dB is better | Measured > limit |
The Technician must know which direction is "good" for each parameter to interpret the plot. A common exam question shows a margin plot and asks whether the parameter passes; the answer depends on which side of the limit line the curve sits, which depends on the parameter.
Margin Interpretation
Margin is the most important number on a certification report after the PASS/FAIL flag. A channel with 3 dB or more of margin on every parameter is a robust install. A channel with less than 3 dB on any parameter is technically passing but borderline — a future MAC, a cable moved in the bundle, or a connector aged by repeated patch-cord swaps can push it to FAIL.
Best practice on the project:
- Margins ≥ 3 dB on all parameters: accept.
- Margins 1–3 dB: inspect and consider re-termination; document the borderline margin.
- Margins < 1 dB or FAIL: re-terminate or repair and re-test.
Some project specifications mandate a minimum margin (e.g., 3 dB) above the TIA limit for acceptance. The Technician must read the project spec, not assume the TIA pass/fail alone is the acceptance criterion.
Frequency Step and Test Band
The tester sweeps from a low frequency (typically 1 MHz) to the category's top frequency:
- Cat 5e: 100 MHz
- Cat 6: 250 MHz
- Cat 6A: 500 MHz
- Cat 8: 2000 MHz
The frequency step (1 MHz, 3 MHz, etc.) determines measurement resolution. TIA-1152 specifies the test step. Modern testers use a stepped sweep with enough resolution to catch narrow limit violations.
Documentation Requirements
Every certification test must be stored with:
- Cable ID (matching the project's labeling scheme per TIA-606)
- Link model (PL or CH)
- Cable category tested
- Test standard and revision
- Tester serial and calibration date
- Operator name or ID
- Date and time
- PASS/FAIL per parameter and overall
- Margin per parameter
The full set of tests, exported as a PDF or CSV, becomes part of the project's as-built documentation. Owners use this package to verify the install meets contract and to baseline future re-certification after MACs. Lost or incomplete test records are a common project dispute; the Technician must produce the records at project closeout.
TIA Standards Governing Certification
| Standard | Scope |
|---|---|
| TIA-568 | Cabling categories, link models, pinout |
| TIA-1152 | Field test requirements, tester performance |
| TIA-606 | Administration and labeling |
| TIA-607 | Bonding and grounding |
| ANSI/TIA-568.0 (Generic) | General cabling requirements |
| ANSI/TIA-568.2 (Copper) | Copper-specific requirements |
TIA-1152 is the standard the field tester certifies against — it defines what "certification" means in terms of tester accuracy, sweep, and limit curves. A test result labelled "TIA-568 Cat 6A Perm Link" is implicitly a TIA-1152 test.
Hands-On Exam Task
A typical hands-on task: certify a Cat 6A permanent link. The candidate selects the correct link model, attaches the correct adapter, establishes the reference, runs the test, and reports the result with margin. The evaluator watches for correct model/adapter selection — a candidate who runs a channel test on a permanent link adapter fails the task even if the result shows PASS, because the test method is wrong.
Which link model includes patch and equipment cords at both ends and has a maximum length of 100 m?
A certification report shows a Cat 6A permanent link with a NEXT margin of +0.4 dB. What is the correct interpretation?