Copper Certification Testing — Every Parameter

What "certified" means on twisted-pair, per TIA-568.2-D / TIA-1152-A: every parameter in the sweep with its meaning and its real limits — the paired channel/permanent-link values, not the single numbers that circulate — plus the tester accuracy levels and the three-tier tester taxonomy. Voluntary standards that become contractual on any spec'd job.

The certification parameter sweep

TIA-568.2-D
Every parameter with its meaning and limit. Scalar limits are shown as permanent link / channel pairs; frequency-swept parameters test against limit lines across the category's full bandwidth.
ParameterWhat it measuresLimitStatus
Wire mapOpens, shorts, reversals, transpositions, split pairs, shield continuityAll 8 conductors correctPass/fail
LengthElectrical length from propagation delay and NVP≤ 90 m permanent link / ≤ 100 m channelPass/fail
Propagation delayTransit time end to end≤ 498 ns link / ≤ 555 ns channel (@10 MHz)Pass/fail
Delay skewFastest-to-slowest pair difference≤ 44 ns link / ≤ 50 ns channelPass/fail
Insertion lossSignal power lost, swept across frequencyLimit line vs frequencyPass/fail
NEXT / PSNEXTNear-end crosstalk, pair-to-pair and power-sumLimit line, both directionsPass/fail
ACR-F / PSACR-FFar-end crosstalk normalized for attenuation (formerly ELFEXT)Limit linePass/fail
Return lossReflections from impedance mismatch, both endsLimit linePass/fail
DC loop resistanceRound-trip pair resistance (the PoE parameter)≤ 21 Ω link / ≤ 25 Ω channelPass/fail
ACR-N / PSACR-NSNR headroom (NEXT minus insertion loss) — computedRecorded for informationInformative
DC resistance unbalanceWithin-pair and pair-to-pair imbalance (4-pair PoE, new in 568.2-D)Limits exist; field test optionalInformative
TCL / ELTCTLBalance — common-mode noise rejectionLimits exist; field test optionalInformative
PSANEXT / PSAACRF (Cat6A)Alien crosstalk from neighboring cables — the noise DSP cannot cancelSampled: worst-case links, 6-around-1Pass/fail
“Informative” rows have published limits in 568.2-D but field testing is optional per TIA-1152 — specs can and do promote them to required.

Tester accuracy levels (ANSI/TIA-1152-A)

TIA-1152-A
Minimum tester accuracy level by category. 'Level V' is the IEC 61935-1 ladder, not TIA's — the Cat8 grade is Level 2G.
Accuracy levelCertifies
Level IIeCat5e (100 MHz)
Level IIICat6 (250 MHz)
Level IIIeCat6A (500 MHz)
Level 2GCat8 (2,000 MHz)

Verification vs qualification vs certification

TAXONOMY
The three tester tiers. Only certification produces standards pass/fail and registers manufacturer warranties.
TierWhat it proves
VerificationWiremap + length + distance-to-fault — confirms termination, measures no performance
QualificationAdds "will it run 1000BASE-T?" — application SNR assessment, no standards pass/fail
CertificationFull swept-frequency sweep of every TIA parameter against limit lines — the only tier that certifies and registers warranties

Reading a test report like a dispute is coming

The three places link disputes actually live: channel-vs-link limit confusion (check which configuration the tester was set to), marginal passes with asterisks (a *PASS is within the tester's accuracy band — re-test before arguing), and alien crosstalk on Cat6A (a sampled parameter — confirm the sample plan was worst-case bundles, not convenient ones). DC loop resistance has become the parameter to watch as PoE loads grow — it is the one that couples this chart to the bundle ampacity chart and the 28 AWG cord rules on the distance chart. The fiber equivalent of this page is the fiber testing reference.

Common questions

What parameters does a Cat6 certification test measure?

Wiremap, length, propagation delay and delay skew, insertion loss, NEXT and PSNEXT, ACR-F/PSACR-F (the parameter formerly named ELFEXT), return loss, and DC loop resistance — each swept across frequency against the TIA limit lines. Cat6A adds alien crosstalk (PSANEXT/PSAACRF), measured on a sampling of worst-case links rather than every run. ACR-N, balance (TCL/ELTCTL), and DC resistance unbalance are recorded but informative.

What is the difference between channel and permanent link limits?

The permanent link is the in-wall portion a contractor certifies; the channel adds the cords. The paired scalar limits differ accordingly: propagation delay 498 ns link / 555 ns channel, delay skew 44/50 ns, DC loop resistance 21/25 Ω. Quoting the channel number against a permanent-link test (or vice versa) is the classic way a passing link gets argued into a "failure."

What is the difference between a verification, qualification, and certification tester?

Verification proves the pairs are terminated (wiremap, length). Qualification adds "will it carry gigabit?" — an application-level judgment with no standards pass/fail. Certification sweeps every TIA parameter against the limit lines and is the only tier that produces a standards-compliant pass/fail — and the only results manufacturers accept for warranty registration. Bids that say "tested" without saying which tier are underspecified.

What tester level is required for Cat6A?

Level IIIe accuracy per ANSI/TIA-1152-A (Level III covers Cat6; Level 2G covers Cat8). One trap: "Level V" is not a TIA designation — it belongs to the IEC 61935-1 ladder, and vendor copy sometimes conflates the two systems.

Run your whole job on the same numbers

These NORDIX tools are a taste of the full platform — bid pipeline, estimating, and job costing that carry your numbers from the first bid to the final invoice. Our team sets it up for your shop and walks you through your next real job.

Request access →