PROFINET EtherCAT Cable in Drag Chains: Cat6A Checklist

Cat6A is a static bench rating. Use this 7-point checklist to separate drag-chain PROFINET/EtherCAT cable that survives flex from one that only passes on the bench.

A Cat6A test report tells you the cable passes a static channel test on a bench. It does not tell you the cable will survive a few hundred thousand cycles in a carrier. The purchase decision has to rest on cycle-qualified flex data: flex life quoted at a stated radius, travel speed and cycle count; impedance and return loss measured at the declared minimum bend radius; a real torsion specification if the axis rotates; the shield construction and how coverage is held under flex; and a jacket compound matched to the fluids on your machine. This article is a qualification checklist for that data, not a guarantee — no third-party flex-life test results were available for it.

A Cat6A test report and a drag-chain test report are not the same document

Category ratings describe a static channel: conductor construction, pair geometry and transmission performance at a fixed installation. They say nothing about life expectancy under continuous flex, torsion or chain acceleration. That is why the failure pattern is so familiar — the link is clean at commissioning, then CRC counters start climbing months later, in the same segment, on the same machine, with no change in the program.

The rules that govern the moving installation sit in a different layer from the category label. PROFINET cabling is addressed in IEC 61784-5-3, with broader building-cabling context from ISO/IEC 11801-3 and IEC 61918, as set out in this Siemens PROFINET Cat6A selection guide. A Cat6A datasheet alone does not answer the moving-installation clause, and no category certificate will state a cycle count.

Treat everything below as an evidence request. If a supplier cannot answer a line item with data taken under stated conditions, that item is unquantified — and your own sample test in your own carrier becomes the only remaining source of proof.

What actually fails inside the chain — and where

Before writing the RFQ, be clear about the mechanism, because it determines which datasheet fields matter.

  • Impedance ripple and return-loss drift. Repeated bending changes conductor and pair geometry slightly. Small, cumulative geometry changes show up first as degraded return loss and intermittent link errors on a gigabit segment, long before the cable is mechanically broken.
  • Shield degradation. Braid strand breakage and foil micro-cracks reduce coverage and raise transfer impedance. The symptom is not a dead link — it is error bursts correlated with motion or with a nearby drive.
  • Conductor fatigue in the moving span. The damage concentrates in the section that actually travels; that is the section the flex-life figure must describe.
  • Jacket cracking and swelling. Coolant, oil and chip exposure combined with flex cycling attack the jacket, and a cracked jacket exposes the shield to the same fluids.

Note where the damage usually is not: a continuously-flex-rated cable is often damaged in the fixed entry and exit zones, where it is over-bent or clamped rigidly at the wrong point of the travel. When you investigate a failure, separate the moving span from the entry and exit before you condemn the cable.

Check 1 — Does the datasheet quote flex life at a stated radius, speed and cycle count?

A flex-life claim without a radius, a travel speed or acceleration, and a cycle count is marketing. Ask for all three, plus the test bench arrangement: chain type, fill weight, travel length and whether the cable was run alone or in a mixed harness.

Make the supplier commit in writing to one of three duty classes — fixed installation, occasional or flexible use, and continuously flexible in a carrier. Then compare published radius ladders rather than single numbers. One industrial Cat.6 datasheet, SAB Bröckskes' CATLine Cat.6 S, publishes 5 x d for fixed laying, 10 x d for flexible application and 15 x d for continuously flexible use — that three-step ladder is the format to demand from every alternate supplier, so the comparison is like for like.

The same source also shows that the temperature window shifts with duty: it quotes −40 °C / +70 °C for VDE flexible application and up to +80 °C under UL. A wide static range does not transfer to the moving case; ask for the range that applies to the duty class you are buying.

Check 2 — Is impedance and return loss specified as stable over cycles, not just "nominal"?

Ask for impedance-versus-cycle-count data, or at minimum a pass/fail statement at the end of the flex test. Return-loss degradation is the usual route from a healthy channel to intermittent CRC errors on a 1 Gbit/s PROFINET or EtherCAT segment, and a nominal value at time zero will not warn you.

Three conditions make the difference between data you can use and data you cannot:

  1. Routing during measurement. Return loss and near-end crosstalk should be measured with the cable routed at the declared minimum bend radius, not straight on a drum. A straight-run number says nothing about the chain entry and exit curve.
  2. Termination. Confirm which connector the data was taken with — industrial RJ45 or M12 X-coded. The mated pair contributes to return loss, and your channel includes the connector.
  3. End-of-test state. Data from a new sample is not the same as data after the claimed cycle count. Ask which one you are being shown.

If cycle-dependent data does not exist, downgrade the decision: buy a prototype sample and test it in your own carrier before series release.

Check 3 — Shield construction, and what happens to coverage under flex

Confirm the shielding scheme in writing. Foil-plus-braid constructions — SF/UTP or S/FTP — are the norm for industrial PROFINET cable, in contrast with the unshielded or simple foil constructions typical of office cable, as compared in this PROFINET versus standard Ethernet cable comparison.

Then ask the question most datasheets avoid: how is shield coverage or transfer impedance maintained after flex? Braid damage and foil micro-cracks are a plausible mechanism for motion-correlated error bursts, and the supplier should have a stated position on it rather than a photograph of a braid.

Finally, confirm the shield-termination method the cable is designed for — a 360° clamp or a pigtail. A good cable with a poor termination fails EMC exactly like a bad cable, which is why this item belongs in the same conversation as the cable spec.

Check 4 — If your axis rotates, is there a torsion specification at all?

Torsion is a separate qualification from bending. "Suitable for robots" is not an answer. Ask for the torsional angle per metre, the cycle count and the test length.

Use a supplier classification scheme as vocabulary, not as a verdict. The selection guide cited above distinguishes a Type S drag-chain classification from a Type S-Torsion variant used for a robot dress-pack axis-6 rotation, and maps a Type R variant to outdoor or wash-down routing. That is a useful way to name what you need — then demand the underlying numbers.

For combined bend-plus-twist motion, ask whether the flex-life figure was measured with the twist superimposed or in pure bending only, and ask for the recommended installation window along the cable where the twist may occur, so the mechanical designer can check it against the actual axis geometry.

Check 5 — Jacket compound: PUR versus PVC, and the fluids on your machine

PUR is the usual choice for continuously flexed, harsh-environment industrial cable; PVC is typically offered for ordinary working environments, again per the comparison cited above. The decision hinge is your actual coolant, oil, grease and chip exposure — not the phrase "harsh environment".

Ask for the fluid-resistance list with the test media and immersion conditions the supplier actually tested. "Oil resistant" without a medium, a temperature and a duration is not a specification.

Cross-check fire and halogen data if the machine ships into a building with cable-routing requirements. Published examples on the SAB datasheet include halogen-free to IEC 60754-1 and flame retardant, self-extinguishing behaviour to IEC 60332-1-2, both cited with their VDE equivalents. Note the scope: these are jacket-material tests, not a statement about flex life or channel performance.

Last item on this check: confirm the jacket is rated for the temperature rise inside the carrier at your duty cycle, not only the ambient shop temperature.

PROFINET versus EtherCAT: where the cable selection question actually differs

Do not assume one cable serves both protocols by default, and do not let a supplier answer "both" without evidence for your channel.

  • PROFINET. Cable selection is tied to a published cabling framework (IEC 61784-5-3) and to the conformance class the channel is intended to support. Ask which type designation in the supplier's classification your part corresponds to, and which conformance class the channel is intended for. The type designation is the supplier's own naming, so ask for the data behind it.
  • EtherCAT. The relevant cable parameters are the ones your segment length actually depends on — loop resistance, attenuation and the connector used for the measurement. Ask for those at the length you intend to run, taken with the termination you intend to use.

This article does not reproduce either specification. Treat impedance, attenuation and length limits as verification steps to be confirmed against the current edition of the applicable document and the supplier's declaration of conformity. That is a gap you close in the RFQ, not a gap you fill from a blog post.

Put it in the RFQ: the checklist that turns a datasheet into a decision

Test itemWhat to ask the supplierWhy it changes the purchase decision
Flex lifeCycle count at a stated bend radius, travel speed or acceleration, travel length, chain type and fill weightWithout radius and cycles the claim cannot be compared to your design life or to a competing quote
Impedance and return loss over cyclesValues at time zero and at end of flex test, measured at the declared minimum bend radiusReturn-loss drift is the mechanism behind late-appearing CRC and link errors
NEXT at minimum bend radiusMeasurement routing and the termination used for the dataStraight-run crosstalk data does not describe the chain entry curve
TorsionAngle per metre, cycle count, test length, whether superimposed on bendingDecides whether the part is usable on a rotating axis at all
Shield constructionFoil/braid scheme, coverage or transfer impedance after flex, termination methodExplains motion-correlated error bursts and determines the EMC outcome with your clamps
Minimum bend radiusThree figures: fixed, flexible, continuously flexibleSets the carrier geometry and the entry/exit clamping points in your design
Jacket compoundPUR or PVC, fluid list with media and immersion conditions, jacket temperature range at your duty cycleDecides jacket life in coolant, oil and chip exposure

Red flags: a cycle count with no radius; "continuous flex" supported only by a static Cat6A report; torsion described qualitatively; oil resistance with no test medium; a flex-life figure with no statement of whether it was measured on the cable alone or in a loaded harness.

One rule worth quoting internally: if the datasheet cannot state radius, speed and cycles together, the flex claim is unquantified and a sample test is mandatory.

Still throwing errors after a compliant cable is installed? Check these before blaming the cable

  • Routing discipline. Run the data cable separated from power in the chain, never strapped to it with cable ties, respect the chain manufacturer's minimum bend radius, and keep the chain straight at entry and exit — the drag-chain routing guidance summarised in the same selection guide cited above.
  • Entry and exit zones. Check for over-bending or rigid clamping in the fixed zones, the most common place for a genuinely good cable to be destroyed.
  • Shield termination and equipotential bonding end to end. A broken 360° clamp presents identically to a bad cable.
  • Triage order. Swap the segment for a known-good lead, then move the suspect lead to a different carrier position, then re-terminate. Record which change cleared the error counter before you re-specify anything.

FAQ

Can I run a standard Cat6A patch cable in a drag chain if it passes the channel test?

A channel test is a static measurement. It does not qualify the cable for continuous flex, torsion or chain acceleration, and it produces no cycle data. Use it as a baseline, not as approval for a moving application.

Why does my PROFINET line show CRC errors only after a few hundred thousand cycles?

Because the degradation is cumulative and mechanical: pair geometry and shield integrity shift gradually under repeated bending, and return loss or shielding performance crosses a threshold before the cable breaks. It is a wear mechanism, not a defect that appears at commissioning.

What bend radius should I specify for a continuously flexed Cat6A PROFINET cable?

Specify the supplier's continuously-flexible figure, not a general industrial-cable number, and check it against the chain manufacturer's own minimum radius — the tighter of the two governs. Ask for fixed, flexible and continuously flexible figures so the entry and exit zones can be designed separately.

Is the same cable suitable for both PROFINET and EtherCAT?

It can be, but not by assumption. Ask the supplier to confirm the construction against both sets of requirements and to supply the electrical data at your segment length with your termination. A "both protocols" answer without data is not a qualification.

Does a PUR jacket alone make a cable drag-chain rated?

No. The jacket compound addresses fluid and abrasion exposure. Drag-chain suitability comes from the conductor stranding, the shield construction and the tested flex life — a PUR jacket on a static construction is still a static cable.

Request a drag-chain test report and a quote for PROFINET/EtherCAT drag-chain cable

Send six numbers and one fluid list: carrier radius, travel length, speed and acceleration, target cycle count, torsion angle if the axis rotates, protocol and data rate, plus the coolants and oils the cable will see. You get the matching flex data for review — radius ladder, cycle count, impedance and return-loss position, jacket fluid list — and a quote for the drag-chain / continuous-flex industrial Ethernet cable range against those inputs.

If you are still at discovery stage, take the sample route: one assembly, run in your own carrier, with the acceptance criteria from the checklist above agreed in writing before it ships. That converts an unverifiable datasheet claim into a measured result on your machine — which is the only evidence that finally settles the question.

Request a quote

Share a model or describe your application. Include a drawing link if you have one.