On a typical project the cable is written into the specification, the patch panel is chosen by brand and the keystone jack sample is checked on the desk. The patch cords are added to the order last, often in the final week, at whatever unit price the quote shows. They make up 10m of the 100 m channel the tester measures, and the connection at each end of that channel is made through their plugs.
The team that buys in this order has a reason. Patch cords are a small line in the budget. The permanent link test has passed, the report is filed, and the expensive part of the job looks finished. What is left is plugging in the equipment.
Users plug into the channel.
A signal from a laptop to the switch passes through one patch cord before it reaches the wall outlet and another one after it leaves the patch panel in the rack. Both cords write part of the channel result. The sections below put numbers on the allowance the standards give patch cords, the build differences that show up in measurement, and the documents worth asking for before you buy.
Channel and permanent link: what each test measures
ANSI/TIA-568.2-D and ISO/IEC 11801-1 cap a copper channel at 100 m. Up to 90 m of that is the horizontal cable installed in the building, the permanent link. The remaining 10m covers every cord in the channel combined: patch cords in the rack, equipment cords and the work area cord. The channel model allows up to four connections.
Field certification has two separate measurements. A permanent link test covers the fixed part, from the patch panel to the outlet, and leaves the patch cords out. A channel test covers the whole path from one active device to the other, cords included.
Fluke Networks, a maker of field certification testers, states the rule plainly. A passing permanent link becomes a passing channel when you add compliant patch cords. Once the cords in a tested channel are swapped, the channel result no longer describes what is installed. The same source points out that patch cords are handled and moved more than any other component, which leaves them the most exposed to damage, and that their quality is overlooked because they are treated as a commodity.
The permanent link report documents the building. The channel is what users work on every day, and patch cords make up both of its ends.
Why the standard gives 10 m its own limit
A patch cord is built to bend, move and be plugged in hundreds of times. Its conductors are stranded for that reason, while horizontal cable uses solid copper. Flexibility has a cost. A stranded conductor has higher insertion loss than a solid one of the same gauge. That is one reason the cords are held to 10 m. The channel budget assumes the lossier section stays short.
The same logic applies in every category. A Cat.6 channel is tested up to 250 MHz and a Cat.6A channel up to 500 MHz. Insertion loss rises with frequency, so the extra loss of stranded cable counts for more at the top of the band.
Thin cords tighten the budget further. ANSI/TIA-568.2-D recommends a de-rating factor of 1.95 for 28 AWG patch cords. Fluke Networks works out that with 10 m of 28 AWG cords in a channel, the permanent link has to come down from 90 m to 82.5 m. For DC loop resistance, the standard recommends keeping the total length of 28 AWG cords in a channel under 15 m.
Same label, different results: three kinds of patch cord
You will find three kinds of patch cord on site, and all three can carry the same category on the bag. The difference shows up in the measurement.
| Comparison | Factory-terminated, with test report |
Factory-terminated, no report |
Cut and plugged on site |
|---|---|---|---|
| Termination | Controlled line; pair untwist length held constant | May come off a controlled line; result unknown | By hand; untwist length varies from one technician to the next |
| Basis of the category claim | Measurement report to the standard | Label on the bag | Labels on the plug and the cable used |
| Effect on the channel test | Channel result is predictable | Result unknown until tested | Installed untested, it can become the weak point of the channel |
| When replaced | Swapped for the same documented cord; channel behaviour holds | A new cord brings a new unknown | Every new cord is new handwork |
On smaller screens, scroll horizontally to compare all three types.
The real risk of a field-made cord sits behind the plug. Near-end crosstalk (NEXT) builds up mostly at the plug and jack connection at the near end of the channel. The further the pairs are untwisted before they enter the plug, the higher the crosstalk. A factory line holds that untwist length constant. On site, each termination gives its own result, and two cords made by two technicians will not measure the same.
An undocumented factory cord sits in the middle. It may come off a controlled line, but its category claim rests on the label alone and nobody knows the result until the test.
What the last line on the order costs
When patch cords are bought last, nobody asks for the test documents that were required for the cable and the panel. Close to handover, whatever is in stock gets bought. If a channel test was run, the cords get swapped afterwards and the report no longer describes the cords in the rack.
The bill arrives in the field. When a channel fails, the crew opens the panel and the outlet first, because they look for the fault in the fixed part. If the cord is the problem, those hours are lost and the same link is tested a second time. If the problem shows up after the active equipment goes live, it comes back as dropped links and slow connections reported by users, and by then nobody goes back to the cord list from handover.
Which test, which cord at handover
Write the permanent link test into the specification for acceptance. The report on the fixed part stays a valid reference for the life of the building, and it keeps its meaning when cords are replaced. Run channel tests at handover on a sample, using the cords that will actually stay in place. Each sampled link is tested with the cords at both ends.
Ask for a separate document for the cords themselves. Patch cord test adapters measure a cord against the standard on its own, apart from the permanent link. A cord that ships with a factory measurement report moves that step to the purchasing stage and shortens test time on site.
Used together, the two steps document both halves of the channel: the fixed part through the permanent link report and the moving part through the cord’s own measurement.
Length and colour plan in the rack
Cord length takes its share of the budget too. If 2 m cords go where 50 cm would reach between a 48-port panel and the switch, every link carries 1.5 m of loss it does not need and the rack fills with coiled slack. The coils block airflow, cover the port labels and slow down the search for the right cord during a fault.
Draw up the length list when the rack layout is drawn. Set separate lengths for panel to switch, rack to rack and the work area, and put the quantity of each on the bill of materials. Define the colours on the same list. Separate colours for data, IP phones, cameras and uplinks tell the technician which cord to pull without reading a label.
Label both ends under the ANSI/TIA-606 administration standard. These details take the patch cord off the last-minute list and put it on the same line of the project file as the cable and the panel.
What to ask for when you buy
- The cord category and conductor build (AWG, stranded) stated on the datasheet.
- A measurement report to ANSI/TIA-568.2-D or ISO/IEC 11801-1 behind the category claim.
- Patch cords quoted in the same offer as the cable and panel, under the same system description.
- If 28 AWG cords are planned, the project’s long permanent links checked against the 82.5 m limit.
- A colour per service, fixed at the start of the project. A cord plugged into the wrong port on site opens a new fault ticket.
Zemecs patch cords
Zemecs quotes its T221 series patch cords as part of one system with the cable, patch panels and keystone jacks. Contacts are plated with 50 µin gold, and the cords come in nine colours so a colour plan per service can be set at the start of the project. Zemecs products carry documents from independent bodies, including ETL and Force Technology. Zemecs gives a 25-year performance warranty on the whole system, patch cords included (terms apply).
AMCE Hospital in Abuja, Nigeria, shows what the system approach looks like in service: 144 km of Cat.6 and 5,906 copper nodes, each one turned into a channel by a Zemecs patch cord in the rack and another at the desk. Cable, panels, outlets and cords all come from one system, and the site has recorded zero failures.
Frequently asked questions
If the permanent link test passed, do I still need a channel test?
The permanent link test documents the fixed part of the building and is the right basis for acceptance. According to Fluke Networks, a passing permanent link makes a passing channel with compliant patch cords. What decides the outcome is proof that the installed cords are compliant.
Can patch cords be made on site?
They can. Each field-terminated cord is a separate piece of handwork, and installed without a measurement it leaves the channel result unknown. Plan a test with a patch cord adapter for any cord made on site.
Can I use thin 28 AWG patch cords?
Yes. ANSI/TIA-568.2-E recommends a de-rating factor of 1.95 for them. With 10 m of 28 AWG cords in a channel, the permanent link is limited to 82.5 m, and the total length of 28 AWG cords should stay under 15 m.
Can a patch cord be one category below the cable?
A channel performs at the class of its lowest-rated component. A Cat.6 cord on a Cat.6A permanent link means the channel cannot be certified as Cat.6A, and the measurements up to 500 MHz run into that cord’s limit.
How much patch cord can a channel hold?
In the standard channel model, all cords in the rack and the work area add up to 10 m, and the horizontal cable is limited to 90 m. Fluke Networks recommends 5 m for equipment cords.
The last 10 metres
Ten of the 100 metres in a channel test are patch cord, and every connection a user makes each day runs through them. A team that buys them with the cable and panel, in the same offer and against the same documents, gets back in the field what it spent at the table.
Send the bill of materials for your next project to zemecs@turera.com and we will quote the patch cords together with the cable and panels.
Lose at the table or lose in the field. Or win twice.
Sources
ANSI/TIA-568.2-D and ANSI/TIA-568.2-E — Balanced Twisted-Pair Telecommunications Cabling and Components Standard.
ISO/IEC 11801-1:2017 — Information technology: Generic cabling for customer premises — Part 1: General requirements.
ANSI/TIA-606 — Administration Standard for Telecommunications Infrastructure.
Fluke Networks — Cabling Mistakes #1: Specify Channel Testing While Installing Permanent Links.
Fluke Networks — The Skinny on 28 AWG Patch Cords.