Skip to content
Ship to WorldwideUSDLanguageEN中文
Guides

Sourcing USB and USB-C cables

Two cables with the same plugs at both ends can differ enormously in speed and power, and nothing on the outside tells you which is which.

A person solders a green circuit board attached to a white plastic casing on a cluttered workbench, with tools and a mug nearby.
On this page
  1. The four properties a connector does not tell you
  2. What actually makes one cable cost more
  3. Certification, logos and what you may print
  4. Specifying, sampling and inspecting
  1. The four properties a connector does not tell you

    Two large, old industrial machines with prominent wheels stand in a derelict factory hall with tall windows and peeling walls.

    Power. The current the cable is built and rated to carry, which depends on the conductor gauge of the power pair and on the connector terminations. Above a threshold defined by the standard, a USB-C cable must contain an e-marker chip that declares its rating, and without it a charger will not negotiate the higher current at all — so a high-power charger sold with an unrated cable quietly delivers less than the box promises.

    Data. A cable may carry only basic-speed data, or multi-gigabit, and the difference is real construction: additional shielded pairs, tighter impedance control, better shielding, shorter maximum length. Specify the rate in gigabits per second rather than by a marketing generation name, because the naming has been revised repeatedly and reused.

    Video. Carrying a display signal over USB-C requires the cable to support the relevant alternate mode, and a charging-grade cable will not do it. If your product claims a display connection, that is a separate line in the specification and a separate test.

    The e-marker. A small chip that reports the cable's capabilities to the devices at each end. It is required above certain power and data levels, and it is the component most often omitted in a cost-down while the packaging claim stays the same.

    Length is a constraint, not a preference. Higher data rates and higher currents both get harder with length, so a specification that asks for the maximum of both at three metres is asking for a cable that either does not exist or does not work.

  2. What actually makes one cable cost more

    Conductor gauge. Thicker conductors in the power pair mean less voltage drop and less heat at high current. It is the largest single material cost and it is invisible from outside, which is why it is the first thing reduced.

    Shielding and pair construction. Individually shielded pairs, drain wires and a proper overall braid are what make high data rates work and what keep the cable from interfering with radios. A cable that passes at a bench and fails beside a phone has a shielding problem.

    Connector construction and termination. How the wires are attached to the connector, whether the joint is supported, and how the overmould grips both. Nearly every cable that dies in the field dies at a connector, and this is where it is decided.

    Strain relief and jacket. Overmould geometry, jacket material and any braiding. Braided cables look premium and the braid is not itself durability — the strain relief geometry underneath is.

    Bend-cycle claims should be treated as marketing unless you define the test. "Tested to N bends" figures in this category come from non-comparable rigs with different angles, weights and radii. If durability matters to you, write down the angle, the load, the radius and the pass criterion, and have the test done to your definition.

  3. Certification, logos and what you may print

    The USB logos are licensed, not decorative. Using the certification marks requires the product to be certified through the standards body's programme and the company to be a licensee. Printing a mark on an uncertified cable is a trademark problem as well as a quality claim you cannot support, and marketplaces increasingly ask.

    Radio and safety approvals still apply. A passive cable is not a radio, but the finished product placed on your market still carries the destination's general product safety obligations, and a cable that overheats is a safety matter rather than a quality complaint.

    Material declarations. Restricted-substance requirements apply to the jacket, the plating and the solder, and the declaration must name your model. Where a market requires a substance declaration for the article, the supplier's generic statement is not the same as a report for the product you are buying.

    And check what the packaging claims. Claims about speed, power and compatibility on the box are claims by the party placing the product on the market — which, once your brand is on it, is you. A cable whose packaging over-claims is a return rate and a listing takedown waiting to happen.

  4. Specifying, sampling and inspecting

    Huge stacks of wooden pallets fill an expansive outdoor storage yard, with trucks parked and a forklift moving more.

    Write the specification as four numbers plus a construction. Current rating and whether e-marked; data rate in gigabits; video support yes or no; length; then conductor gauge for the power pair, shielding construction, connector shell and plating, jacket material, and the overmould and strain-relief detail.

    Ask for the cable's own test data, not the connector supplier's. Continuity and resistance, high-current temperature rise, data-rate verification at your length, and the CC resistor value on A-to-C constructions.

    Then run three checks on the samples yourself. Measure the CC pull-up on any A-to-C cable. Run the cable at its claimed current for half an hour and feel the connector and the cable — a cable that is warm at the plug is telling you about its termination. And verify the data rate with the devices you actually sell alongside, not with a benchmark utility on one machine.

    Cut one open. Conductor gauge, the number and construction of shielded pairs, the braid, the drain wire and how the wires are terminated inside the shell. A cross-section explains a price gap in one photograph.

    Every supplier on this site is a manufacturer, and the business scope (经营范围) check is one of the things confirmed before a listing goes live — which matters more than usual here, because cables are a category where the assembler, the wire mill and the connector maker are frequently three different companies.

Questions people actually ask

Are all USB-C cables the same?

No, and the connector tells you nothing. Cables differ in current rating, data rate, video support and whether they contain an e-marker chip — properties that require different construction and cost very different amounts to make, while looking identical. Specify each one explicitly.

What is an e-marker in a USB-C cable?

A small chip that declares the cable's capabilities to the devices at both ends. It is required above certain power and data levels, and without it a charger will not negotiate the higher current at all — so a high-power charger bundled with an unrated cable silently delivers less than its packaging claims.

How can I test a USB-A to USB-C cable myself?

Measure the pull-up resistor on the configuration channel: it must be 56 kΩ. A smaller value tells the device at the C end that it may draw far more current than a legacy A port can supply, which is how cheap cables damage chargers and laptops. It takes seconds with a multimeter and should be an incoming-inspection check.

What makes one cable more expensive than another?

Conductor gauge in the power pair, the shielding and pair construction that make high data rates work, the connector termination and how the joint is supported, and the strain relief. All four are invisible from outside, which is why they are the first things reduced.

Can I print the USB logo on my cables?

Only if the product is certified through the standards body's programme and your company is a licensee. The marks are licensed rather than decorative, so printing one on an uncertified cable is a trademark issue as well as an unsupported quality claim — and marketplaces increasingly ask for the evidence.

Are "tested to 10,000 bends" durability claims meaningful?

Not unless you define the test. Bend-cycle figures in this category come from rigs with different angles, weights and radii and are not comparable between suppliers. If durability matters, specify the angle, the load, the radius and the pass criterion, and have the test run to your definition.