Mfcttrf

How do we test and determine if a USB cable+connector is version 2, 3.0 or 3.1, without using port, PC and/or having data transfer but instead with basic physics tests e.g. using multimeter?

First, there is no real thing as "version" of USB cable. USB cables must conform to certain electrical requirements to be qualified to work reliably at corresponding data rate.

In case of USB 2.0 the task is relatively easy - the cable/connector has only four wires, GND, D+, D-. VBUS. And maybe one locally-connected ID pin inside mini- and micro-USB plugs.

You probably want to distinguish if a cable is good for USB 2.0 HS, or only for FS. This can be determined only by measuring the cable+connector electrical transfer characteristics, primarily the differential characteristic impedance, and the cable's ability to deliver sufficient signal at far-end of the cable (when loaded with a standard 90-Ohm diff. load) If the cable characteristic impedance along the link (especially along solder wires inside connector molds) meet the specification (90 Ohm +-10%), and cable loss doesn't violate the far-end eye diagram template,

the cable is good for use in USB 2.0 480 Mbps transmission applications.

Unfortunately, a multimeter is far from being an adequate tool for doing this kind of measurements. To make these measurements you need (a) TDR (Time-Domain Reflectometer) or VNA (Vector Network Analyzer), and (b) an oscilloscope with a differential probe with effective bandwidth of no less than 2 GHz.

You also will need a test fixture that smoothly connects the USB female connector with coaxial cables used by high-speed instruments, which typically use SMA connectors.

For USB 3.0 signals (Gen1 5 Gbps) and USB 3.1 Gen2 (10 Gbps) the approach is essentially the same, just the frequency requirements are a lot higher - the oscilloscope must go up to 16 GHz, which dramatically increases the cost of test equipment. In addition to measuring the characteristic impedance, the USB 3.x cables must meet additional requirements to have low cross-talk, since USB 3.x cables carry at least two pairs of signals going in opposite directions, so their interference should be minimized.

In general, USB Implementer's Forum (USB-IF.ORG) has a library of documents including Electrical Test Specification for cables, here is one of the documents.

This is an interesting problem, as there are now a proliferation of similar-looking cable types with the introduction of USB 3.x and the Type C connector. Determining how many pairs are connected, and if it supports flip-over and other compliance issues is non-trivial. There's a lot of mis-marked junk out there and having a reliable way to sort it out would be useful.

Have a look here to see how dire the situation really is: https://www.engadget.com/2016/02/03/benson-leung-chromebook-pixel-usb-type-c-test/

I'm going to assume you're interested in testing USB C cables, since it's possible to suss out previous USB 2.0 vs. 3.x types by the connectors types they have on them. Then again, a test for these non-C cables might be useful too. If you work out a test for Type C, the others are subsets.

Start by getting the actual USB 3.2 spec from USB-IF, here: https://www.usb.org/document-library/usb-32-specification-released-september-22-2017-and-ecns

As a basic test, using a breakout board at each end and checking end-to-end with a multimeter would be the way to do it. You would check for continuity on each possible signal pair:

• USB 2.0: DM/DP


• USB '3.0 / 3.1' (one SS pair): DM/DP, SSTX+/SSTX-, SSRX+/SSRX-


• USB 3.2 Gen 2 (two SS pairs): DM/DP, SSTX1+/SSTX1-, SSRX1+/SSRX1-, SSTX2+/SSTX2-, SSRX2+/SSRX2-

As an improvement you could use a TDR reflection method, like some LAN cable testers use. This would also tell you if the cable is the right impedance.

More about TDR testing here: https://www.electronics-notes.com/articles/test-methods/time-domain-reflectometer-tdr/what-is-a-tdr.php

Example LAN cable tester here: https://www.cablestogo.com/product/39002/lansmart-tdr-cable-tester

Certifcation-level testing here: https://www.keysight.com/upload/cmc_upload/All/USB_3_0_test_solution_overview.pdf

If you're thinking to make a product I think making a continuity tester board would be useful. The tester would terminate the cable at one end, and test for continuity at the other using a microcontroller to measure it. This would be able to reliably distinguish if the cable has the right number of of pairs.

Example commercial USB test product: https://www.totalphase.com/blog/2018/09/tools-can-use-test-analyze-functions-power-delivery-usb-type-c-cables/

It would be good to also check shield continuity as a separate test. It's supposed to be isolated from signal ground, but still present, and it's an area where many shoddy vendors will cheat.

(To the other posters - think about how you would do incoming inspection / acceptance on cables you bought for your product. You'd do something like this, and maybe even take the cable apart too.)