0
\$\begingroup\$

Some time ago, I purchased the FNB58 USB multimeter from FNIRSI with the intention of testing some of my USB cables. Ideally, I wanted to throw away those that are redundant and/or those that pose a safety hazard.

Using ChatGPT and a few Wikipedia articles as my guides, I decided on the following program:

  • I would first test the resistance of the cable in ohms.
  • I would then compare the difference in voltage, current, and power.
  • Finally, I would check whether the protocols supported by the charger carry through the cable.

The first thing I discovered was that there were basically two types of cables - those that came with electronic tools and did not allow the protocols to propagate, and those that came with smart devices that did. The difference in resistance could vary wildly - one cable I tested had a resistance just short of 300 milliohms while another had about 2 ohms. For all of my tests, I used a 5 volt charger with an attached device that consistently drew 0.5 amps - increased resistance in the cables appeared to affect the voltage, but not the current.

Here’s where I need help: At what point do I decide that a USB cable is faulty and poses a hazard? Is there a resistance threshold I should enforce? Also, is there a way to measure data integrity?

\$\endgroup\$
6
  • 8
    \$\begingroup\$ forget everything ChatGPT told you. You're not the first to learn that ChatGPT has no idea of anything, and can just create sentences that sound well, but are completely detached from reality. ChatGPT was built to create sentences without understanding anything. I really don't understand why people insist on asking it questions, and I certainly won't invest work in debunking its nonsense: that takes real human time, whereas writing that nonsense took zero human effort. \$\endgroup\$ Commented Jul 14, 2023 at 10:44
  • \$\begingroup\$ This is not to say that your protocol isn't sensible. It's just that your first and your second point are 100% redundant. Ohm's law. \$\endgroup\$ Commented Jul 14, 2023 at 10:45
  • 1
    \$\begingroup\$ Thgere really shouldn't be a difference in cables between different types of devices. It's either standards-compliant and can carry the same data, or not. So, what you describe sounds like you got some non-standard cables. \$\endgroup\$ Commented Jul 14, 2023 at 10:46
  • \$\begingroup\$ Please elaborate - how do you mean, no difference? Couldn’t a cable transmit data incorrectly? Couldn’t a cable charge a device more slowly on account of lower voltage or lower current / greater resistance? Keep in mind: I’m really new to this 😅. Thank you for your prompt response! \$\endgroup\$ Commented Jul 14, 2023 at 11:48
  • 1
    \$\begingroup\$ @MarcusMüller Sure USB is a stadard, but do you think all cheap chinese cables are made compliant or that cables don't wear out and get damaged so they are not compliant any more even if they once were. Also devices may come with cables that are somewhat device specific such as small flexible and supports 500mA or whatever early stadards had, which obviously won't work well with new devices that support charging at 2A. \$\endgroup\$
    – Justme
    Commented Jul 14, 2023 at 13:50

2 Answers 2

3
\$\begingroup\$

In my experience, the most frequent cause of failure in USB cables is breakage of the conductors near the plug. The cable may work if the cable is positioned in one way, and fail if the cable is positioned in another.

To test for such failure, connect the cable between two devices, such as a computer and a phone, and wiggle the cable in various ways to see if the connection breaks when the cable is in some particular position. I would throw away any cable that does not maintain a consistent connection.

In my experience, the second most frequent cause of failure is the buildup of dirt within the plug of the cable. This causes the plug to not engage completely and to appear "loose". Soak the plug in isopropyl alcohol (IPA) and then using (dry) compressed air, blow any debris from the plug. The cable may work as good as new after such a procedure.

\$\endgroup\$
4
  • \$\begingroup\$ That’s certainly an approach I didn’t immediately think of 🤔. However, I can assure you that I am a very clean person, almost obsessively so. Moreover, simply wiggling a cable around doesn’t seem like a reliable way of determining whether a cable is faulty. Obviously, I won’t use cables that don’t sit correctly. But that alone won’t guarantee that a cable will play nice my electronics. \$\endgroup\$ Commented Jul 14, 2023 at 11:58
  • 1
    \$\begingroup\$ @AndrewJackson Wiggling cables is often the way to find a fault. Broken internal connections on a connector and/or mating connector can be hard to diagnose. Our company does the wiggle test as part of an incoming test for moulded cable/connector assemblies (both brand new and from the field) whilst monitoring on an automated cable test fixture. \$\endgroup\$
    – qrk
    Commented Jul 25, 2023 at 17:58
  • \$\begingroup\$ Thinking about it some more, you’re both quite right - I’ve had cables that, oddly enough, work when bent one way, but then don’t work when bent the other way. Nevertheless I still think that it’s only part of comprehensive test. \$\endgroup\$ Commented Jul 26, 2023 at 13:23
  • \$\begingroup\$ Wiggling cables... With so many USB standards with the insane amount of electronics that can go into them or not (check out this video using using Lumafield's CT scanning youtube.com/watch?v=AD5aAd8Oy84) may not be nearly enough to "test" a cable. When a cable was a thing you plug into your guitar and an amp, it was REALLY CLEAR when you had a bad cable. Now? Not so clear. Especially when the software sits there thinking for a while before actually displaying something relevant, so the latency may make it really hard to know if the wiggling was the cause! \$\endgroup\$
    – iJames
    Commented Mar 16 at 15:49
2
\$\begingroup\$

In case it might be helpful to others facing a similar problem, I have come up with a better way of determining when a cable is faulty and should be replaced.

  1. First, you should obviously check whether it provides any charge at all by simply plugging it in. If this doesn't work, the cable should be thrown away immediately.
  2. The next step would be to test the resistance of the cable using a cable tester. I found the FNB58 to be a very handy device to have for this purpose. I haven't come across a specific threshold, past which a cable should be considered faulty. In my experience, anything under 2 Ohms should be fine though.
  3. Next, it would be a good idea to see what the cable is meant for. If it's just meant to charge a device, then some pins in the cable responsible for data transfer may not even be connected. Again, it is possible to test this with the FNB58 - I used it to run protocol detection tests on my charger, which supports the vast majority of these protocols. However, if I connect to the charger with a simple charging cable, some of the protocol tests won't pass.
  4. Lastly, I would check how many Volts the cable can support. I did this by connecting the FNB58 to a powerful charger on one end and a particularly power hungry device on the other. Typically, the cables for most consumer electronics will support 5 Volts at 500 mAmps. A cable that deviates from this should prompt further investigation.

EDIT: Now that the Thunderbolt 4 and USB 4 specifications are out, it is particularly important to record both the amount of power that can pass through it (in watts) as well as the specific protocol that it's meant for. One cannot simply assume that all USB-C cables are the same!

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.