1

A bit of context: I have a 6-port/60W USB power hub which seems good to power a few raspberry pis. I've tried it with a device in each port (raspis and a banana pi) and all seemed right - pretty stable voltage at 5.13 volts, I'm not sure what else to measure.

However, watching a movie in a raspi2 with openelec and nothing else connected to the hub I got the power/temperature warning icon, and the temperature was normal.

I've read that a common reason for this icon maybe a low quality cable. Thus my question: is there an objective way to measure quality of an USB cable?

  • This might be a better fit for another site, perhaps Superusers – user62733 Feb 26 '17 at 16:41
  • 3
    @DoritoStyle We (mods) had a very brief chat about it and this does seem to fall under the headings of both peripherals and related electronics. Almost everyone using a Pi is going to be using a USB cable to power it. – goobering Feb 28 '17 at 15:05
3

The simple answer is that there is no simple way to tell.

The only objective test (which doesn't involve destroying the cable) is to measure the loop resistance, which should ideally be less than 0.25Ω.

Measuring low resistances is, unfortunately, a non trivial procedure. You need a test jig, with appropriate connectors, a current source and voltmeter. (Directly measuring resistance with a multimeter does not work reliably).

A simpler, probably more practical approach, is to measure the combined Power Supply and cable. You need a 5Ω resistor (at least 10W) as load, then measure voltage across the resistor. This is the method I use in testing Power Supply suitability, although I use a custom made cable with very low resistance.

  • Why direct measure wouldn't work? – raven Mar 6 '17 at 23:03
  • 1
    This is an EE question. You are trying to measure ~0.1Ω. Multimeters generate a small current and measure voltage. Unfortunately contact resistance, voltage generated by dissimilar metals introduce errors which exceed the value you are trying to measure. The process I suggested is the standard method of measuring low resistance, and requires 4 contacts. Available standard resistors have separate current and voltage connections. – Milliways Mar 6 '17 at 23:09
4

Yes, but you'll probably have to take the cable apart to say for sure.

The Micro-USB Cables and Connectors Specification / Revision 1.01 is the relevant spec document. Things which are out of spec may cause problems. Among the more common issues to check for are:

  • Appropriate wire gauge (see: 6.6.2 Construction / p111)
  • Good solder joints at both ends
  • Intact shielding, properly terminated to shells at both ends
  • Properly aligned, straight pins at the micro-B end

It's rare (although not unheard of) to find brand new cables with soldering issues these days. I think the most likely problem with a micro USB cable which doesn't supply enough power is that it's been made using 28 gauge wire for both data and power lines. It's preferable, for higher current applications, to use thicker (likely 24 gauge) wire for the power lines, and 28 gauge for the data lines. As a rule of thumb, aim for the thickest cables you can find.

  • Great! I've tried a cable I had advertised as 20 gauge and the icon disappeared. – raven Mar 1 '17 at 7:04
3

One fact about USB cables people tend to forget is that voltage drop is proportional to cable length. So, if you know your devices will lie in the vicinity of the hub, prefer short cables to power them. A nice expensive 24 AWG cable 2 meters long will have the same resistance (and the same voltage drop) as a flimsy 30 AWG cable of 0.5m.

Anecdotally, I have a cheap and super-thin USB cable I bought for $0.2 or so, which is only 15 cm long. It works even with my external HDD drive, which refuses to work with most low-quality cables.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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