USB Cables... Why are some of them bad for Raspberry Pi?

Question

Yesterday I bought another Raspberry Pi, and installed Rasplex on it. While using Rasplex, I got this colored square on the top-right corner, indicating that the power supplied to my Raspberry Pi is not sufficient, although I was using an Anker power supply, 36 Watts, which is supposed to be more than enough.

With some experimentation, I found out that out of a few Micro-USB cables, some had this problem, and some didn't. Then I found some forums confirming that USB cables can be a cause for the problem. Right now I'm using an original Nokia cable, and it works fine.

My questions:

1. How does a cable technically matter here? Is it that its resistance is too high that a strong voltage drop happens across it? If that's the case, why can't manufacturers just fix this issue? How hard can it be?

2. How can I know whether a USB cable is good before testing it with my Raspberry Pi? Is there a specific way to test it? What about other USB cables, like Mini-USB? Can I measure some resistance or capacitance there to figure it out?

Answer 1

Everything for sale is built to a price.

Tha manufacturer wants to make a profit.

There is no need for many USB cables to carry more than a fraction of an amp so they would be built with thin wire. Thin wire is cheaper and lighter than slightly thicker wire. All other things being equal thinner wire has a higher resistance and can carry less current and will drop more volts.

Answer 2

Cables do make a difference once you start to draw hundreds of mA or several A.

You can often tell how much current a cable can carry simply by its looks. Anything thick and stiff is good, thin and overly flexible may cause trouble. Check out these pictures to see what I'm talking about:

VS

Cable length also has a similar effect: Shorter cables are much better for high current draw compared to longer ones.

Answer 3

Popular things like the Raspberry Pi surround themselves with Urban Myths. One such is the USB cable Myth.

That is not to say there is some grain of truth, there usually is, but people jump to conclusions.

1. USB2 is specified to supply 500mA max.
2. USB power is 5V ±0.25V

To remain in spec there should be less than 0.25V drop which corresponds to loop resistance of 0.5Ω.

It is possible to measure low resistances, but it is not straightforward. There is the problem of making contact with the small connectors. Then reliably measuring without contact resistance. In practice a test rig which feeds current through the cable and measurement of the voltage drop across the cable is necessary.

It is much easier in practice just to measure the voltage at the Pi end. This is not the whole story, as you ideally need to measure the voltage at the supply end.

There are confounding factors.

The Pi and many modern smartphones are designed to draw more current than the 500mA max. Phone manufacturers can supply higher current chargers, either by non-standard means or by adopting the new USB Charger spec, which permits higher currents, but permits voltage to drop to 3.6V. These are OK for charging smartphones, but NOT for voltage sensitive devices like the Pi. (I suspect the referenced charger is one such).

The microUSB connectors are easily damaged. The typical environment in which they are used by young people is conducive to such damage. This leads to poor or unreliable contact, and potential voltage drop.

Use of an older USB2 data cable, particularly one not designed for charging can lead to problems. If this is combined with an unsuitable charger the problems compound.

Answer 4

To actually answer the point 2 of the question: Just read the cable specification!

Resistance is proportional to the length of the conductor, conductivity of the material and the diameter of the conductor. Since the current our power supply provides should be DC-ish, we don't need to worry about skin effect. That also means that we don't have to worry much about the capacitance and inductance of the cable.

The conductivity of the material will often be fixed, since cables should usually be copper.
As many others have noted, length should be as short as practically possible.

That leaves us with the diameter.
It's often actually written on the cable itself. The outer jacket of the cable will have inscriptions on it and usually you'll see temperature and voltage ratings, insulating material and so on, stuff many people don't bother to read.

Among those things, you'll see the description of the actual cables used within out bigger cable. For example, one cable I have says 28AGW/1P 28AWG/2C.
This means that internally, the cable has one 28 AWG twisted pair and two 28 AWG conductors. The AWG stands for American Wire Gauge. With AWG, higher the number, thinner is the wire you get.
Therefore you should look for cable which has a low AWG number for the /2C part. For example, you may see 26AWG/1P 24AWG/2C or maybe even 24AWG/1P 20AWG/2C (I haven't seen such micro USB cables myself, but I've heard legends of them).

It's a shame that USB cables usually give little information about the quality of the connectors used, because they will also have some impact on the resistance of the cable.

One other trick, that's indirectly related to cable, is to carefully read specifications of the power supply. USB power supplies will usually be 5 V nominal, since older USB expects 5.00 V± 0.25 V. What some power supplies do is try to take into account voltage drop of the cable when having their output voltage designed and being really tight on the high-end side of their internal voltage tolerances. For example, I have a Samsung EP-TA10EWE power supply which has nominal output voltage of 5.3 V. By the time this reaches the device, we'll almost certainly lose the 0.05 V and be in the USB specification again.

Also please keep in mind that although your particular power supply has rating of 36 W, the website you linked to also has rating of 2.4 A per port!

Answer 5

I'd reiterate @joan's answer, but the trouble you have is not something that can be avoided just by looking at the package - other than as an assessment of the care the manufacturer would have made in producing it (the country of origin may have some bearing on this! 8-) )

You may be able to measure the resistance of each contact in the lead from one connector to the other (if the two wires that carry the supply {the outside pair on a standard USB connector} have the same resistance the "volt-drop" in volts you get will be twice that resistance in ohms multiplied by the current in amperes) - but you would have to open the retail packaging and have a means of reliably making contact - I can't see a retailer really want to permit you to rip open the box in their place of business - and the probes on the average Digital Multi-Meter are a bit large to make contact with the terminals on, at least, the micro-USB connector. One thought is that leads for use with an Apple iPad or other tablets may be more likely to have the thickness of wire needed as they are "known" to need high current supplies - just expect to pay for that branding!

It might be possible to get a refund from your supplier depending on where you are in the World, for instance, in the UK we have this legislation known as The Sale of Goods Act which may offer you the chance to seek recompense from a business selling you something that is not fit for its intended purpose - you will just have to be able to prove that your intended use is a reasonable one, but getting the supplier to see that can pose a challenge for some values of "supplier"!

Edit: I just noted that you referred to the "low-voltage warning square" which is a feature on the later models (Pi+ and Pi2) which I understand have better (lower) current requirements than the originals. So if a lead is not good enough for those later models it is even less likely to be safe for the earlier ones without that warning.

Answer 6

There's also the problem of cable length. One of my cables delivers 1.5 A while a similar cable which is twice as long delivers only 0.7 A. More length means more internal resistance thus a lower amount of Ampères delivered.

Answer 7

Others have already answered most of this, but I think this page on testing USB cables from someone who actually carried out detailed measurements of cable performance contains more insight. Quoting: "It's essentially impossible to reliably power a Pi off a cable over 1m long using a USB cable. You simply cannot run 5V through that length of thin wire at the current levels required by a Pi, the losses are too high. If you do need to power one more than 1m from the power source, run 12V through a proper barrel-jack power cable and use a UBEC next to the Pi to take it down to 5V".

They've also tested a ton of different micro USB cables, ranging from very good (a tiny handful) to mediocre through to downright awful (one that's described as "a long thin resistor, not a cable"). Worth reading as a longish technical answer to the question.

Answer 8

At work, I play a lot with Raspberry Pi, USB devices and mobile phones.

I think your questions have been answered by others in this thread; USB cables are manufactured either for a specific purpose (a certain phone model for instance), or according to USB specs.

My first advise: if you buy a high power USB charger, get a specific cable for it. Don't be cheap on this, there are plenty other ways to save money. And get the shortest cable that still work for you.

But reasonably, you should not need to have a 2.4A USB charger for running the Raspberry Pi. What USB devices do you have hooked up to the RPi?

In our application, we try to put as many mobile phones on a Raspberry Pi as possible. It turns out that 2 mobiles is the maximum, if at most one of them is a modern Android or iPhone. That's because when they start charging their batteries, they really suck up all the juice they can get. And in this situation, never pull out/plug in a mobile while the RPi is running. This will kill your SD card, eventually. If not more.

My second advice: If you need that much power for your USB devices, get a self-powered USB hub that does not back feed power to the RPi. Supply power for RPi and devices from different PSUs.

My third advice: Get a new power supply after a while. Some of our RPi's are on 24/7. All year. We've bought expensive 2.1A power supplies that eventually go belly up because they are not built for that kind of application.

Answer 9

I was shocked once I found out, that some USB cables would not provide as much power as others to the Raspberry Pi.

We have an Android phone connected to the Pi, and when its battery is mostly flat and its browser is rocking on 4G/LTE, then your battery could be drained faster than it can be recharged by the Raspberry Pi. So I bought a simple USB power meter and found out, that one specific noname USB cable was really a bottle-neck - the meter wouldn't go any more than 700 mA. I am using Anker cables now, and the meter is somewhere around 1,3 A with the phone plugged into the Pi.

In my setup, there is actually 2 USB cables: 1) is supplying power to the Pi from a USB charger, and 2) is connecting the Phone to the Pi (data via adb and charging the phone). Obviously the cable 1) has approx. 250 mA more load (the avg. current of the Pi).

Also, I would like to mention these settings in /boot/config.txt found on the Pi:

safe_mode_gpio=4
max_usb_current=1

Make sure to read all details about it here