Two DealExtreme USB power adapters fail to charge MP3 player

[nzo]

Hi all...

I was going to build a small 5v USB power adapter to charge my generic MP3 player when I'm away from my desktop PC.

Someone here suggested it was not worth building, since adapters are so cheap to buy via DealExtreme. So I thought stuff it and bought two of them: this one and this one.

I was surprised to find that neither of them charged the player.

The PC USB feed works perfectly and the player charge icon shows it is being charged.

When I use either of the wall adapters, the player LCD screen lights up briefly, bypasses the charge icon, goes straight to the play interface and "nothing" happens.

I prised open both adapters and there is definitely +5v between pin 1 and 4 of each USB socket. Thinking the problem may be the micro-mini resistors on pins 2 and 3, I removed them but charging still does not happen.

What is it about the PC USB feed that causes the player to charge but not the adapters?

Got me beat. Any suggestions much appreciated.

[IanB]

I think the MP3 player has to believe it can draw current from the USB cable. Either it can negotiate with a real computer on the other end, or it checks for particular voltages on pins 2 and/or 3 (that's what the resistors are for, they divide the 5 V down to a precise intermediate signal voltage). Most likely in your case the MP3 player does not find the voltages it is expecting to find. I know iPhones are very fussy about this, but it is news that a generic MP3 player may also be fussy.

[westfw]

What brand of MP3 player?  Some (apple) are infamous for playing a bit loose with USB standards so that they'll only recognize "real" OEM charger units.  Adafruit has a lot of experience getting their "mintyboost" charger to work with various devices, and they have posted a bunch of info about the hows and whys: http://www.ladyada.net/make/mintyboost/icharge.html (and other mintyboost documents.)

[alm]

Not sure if I'd call it playing loose, does the USB standard require devices to do anything if they get supplied with 5V without a proper USB host? I think USB would require negotiation with the host first. Just connecting a dumb 5V regulator to the USB power pins is outside of the USB spec, as far as I know. There is a USB battery charging spec, but I doubt that a cheap DX charger will conform to it. There is a slight chance that Apple does, though .

[nzo]

Sorry westfw, I haven't the foggiest idea of the brand. The player came with a sparse leaflet with no brand ID. It is 2GB, with mini USB socket. Looks like a typical Chinese Apple knock-off.

[david77]

I have seen similar behaviour with some satnavs. They charge perfectly well with the original wall wart -> USB or the original cigarette lighter -> USB adapter and also from a PC's USB port using a standard USB cable. But some do not charge from generic USB power adapters. I have never had the chance to investigate this odd behaviour as the satnavs usually are customers units.
I'd say alm is probably right.

[0xdeadbeef]

There are indeed a few devices that will only charge after they negotiated with the USB host. E.g. a Playstation 3 controller will not charge when connected to any normal USB power supply, but only when connected to a USB-Slot of a PS3, PC or any other "real" host. A similar behaviour can be observed for iPhones/iPods. While they will charge from cheap USB power supply, they won't charge from an unpowered USB hub. AFAIK the iPad will need a real USB host as it wants to negotiate the current it can draw.

[alm]

Some just need a resistor between the data wires. Not sure if the USB battery charging standard is similar.

[westfw]

Not sure if I'd call it playing loose, does the USB standard require devices to do anything if they get supplied with 5V without a proper USB host? I think USB would require negotiation with the host first.

The USB host would be within spec to refuse to supply more than 100mA without negotiation, and presumably it would be "ok" for the device to refuse to charge unless it negotiates with the host to ensure it can provide the higher current.

It's the use of resistors on the data lines to identify a dumb charger as "good" that I call "playing loose."  "We'll charge if we connect directly to a Mac and it agrees to charge us, OR if these magic undocumented pullup/pulldown values are present."  Grr.

[IanB]

It's the use of resistors on the data lines to identify a dumb charger as "good" that I call "playing loose."  "We'll charge if we connect directly to a Mac and it agrees to charge us, OR if these magic undocumented pullup/pulldown values are present."  Grr.

Some USB chargers can supply 1 A or 2 A. It would be unreasonable of a device to try drawing 2 A from a supply unless the host gave some indication that 2 A was available.

[westfw]

Some USB chargers can supply 1 A or 2 A.

This is also "playing loose with the USB specification.)  The spec says 500mA max, period.  (actually, I don't know if the spec allows for a charger-only "host" at all (ie with nothing connected on the data lines.))

And USB.ORG thought they were powerful enough, and had a reasonable enough spec, that it wouldn't get as widely violated as the old serial and parallel ports.  Hah!

[IanB]

Some USB chargers can supply 1 A or 2 A.

This is also "playing loose with the USB specification.)  The spec says 500mA max, period.  (actually, I don't know if the spec allows for a charger-only "host" at all (ie with nothing connected on the data lines.))

I would argue that when you are supplying power from a dedicated adapter and have no data you are not really using USB, you are just using a power cable with USB connectors on it. Those devices that might draw 2 A from a dedicated charger will limit themselves to 500 mA when connected to a real computer or USB hub. They comply with the standard when connected to standard conforming devices.

[NiHaoMike]

The data lines are just connected to each other to identify a "standard" USB charger. (Apple, however, uses resistive voltage dividers.) 2A or so appears to be the actual current rating of a standard USB connector since that's the highest I've seen in common use. I'm not sure why they decided to pick 500mA for 1.1 and 2.0. They bumped it up to 900mA for 3.0, which really is too little.
http://www.maxim-ic.com/app-notes/index.mvp/id/4803

It's also possible to use several ports to get enough current, but it's apparently too weird to be accepted by mobile device manufacturers.

[westfw]

too little.

How thick did you want your cables to be?
http://genuinedealz.com/voltage-drop.html  (0.5V drop over 10ft of 24g wire; pretty significant.   I see advertisements for 20g USB cables, but I wouldn't want to use them everywhere, and...  The USB folk took all sorts of this stuff into consideration; you're not supposed to need "special" USB cables for different devices.  Any official spec USB cable should work with any device.  (Of course, there are lots of non-spec cables out there these days too.)  (Your mouse is allowed to have a tiny, flexible cable because it's permanently connected.)