We continually make minor tweaks to the USB output of our batteries to make sure we charge as many devices as possible. We pay particular attention to Apple products and now, with the introduction of the iPad, it has become slightly more complicated to have a one size fits all solution. This post tells you the Voltage that we use on each pin in current versions of our USB battery packs. We make these battery packs available standalone and as part of our portable solar chargers. Hopefully, this will be useful if you’re trying to make your own USB charger.

The USB receptacle or female connector has 4 pins:

Pin Wire Color Signal Name Description
1 Red VCC +5VDC
2 White D- Data -
3 Green D+ Data +
4 Black GND Ground

In all cases, pin 1 is 5V (with a little wiggle room, spec allows from 4.75 to 5.25 V, and we aim for the upper end of that range to compensate for voltage drop due to high current draw) and pin 4 is 0V. But, depending on the configuration of the two data pins AND the current draw of your device, a USB output may charge all, some or no devices.

We ended up using the following Voltage configurations for current versions of our batteries:

Battery USB Port Current D- D+
V11 USB Battery Pack 900mA 2.75V 2.0V
V39 USB Battery Pack “low-power” port 900mA 2.75V 2.0V
V39 “high-power” port 2,000mA 2.0V 2.75V
V60 Universal Laptop Battery 900mA 2.75V 2.0V

Here’s the logic and the tradeoffs associated with each of the decisions.

The V11 and the V60 both have a single output port so we need for those ports to charge every device (note: as of early 2012, all V11s have 900mA output instead of the previous 650mA). Only the 2.75/2.00 V (D-/D+) output charges both the iPhone and the iPad consistently across all models. If we went with 2.00/2.00 then it would charge iPhones but not all iPads. The risk with this configuration is that once or twice in hundreds of tests the iPhone 4 says “accessory not recognized” part way into the charge cycle. The solution is to disconnect and reconnect the charger. We don’t expect this to occur much in the field, but it is the tradeoff we needed so that you will be able to charge your iPad with its own charging cable from the V11. If you get have problems charging an iPad from our current V11 battery, you can use our iPod adapter which sets the right Voltages.

We kept the same 2.75/2.00 configuration for the V60 with its 900 mA output, but didn’t see any issues surface in testing with the iPhone at this higher current.

On the V39, included with our iPad solar charger, we have both a high and low power port so we chose the same 2.75/2.00 output for the low-power port and 2.00/2.75 (the same voltages supplied by an iPad charger) for the higher power port, which is specifically designed for tablet charging.

Update: The iPhone 5 works across all three pin configurations.

Note: A general observation with the Apple products is that they will attempt to draw the maximum current if and only if the voltage level on Pin 1 with respect to the current draw remains constant between 5 and 5.25 volts.  Often, if a charger is unable to supply the proper amount of current the voltage output will drop.  For instance, a 5.25V 1A power supply may only output to 4.5V when under a 2A load.  For Apple products, when a device is presented with any one of the voltage configurations (500mA, 1A, or 2A) and then presented with a range of Pin 1 voltage levels from 4.5V to 5.25V, the actual current drawn by the device varies.  In our tests, we found that when the voltage sent to the iPad 2 was 4.5V, the iPad 2 drew only about 1A, but steadily scaled up the current to 2A as the voltage was incrementally increased to 5V.  Additionally, we noted Apple devices will not accept charge from power supplies with Pin 1 voltages of less than 4.5V or greater than 5.5V.

About The Author

28 Responses

  1. ScottInNH

    Nice writeup on the voltage combinations.

    Is it not possible to support multiple USB charge modes using switches? That is, switch position A routes power through one resistor combination (mode), and switch position B through another set of resistors?

    This would seem to be getting your cake and eating it too. Or is it not that simple?

    Reply
    • admin

      Thanks.

      You certainly could, but our sense from testing was that we didn’t need to make that step. If you’re talking about a manual switch, we would like to stay away from that as it makes the battery interface more complicated and adds cost. You could do it automatically after sensing the current draw as well, but our experience with these designs is that the implementation takes way longer than we expect and downstream problems always crop up. Better to keep it simple if at all possible. We’ll also see what happens with the iPhone 5…

      Reply
  2. Nate

    2A is quite a lot of current to shove over 24AWG or whatever the power conductors in USB typically are. Of course you’re going to see voltage drop, even over just a few feet.

    However, USB-like cables are available with additional wires, so it might be interesting to use another conductor as a “sense lead”, connected to V+ at the output connector, allowing the regulator to compensate for voltage drop in the cable.

    (You’ll see the same drop over the GND conductor too, so either use a second sense lead, or double the compensation.)

    Reply
  3. HansSchulze

    I would love to see a quote of the USB spec that allows or supports these levels.

    Reply
    • tommy

      Hans,

      We would too. USB 3.0 spec only supports the 900 mA output that we use as our default. Unfortunately, in order to keep up with the demands of devices that are getting ever-power-hungrier (particularly tablets), manufacturers have introduced their own ‘standards’ that don’t actually fit the USB spec. From their perspective, it allows their devices to charge faster and also forces customers to buy their chargers. A selection of recent tablet AC adapters reveals a number of different resistor configurations (and resulting data pin voltages).

      As a solar charger company that wants to remain competitive and keep customers happy, we are forced to constantly evaluate new devices and redesign our products to maintain compatibility. In an ideal world, a standard meets everyone’s needs and everyone uses it. It would certainly make our jobs easier to stick to a spec and know it will charge everything. However, that’s just not the case – hence our V39 battery featuring a non-spec “high-powered” USB port.

      Thanks,
      Tommy

      Reply
      • HansSchulze

        There seems to be a new USB power spec here
        http://www.usb.org/developers/powerdelivery/
        that talks about raising the power ceiling to 100W.
        I haven’t read it yet.
        It seems that some devices use pin1 as a Kelvin sense point to counteract the inability to use >24GA wire on the micro usb connector. Even with 24GA we see almost 0.1V/A/ft of drop across the power pair. Since the target voltage cannot drop by more than 10%, you need to make sure you find ways to beef up connector or double up the 24GA wires.

  4. NigelF

    Have you had a chance to have a look at the new 4th generation iPad which comes with a 12 Watt charger, rather than, I believe, the 10 Watt version provided with the iPad 3 – or is the extra capacity in the new charger to simply provide a bit more head room in case the iPad tries to take a bit more current?

    Reply
    • admin

      Hi Nigel,

      Our output is still at 10 Watts. I think they bumped this up because there were reports of the retina iPad not charging with 10 Watts if it was in heavy use (meaning lots of power consumption). We have not revised up, as we’ve found that the iPad can charge all the way down to 0.9A / 4.5 Watts. As you go lower than 12Watts, the risk is that in some (but not most) situations you may be providing less power than the iPad retina is drawing. At a minimum you are keeping the iPad steady or slowing the decline. However, we find that 10 Watts still works very well and 4.5 Watts will do, especially if charging in sleep mode or if your goal is to extend the runtime.

      Jeff

      Reply
  5. Armand

    Hello,

    I have done some testing and found a single resistor of less than 300K across usb data pins will trick the ipad and iphone (even the latest iphone 5), into thinking its a high output charger. Do you see any reason not to go that route?

    Reply
    • admin

      Hi Armand,

      If it works for you, go for it. The one caveat that we have is that we’re trying to not only do the iPad, iPhone, but also all other devices that charge from USB (Samsung, HTC, Blackberry phones, etc.). For us, the pin Voltages we ended up with work well across those devices as well.

      Jeff

      Reply
  6. Nigel Farmer

    I’m probably doing something really, really dumb, but I found my old USB breakout box and out of interest got my old iPhone 3GS charger out and started measuring voltages on the USB output(all with respect to pin 4 (Black)
    Pin 1 (Red) 5.05V
    Pin 2 (White) 2.64V
    Pin 3 (Green) 1.95V

    So OK, allowing for its age (it’s an original) and a bit of change in resistance, broadly as expected.

    I then got my wife’s iPhone 5 (ie Lightning, not 30 pin connector) charger and found:

    Pin 1 5.0V
    Pin 2 0V
    Pin 3 0V

    I was of course expecting the same as for the 3GS charger (the appearance is the same but the model number and description are too small for me to read….)

    Lastly I got my 12 Watt, iPad 4 (Lightning interconnect) charger and found:

    Pin 1 5.1V
    Pin 2 2.66V
    Pin 3 2.66V

    It would be great if anybody else who has similar kit could check to confirm or discredit my results; suggest what I’ve ignored or doing that’s really, really silly…

    Otherwise, have Apple moved the goalposts again?

    Nigel

    Reply
    • Eric

      Nigel:

      In your iPhone 5 charger, are pins 2 and 3 shorted together, floating separately at 0V, or shorted to ground?

      FWIW, my old 3GS charger is close to yours: 5.11V, 2.74V, 2.05V. I also have a Belkin charger from that era that’s nearly identical: 5.11V, 2.74V, 2.03V.

      —–

      Voltaic:

      I’d also be interested to hear your take on Apple’s new data pin voltages.

      My iPad 4 charger measures 5.19V, 2.77V and 2.78V. However, the old 2.00V/2.75V arrangement still works with my iPad 4.

      My question is, will Gen 1-3 iPads charge from the new 12 watt charger, and if so, can we conclude that the discrimination is done by software and can be changed by Apple through iOS updates?

      Reply
  7. Pete

    I bought an adaptor at Sun n Fun to charge my ipad 2 in my plane. It charges at a high rate. I plug it into the cigarette jack. When I plug the ipad in, the squelch on my radio goes on and I hear white noise. I unplug the Ipad and the noise goes away. Any suggestions?

    Reply
    • admin

      Does it happen with other loads, like for example your phone (that presumably draws less current) or only the iPad2?

      Reply
  8. dennyhayes

    What I can’t figure out is why putting a 2 amp 5 VDC power supply on the number 1 and 4 pins of a USB socket will not charge my iPhone. My VOM shows the 5 VDC is there, even with the load attached, but neither the Iphone 4 nor 5 recognize it. The cable is fine, because when I connect it to am Apple power module, it charges fine.

    Reply
    • alberth

      Apple i-devices will refuse to take any USB power unless it sees DC voltage on pin2(D-) and pin3(D+) in precisely the way outlined in the article. Even if a solid 5V is on pin1 and ground on pin4.

      If I understand right you identify your
      current sourcing ability via setting voltages
      on pin2 & pin3 as follows.

      current: pin2/pin3
      900mA: 2.75V/2.0V
      2.0A: 2.0V/2.75V
      500mA: 2.5V/2.5V (i think?)

      Reply
  9. Andy

    “If we went with 2.00/2.00 then it would charge iPhones but not all iPads”

    I’m building a system for an extended bike tour using a 18v Unisolar battery charger panel and a duracel usb car charger. I hooked the panel to the charger and got the readings
    Pin 1: 5.1
    Pin 2: 2.0
    Pin 3: 2.0
    Pin 4: 0.0

    Will the following work (with D+/D- at 2.0/2.0) to charge my iphone 4? Is there any chance of damaging my phone by trying it?
    Thanks

    Reply
    • admin

      Hi Andy, as long as those are stable, i.e. there are no spikes, then you won’t harm your iPhone with 5.1 / 2 / 2 / 0. There is a chance it won’t work if the panel is not putting out enough current, but you won’t damage the phone. Good luck & send us pics. /jeff

      Reply
  10. Callum B

    Is there an allowed tolerance? Using a voltage divider calulator, I can get 5v down to 2.747v using an 0R82 resistor as R1 and 1R as R2. Unless someone can get a better combination, would that be alright? Getting it to 2v is fine.

    Reply
    • Ben

      making a divider out of 0.82 ohm and 1.0 ohm will draw a lot of current and the resistors will get hot. Please use resistors in the 10 kohm range (10,000 ohms+)

      Reply
  11. Eli

    Hello,

    I am a bit confused to how do you get the 2.75V/2.00V to the center 2 pins of the USB? What is the process? I am trying to create a portable battery charging device capable with the ipad.

    Thanks
    Eli

    Reply
  12. Robert

    I’m a pilot and use an iPAD in the cockpit running aviation applications that require WiFi and Bluetooth running simultaneously and the display at maximum. 12 watt chargers won’t keep up. I’ve designed some 28v (aircraft bus voltage) to 5v 3 amp supplies but can find the appropriate D+ and D- voltages to sustain higher rate charging. Any ideas? Thanks.

    Reply
  13. Phil M

    I can confirm that the voltages listed above result in the stated charging currents through my own testing. My results see an iPad 2 and iPhone 5 both seeing: 0.5A with 2.0V / 2.0V (D+ / D-); 1.0A with 2.0V / 2.8V (D+ / D-); 2.1A with 2.8V / 2.0V(D+ / D-).

    Both devices also report a 1A charger when D+ and D- are shorted and left to float.

    Reply

Leave a Reply

Your email address will not be published.