iPhone 3GS - "Charging is not supported with this accessory."

[jvschwarz]

jsappo,
Have you charged the 3GS with the Minty Boost? In what configuration? If your correct, totally gnarly dude. But the question still persists, Why only the minty boost? Several non-OEM chargers work.

What I have tried since I last posted are, 3 cells. Helps the voltage 4.97v and D+/- @ 1.97 under load but still "not supported." I reviewed the LT1302-5 data pdf and thought I'd try adding the 4.7 k ohm resistor after c1 and a 330 uF cap between positive and negative like on page 10 but still "not supported." I was hoping that the reduction in undershoot help. Next thought is the boost disable. Any thoughts Ada? Would using a zener diode to control the D+/- voltage help? Since I don't have a signal generator and Ada's 3GS' all work, has anyone tried controlling the 2.0 volts on D+/- while using the Minty? Last thought is inductance, any help?

[adafruit]

jsappo, Have you charged the 3GS with the Minty Boost? In what configuration? If your correct, totally gnarly dude. But the question still persists, Why only the minty boost? Several non-OEM chargers work.

it's been our experience that the 3gs units are not consistent, we can use it for days with a mintyboost and then it will start giving the "Charging is not supported with this accessory" or it will start saying that with other chargers as well. the dock connector replacement solution is something we're going to try for sure.

cheers,
adafruit

[jvschwarz]

Understand. I'd definitely be game to try with a little encouragement. Waiting your report.

[Orwyn]

Unfortunateyl I am the next victim of Apples charging policies.
My iPod Touch 3rd Generation reports "not supported accessory"

Most likely apple has changed voltages for this devicees, too.

Did anyone find out the matching values fpr the Resistors?
The minty boost seems to be ok, my creative Zen charges hapily.

[jonbill]

I've just build v3 MintyBoost. I've got fresh batteries in it. I get a voltage of 4.97v. It works with my wife's iPhone 3G but not with my iPhone 3GS. I get the message Charging is not supported ...

The 3GS is running iOS 5.0.1

I've also tested the voltage unconnected and get the same 4.97v, whereas the tutorial says it should be 5v or a little over. I also tried bridging the data pins as suggested in the tutorial.

I'm pretty experienced with assembling electronics kits. It's unlikely to be a problem with that.

For a full voltage test I get this:

Looking at the bottom of the board, curved edge facing me, numbering the USB pins from left to right 1, 2, 3, 4 I get the following:

Black lead on 1, red lead on 4, 4.97 volts, connected or disconnected

2 to 3, 0v

Black lead on 1, red lead on 2, disconnected = 1.94v, connected = 1.8v

Black lead on 3, red lead on 4, disconnected = 2.9, connected = 3.04

I've tried every cable I have. The original Apple cable and several third party cables.

Very disappointing as I intended to use it to support tethering my iPad which sucks batteries quickly.

The only other things I can think of is to replace the IKEA batteries I am using with Duracells, but I'm not too hopeful.

Any suggestions?

Jon.

[adafruit_support_bill]

The unloaded voltages sound good. But it does sag a bit under load which is probably the problem. Others have had problems with Ikea batteries. They are probably standard alkaline chemistry and not designed for high-drain application such as this. Check the battery recommendations here: http://forums.adafruit.com/viewtopic.php?f=15&t=22377

[Knipex]

Hi there,

I have built up a circuit for charging my iPhone 3GS (bought in Germany) via hub dynamo while riding my bycicle.

The circuit works and creates a constant output of 5V at a speed > 10km/h (approx. 6mph).

I can charge all other devices without problem except my Iphone 3GS. -->The display shows the well known yellow triangle and: "Charging is not supported with this accessory."
So I applied the additional resistors at D+ and D- as described under the following for the MintyBoost "iCharging": http://www.ladyada.net/make/mintyboost/icharge.html
I expected a voltage of 2.0V (49.9K / (49.9K+ 75K) * 5.0V = 2.0V) at D+ -> Ground and D- -> Ground but there were only 1.85V

I think that has something to do with the input impedance of the iPhone so I increased the "49.9K" resistor to 62K which lead to a measured voltage of approximately 2.0V at D+ -> Ground and D- -> Ground.

But still the iPhone does not charge an and still shows following message on the display: "Charging is not supported with this accessory."

jsappo has posted that by renewing the dock connector the failure is gone. But he hasn't wrote about the MintyBoost circuit in particular.

Has any of you guy's had an iPhone 3GS which could not be charged with the MintyBoost (but with other devices) and after changing the dock connector charging progress worked with Minty Boost?

I just don't want to buy a new dock connector without seeing the chance that could fix the problem.

Thank you for your answers!

[Shadowww]

iPhone is pretty sensitive to DC output 'quality' - don't hub dynamos have horribly high ripple?

[Knipex]

Yes, you are right the hub dynamo provides only AC (AC but not sinusoidal) output which can not be used for charging the iPhone.

But by using a bridge rectifier and a smoothing capacitor in front of the 7805 voltage regulator I achieve a constant 5V DC output.

I'm using (with small changes) the circuit shown at the end of the page in the following link:

http://home.vrweb.de/~k-d.redeker/lp/PowerSupply.htm#sb
(German Webpage)

The circuit works - I can charge other USB gadgets but not the iPhone 3GS.

Even if I apply the resistor circuit at D+ and D- mentioned in my comment above.

[adafruit_support_bill]

Does the 5v output hold up with the phone connected? These phones will try to pull a lot of current. If the voltage sags under the load, you will get the "not supported" message.

[Knipex]

No, the 5V at the output remain stable. I'm using a 3W hub dynamo.
At the moment I do only supply the charging circuit including the iPhone. The light of the bicycle is currently not connected to the hub dynamo so the full power could be used for charging.
The oscilloscope shows me a stable 5V output even when the iPhone is connected to the circuit.

[dzfan]

Regarding the mintyboost not working with iPhone 3GS:

I was messing with another SMPS car USB charger that didn't work with my iPhone 4, and I saw that they did not have anything connected to D+ or D- at all (there were no traces connecting to the pins). Apparently as of 2007, the USB standard for a "charging port" (supplying only power with no data) is to have D+ and D- shorted together, but it doesn't say anything about applying a voltage bias to them. The voltage bias appears to be Apple's addition to the standard I think. Perhaps when the 3GS was released, they were still messing with what Apple products would like to see from the charging port.

Although adafruit's design holds D+ and D- at the same voltage level with separate voltage dividers, they're not physically shorted together. This could possibly be a factor.

I experimented a little and saw that shorting D+ to D- alone without biasing them was enough to get my iPhone 4 to start charging, and it drew 500mA. My iPod nano 2G did not like that, interestingly, and drew about 100mA. Then I added some surface mount resistors (47k and 40k, package size 0603) between the leads to get 2.7V on D+ and D-. I used surface mount resistors because those fit rather nicely without having actual pads to lay them on or any extra through holes for new resistors, and I had access to them at work. Mintyboost has through holes for resistors, so you don't have to worry about that (but that's a good way to upgrade an old power supply). I didn't bias them at 2.0 V because in this case I wanted to pull 1000mA from the charger instead of just 500mA.

My observations:
1. It is not necessary to bias D+ and D- separately. They can be shorted together and you can use one resistor divider down from 5V together to bias (only 2 resistors instead of 4).
2. Although adafruit observed that Apple's charger uses D+ = 2.0V and D- = 2.8V, I observed that when (D+ = D-) < 2.4V or so, the current limit was 500mA. When (D+ = D-) >= 2.4V or so, the current limit was 1000mA. Maybe Apple looks at the average voltage?

I don't have an iPhone 3GS, but I would make the suggestion of shorting D+ and D- together and biasing them together at 2.0V and seeing what happens. Maybe the tolerance mismatch between the two voltage dividers is enough to not follow the "charging port" standard of having D+ = D-. Shorting them will guarantee they're the same. I also don't know if Apple devices check for the impedance between D+ and D- instead of just the voltage, which would definitely be much different if they were shorted together vs. not shorted. You might also try shorting them together and not biasing them at all (remove the resistors).

If anyone tries shorting D+ and D- together on the mintyboost and tries charging on an iPhone 3GS, I'd be interested to hear what happens.