This is an edited version of something that's on my Google Plus stream. Hoping to find a couple of answers here among experts.

Backstory:
In the past, I'd cobbled together a single in-car dock for my iPhone 4 that could do all of the following from a single plug:
- Aux-out from the phone to my empeg through the dock connector.
- Charge the phone.
- Charge my Apple-brand Bluetooth headset that I lurve and that they don't make any more.

The last one was the tough one, because there's only one plug that'll do that trick: The one that comes with the headset. This is the plug I'm talking about:

Apple iPhone Bluetooth Travel Cable

That plug is special: No third party company that I know of makes one of those. You have to get it from Apple. The problem is that it doesn't do aux-audio out. You can get a billion different plugs and docks that will charge the phone and do aux-out, but not a single one that will also charge the Bluetooth headset.

So, wanting a super-clean installation in my car with as few protruding bits as possible, that's the plug that I'd modified to be an in-car docking connector. But like I said, the trick is that it doesn't do aux-audio-out. The necessary L/R aux-out pins (3 and 4) aren't even present on the connector. So what I had to do was dismantle the plug, and cannibalize the data pins to become the aux-out pins. In addition to grounding Pin 11 to force the audio to the aux-out... that pin-11-ground trick is what you did on old iPod cables to make the aux-out work.

That all worked! At least at first, before the iOS5 upgrade it worked. It's super-clever. Combined with a super low-profile phone mount on my dash (that I constructed from scratch with my housemate Fishy's help, better than any commercial mount I've seen), it made the iPhone a perfect combo GPS and Pandora-player box (God, I'm glad I'm grandfathered into AT&T unlimited data). It sits just behind my steering wheel and just below my line-of-sight to the windshield, including the headset charger all on the same plug. It was brilliant.

It all worked, and the worst thing that happened was that, each time I plugged it in, I got a warning dialog that I was using an old-school iPod charger that might reduce my cellular reception. I'm sure the dialog box was appearing because I'd cannibalized the data pins, so it couldn't tell it was a nice Apple-brand plug any more.

Then, the iOS5 upgrade hit.

And... what used to be a simple warning message now turns into... the phone won't charge at all in that dock. I think what happened is that the codepath that used to just put up a friendly warning has now changed. Now, instead of putting up a warning, it just refuses to charge if it thinks it's an old-school iPod dock.

So, following a tip on Twitter from a fellow Filker friend, I did the following:

- Cannibalized my second "backup" Apple iPhone Bluetooth Travel Cable, for the pins. (Pins from other less-expensive iPhone plugs were sadly the wrong shape and length to mate with the PCB in that special Bluetooth plug.)
- Using parts from both my Bluetooth cables, assembled a new plug that contained all the necessary bits: The Bluetooth headset charger, the aux-out pins, the grounded pin 11, the data pins, etc.

And boy was I surprised at the results. It charged the phone, it charged the bluetooth headset, and it put audio out on the aux... just not at the same time. As soon as I plugged the USB into some power, suddenly the aux switched back to the iPhone speaker. sigh.

Then I saw this: iPhone Accessory Message... the answer is 68k

And sure enough, all I needed to do was carefully lift pin 21, put a 68k resistor on it, and run that to ground. Now it does all of the things I want: iPhone charge, headset charge, aux-out, all at the same time.

At least it does them in my indoor tests with AC-adapter plugs. Install it into the car with my DC charger, and... and... and... The aux-out only works if I dock the phone before starting the car's ignition. Otherwise it uses the iPhone speaker. So now it's startup-order-dependent. Argh.

Well, at least I have the desired functionality, and a possible work around for when it's not working as desired. For instance, I could power the iPhone charger off the amp remote wire instead of the ignition-switched power. That would at least allow me to work around the startup order dependency if I forget to dock the iPhone before starting the car.

So my questions are:

- Anyone know how many amps an iPhone charger draws? And is that more than the Empeg's amp remote wire can handle? Can I just charge the phone straight off the Empeg's amp remote wire or do I need to do this through a relay?

- Anyone know how to get my iPhone aux-out to work without the startup-order dependency?

I read somewhere that the special iPhone charger can charge at one amp (or maybe 1.5 amp). So in 12V terms about 0.5A-1A

Hm. If I've got one of those generic third-party cigarette-lighter USB charger adapters, do you think that would also draw less than 1 amp? I seem to recall the empeg's amp remote wire was 1 amp.

Last time I looked, DX had 1A and 2A versions of that lighter plug adaptor.

Hm, so it's possible it's right on the border.

Okay, if I've got one of these kind of handheld digital multimeters, what is the best way to *measure* the actual amperage draw of my cig-lighter-plug adapter? Assume that I've got access to the raw wires leading to it and can disconnect them if necessary.

I'm not familiar with that model/type meter, but assuming it can do DC current measurement: disconnect either the positive feed or the ground wire to the adapter and hook up the meter probes to bridge the gap you just created (one to the end of the wire, other to the empty terminal on the adapter).
Which way around shouldn't matter, just gives you a '-' or not on the reading, the absolute value is the same.

Just make sure the meter's range is higher than the current expected.

You can put a resistor in series with the meter to reduce the current to something it can handle.

It's unlikely one of those probe type meters will have a 10A range and maybe something like 200mA at best if at all.

Got a better photo Tony?

Put a resistor in series with the meter? No, that won't work.

As Michael says break either the positive or negative lead (negative would be a bit better in a negative earth car). Put the meter on its highest range (a suitable meter will have a 10A range) and work down from there. Make sure you use the current terminals since they are probably different. Bridge the break with the meter so the current flows through the meter. Also as mentioned, it doesn't really matter with polarity. If wrong the meter will just read "negative" current.

If you only have a volt meter, put a low value resistor in series in the wire where you broke it and measure the voltage drop. e.g. 1 ohm will give a 1 volt drop with 1 amp. i.e. if you measure 0.2V across it, the current will be 0.2A. V=IR.

Depending on the current you may need a 1W resistor or more.

The small voltage drop shouldn't affect the charger.

Is the fear that the meter just won't measure the amperage, or is the fear that I'll let the blue smoke out of my meter if I try?



I tried to google for the model but Radio Shack doesn't make it any more.

Edit: I think it's this one or one very close to it.

I'll see if Fishy has a proper meter, I think he does.

Trying to measure 2A with a meter that can't take more than 200mA would most likely have negative consequences for the meter.


Quoting the page you linked: " Autorange Digital Multimeter for Volt and Ohm. Data hold function"

No mention of Current there, so likely not even able top do current measurement. Looks like you need either another meter or the very low resistance in series trick mentioned in the thread.

Well the fear was both and the latter is somewhat worse than the former but you can still measure current with it.

Measure the voltage across resistor in series with either line. I'd be looking at something like 0.47 or 1ohm. Then calculate from I = V/R (in amps, volts and ohms)

[Shonky knows this already, but perhaps others here don't]

That is exactly what the 10A range on the (cheapo) meters I have here does. The red probe plugs into a special socket that has a sub-ohm resistor in series with the black probe, and it measures the drop across it to calculate the current.

The regular red probe input is fused for less than 1A.. hopefully the fuse might blow rather than the meter frying, but no guarantees on that.

Cheers

Yeah the cheap meters with the 10A socket will be unfused. The lower current range (typically 200mA or 400mA) should be fused. Quality meters like my Fluke 179 though do have fuses for both 400mA and 10A current terminals.

If you have a cheapo meter make sure you get your wiring right if you're using the 10A range. Worst case in a car though should be blowing one of its fuses if you thoroughly stuff things up.

Also you should be able to buy one of these cheap digital meters which is handy for other things very cheaply. Sub $15 should be easy.

iPhones take maximum 1A. If you use a switching DCDC adaptor to get down to 5v (like most of the tiny ones are) count on them being 75%+ efficient, which means you'll be taking at the most:

(5*1)/0.75/12 = 0.55A at 12v.

I believe the empeg might manage that, but I can't remember for sure. You'd be safer using a relay, really.

There was a desktop dock for the original iPhone that fitted the phone and the apple bluetooth headset, AND had audio out. iPhone 4's (and 4S) fit this just fine, and I may even have a spare one...

Yeah, I bought the relay this week and was going to install it during my vacation. Haven't done it yet, but I shall soon.



Yup! I knew about the existence of those, but by the time I got the headset and then wanted one, they didn't sell them at the Apple Store any more.

Of course, I've already done all the necessary work to build the car dock assembly, so even if I could find one of those now I wouldn't really be able to make use of it. Now I'm just curious what I'd done in my connector-modding (see orig post) that's caused this interesting startup-order dependency. Any ideas?

If you can't replicate it at home with AC bricks... this could be an available current related thing. The apple wall plugs (and apple laptops) are rated to 1A current, and when one of those travel cables is detected the amount of current the phone takes is adjusted downwards to allow for the headset to charge too without blowing through the available current.

It's likely that your car charger is either appearing as a china charger (D+/D- shorted) or is identifying itself as a 500mA charger, and this is modifying the behavior incorrectly (it's likely a bug from the sounds of things).

What appears to be the 1A resistor network is posted here:
http://www.instructables.com/id/Build-a-Battery-Powered-USB-Charger/

Is that what actually being compliant to the USB-IF Battery Charging specification is called nowadays?

Peter

The main article in that link describes a circuit and lack of understanding that beggars belief.

For some reason the Internet seems full of incorrect or incomplete information and much misunderstanding when it comes to exactly how the various Apple USB chargers and iOS devices actually sense and manage battery charging.

I have the following information to offer.
When D- is 2V and D+ is 2V, it's a 500mA charging adapter (standard non-Apple USB 2.5 watt charger)

When D- is 2.8V and D+ is 2V, it's a 1000mA charging adapter (recent iPhone/Apple standard 5 watt charger)
When D- is 2V and D+ is 2.8V, it's a 2000mA charging adapter (iPad 10 watt charger from Apple, also compatible with iPhone)

As far as I can tell the resistor networks for the last two are the same as that posted by Robot Lover in the comments to the linked article, with the 1 amp capacity voltage signals on D+ and D- simply being swapped to indicate 2 amp charger capacity.



Recent iOS devices (I cannot comment about others as I did not test them) carefully balance current draw against voltage sag. An iPhone 4, for example, starts by sensing the voltages on D+ and D- and determining if it is connected to a dumb charger or an actual working USB host.

There are some special cases if the iPhone battery is completely drained or the device is turned completely off, but in general the iPhone will begin ramping up current draw while watching the available USB charging voltage at the iPhone 30-pin connector.

If the power source is NOT an actual USB host (it is a dumb charger) then the voltage signals on D+ and D- pins define the maximum current the iPhone (or iPad) will attempt to draw. If the available charging voltage on the 30-pin connector sags too much, the iPhone will trim its current draw to maintain the USB voltage within bounds. Voltage sag can be caused by the USB to 30-pin cable conductors being too thin or too long, and of course by the capabilities of the charger to actually put out the current. Even a hundred millivolts can be a significant voltage swing here.

My measurements indicate that genuine Apple USB chargers are tightly voltage regulated and have output voltages right at the upper bounds of the USB '5 volt' specification, right around 5.250 volts if memory serves. This 'extra' bit of voltage partially compensates for voltage losses along the USB cable and connectors.

Note: The lightning bolt icon in the battery symbol appears whenever the iPhone determines that the connected charging source can charge the iPhone, but does not necessarily mean the battery IS charging right this minute. A fully charged iPhone will actually stop charging the battery and actually allow the battery charge to decline a few percentage points before 'topping it up' again. The lightning bolt icon will appear during the entire time.

The lightning bolt also does not tell you how quickly the battery is being charged. As far as I know there is no way for the iPhone to actually display the charging rate.

Note: I believe, but cannot find the link right now, that the most recent iterations of Apple computers (late 2011 models) can provide the higher 10 watt (2 amp) charging current that the iPad prefers and also support the USB protocol extensions to negotiate the larger current feed to the iPad.

I can't remember the voltage levels used to indicate the charger spec, but that is indeed how it works.

The iOS devices don't "carefully balance current draw against voltage sag"; they will try to take the amount of current specced by the adaptor (though a phone won't take over 1A), but *may* back off if they see undervoltage events. A lot depends on the exact phone as there are different current controls schemes in each generation. Some generations had digitally controlled current limits and others had analog limits that are more predictable.

For "china bricks" (D+/D- shorted together at the charger end) there's an iterative process of stepping up the current limit, letting it settle, then trying for higher - the china spec says anything from 300mA to 1.5A is available so the device has to experiment.



Well, almost. At the input to the charging buck, which tends to be about 200 milliohms away from the 30 pin. Also, they don't charge when off - it'll always start the CPU if sufficient external power is available. If you plug an iPhone in then turn it off, it's only pretending to be off (it's actually gone off, then come back up in a fake-off mode).



It's not that simple, as current changes across charging states and the system load always takes priority over charging current. Turning on the screen, CPU load, camera current, etc can cause the current to spike (though never above the configured maximum based on the charger ID) can cause instantaneous brownouts on the USB voltage, which are ignored in some cases.



That's correct.



Correct, it's not exposed.



Very possibly. I know the unibody macs added the 1A charging support in ... 2009? Whenever those came out.

That is a spectacular set of information. Awesome.

My current issue isn't the charging right now, though: I have no trouble getting the thing to charge these days, the problem is getting it to put the audio out on the dock pins. My startup-order dependency is that it always charges, it just won't put audio out on the dock pins unless I dock it before applying power to the charger. If I do it the other way round, then the audio comes out through the iphone speaker. This was, I think, reproducible with indoor apple chargers too, but of course only on my modified plug (with pin 11 grounded and the 68k resistor on the accessory pin).

For completeness and future reference, further review of my old notes from my iPhone charging profiling (1 amp charger style resistor network on D+ and D-) with either an iPhone 3GS or iPhone 4 (not sure which I had at the time but I think it was the 3GS).

If the iPhone was plugged into power and then turned off the maximum charging current would limit itself to 500ma. If the iPhone was allowed to 'boot up' then the maximum charging current was approx. 870ma at 4.85 volts.

The minimum input voltage to start charging was measured as 4.50 volts.
If the initial input voltage was above 5.50 volts then the charging would not begin.

Once charging had begun with iPhone ON the input voltage could rise to 5.60 volts before charging would shut down.
WIth iPhone OFF the minimum charging voltage was 4.25 volts but would only draw 52ma. Current draw remained minimal until input voltage exceeded 4.45 volts.

Interestingly, as the input voltage was raised from 4.50 to 5.50 the input current would decrease from 496ma to 454ma (iPhone powered down).

Minimum input voltage for maximum current draw of 870ma was 4.85 volts with iPhone turned on. Further voltage increases maintained the wattage consumption at about 4200 milliwatts by moderating input current towards 756ma at 5.50 volts.

With the iPhone battery heavily discharged (large red battery symbol) the powered down iPhone would not immediately boot up when connected to the charger and would limit itself to under 500ma of charging current. Once the battery had partially charged the iPhone would start itself and then the charging current would rise further.

If D+ D- were left open then current draw limited to =< 100ma.
D+ D- resistor voltages applied after charging began, current limited to =< 500ma.
D+ D- resistors are connected right from the start and then disconnected the charging current can remain above 750ma.

I stopped collecting data once I had a good enough idea what the iPhone was doing with the D+ and D- signals from the charger and how charging voltage related to current draw.

Clearly the iPhone (3GS in this case, IIRC) is sensitive to both the semaphore voltages on D+ D- and to the actual charging voltage delivered to the 30-pin connector, and of course the available current.

It would be interesting to see if the iPhone 4S also limits itself to 1 amp maximum current draw even when connected to the high output iPad style 10 watt charger. A faster charging iPhone 4S would be nice to have

Perhaps this may be helpful;
http://pinouts.ru/CellularPhones-A-N/iPhone_connector_pinout.shtml

Thinking about your resistor to ground comments, perhaps there is an issue with which ground(s) you have used.

I think there are separate ground pins for power, USB and audio within the iPhone connector. I do not know, but I suspect, that some of these grounds are NOT inter-connected together within the phone itself. Or maybe they are, but it may still be important which ground pin gets used for what.

There may also be different ground points within your vehicle connections for power vs. audio. Sometimes these can have voltage differentials even though they are all 'ground'.

Perhaps what is happening is that the unpowered power input wiring is affecting the sensing of the 'grounded' 68K resistor or other pins you are depending upon. A review of exactly which pins and signals are grounded, and through what, may lead you to a discovery and resolution.

Also review what voltages and resistances are on the D+ and D- pins, as these signal what charger 'type' is connected.

I also found this project connector that might be useful;

http://www.podgizmo.com/p/118/ipod-iphone-ipad-podbreakout-mini

Yeah, the Pinouts.ru site is in fact what I was using as my original reference for the hack. But modding the connector is tough, and you're right, I may have simply used the wrong grounds. The easiest grounds were to the shielding. So that could very well be the problem!

Yes, that's not real off, it's the bootloader which doesn't have the full stack of stuff for charger ID and treatment hence it's a bit more conservative.



That'd be the overvoltage protection circuit...



Again the OVP, but the ~5.5v threshold looks different when current is flowing (ie, there's a drop between where you're measuring it and where the PMU is measuring it).



This is because it's not a linear charger, it's a DCDC that bucks down to just above the current battery voltage, then LDOs into the battery (to keep the battery rail clean). Hence you get a multiplier effect that will reduce the input current as the voltage rises (it'll be vaguely constant power, as you saw with the 1A brick/charging.



That would indeed be nice, but typically battery vendors don't like you charging at a higher rate that 0.7C (this is ~1A charge current maximum with a 1420mAh cell). Due to various tolerances in the system, the typical battery maximum current is around ~0.63C, though individuals units could be +/-10%.

There's one ground for everything, in effect BUT the audio has a ground *sense* pin. This senses the audio ground at the point of consumption (on a dock, that's usually at the audio output jack) and the system drives the signals with respect to that virtual ground. This means that changes in charge current, for example, which would be moving the system ground around (with respect to what the dock thought was ground) don't get coupled onto the audio.

The resistor can go to any actual ground - ie the power supply ground is fine.

Thanks for clarifying and expanding.

One of the reasons for my own original investigation was to understand how one might maximize iPhone charge speed when connected to a computer via a powered USB hub.

Even when the USB hub is connected to a modern Mac which itself could support and negotiate the fast 1amp charging of a directly connected iPhone, when connected through the powered hub the iPhone seems to limit itself to the more pedestrian 500ma charge rate.

Many USB hubs have power supplies that can deliver more than 500ma in aggregate, but I didn't find anything in the Mac or iPhone characteristics that would allow the iPhone to negotiate and draw the additional power from a USB hub.

As far as I know Apple does not sell a simple USB hub of any kind that might support the 1amp or 2amp charge rates. I don't know whether the USB ports on something like the Thunderbolt display can do it, but that would be a $1000 hub!

Must an iPhone or iPad be directly connected to a Mac in order to utilize the higher charging rates? Or is there a way to do so through a (very up to date) USB hub?

Problem illustration: I have (among others) a 2011 MacBook Air which has exactly two USB ports. I also have an iPhone and iPad. If I connect each directly to the computer then I have zero USB ports for anything else.

If I attach a USB hub then any iDevices connected through the hub charge much more slowly. The iPad actually slowly discharges if the screen is on while hub connected since the minimum current to support in-use iPad charging is 1amp.

Which pin number is the audio ground sense pin?

What happens if that pin is not actually connected to anything? Or to the wrong thing?

It should be possible to make a hub to provide more power, though I didn't have much luck finding one for sale when I searched two years ago. The previous 27 inch Cinema Display has a hub in it, and it will provide up to 1A per port. This was my solution during the pre iOS 5 days for charging, both the iPhone and iPad sat on two of the monitor ports. (Now they sit in a dual charging dock in my bedroom and sync via WiFi). The Thunderbolt display has a full USB2 controller and root hub in it, and also provides 1A per port.

One possibility might be a USB 3 hub. The base port is supposed to offer up to 900ma of power, though I'm not sure if the port would offer this to a USB 2 device.

That's because >500mA is not in the USB standard. Apple uses vendor-specific extensions to request >500mA, hence why it only works with apple devices.

There *are* specs for higher power charging from hubs (up to 1.5A) but this requires extra circuitry in the device to do the dance with the data lines that allows you to identify this.

As, for signal integrity reasons, this generally needs to be within the USB PHY, there's a lag of several years between the standard being made and chips that support the standard getting into phones.[/quote]



I'm not aware of any hubs that would support that I'm afraid. The "solution" is an externally powered thunderbolt display

For completeness and future reference, I should include this link to the Battery Charging v1.2 Spec and Adopters Agreement, USB Implementers Forum

From the USB 2.0 Wikipedia entry for USB 3.0;

The 1-amp rated Samsung AC charger supplied with my Galaxy Nexus has pins D- and D+ shorted.

The cheap china car charger I got from DX has proper resistor networks to signal 500mA max per port (2 USB ports). I just now hacked it, adding 200ohms across D- D+, to trick my phone into charging at (up to) 1amp from it, instead of trickle charging as it normally would.

A friend of mine makes a cable that can switch in and out the short between the data lines. This is nice so I don't have to modify any chargers. Dealextreme makes cheap versions of his cable, but he says those are junk.

I took a USB extension cable and modified it, then just swapped that in and out as needed. Same idea.