WL154 - I read the results of your current draw test and believe that I understand most of the implications, one of which is that you might benefit from a short primer on how LIon fast charging works, so that we can continue to make some headway on this problem my friend:
If an LIon battery is below 70 - 80 percent capacity (depending on exact chemistry), it is always charged starting in Constant Current (CC) mode, until the voltage gets to about 4.0 to 4.2v (again these numbers vary a bit depending on chemistry and whether you are optimizing for battery capacity or longevity). All LIon fast chargers do this, whether they are 10W or 80KW
So if you saw your (calculated) current to the iPad start at 1.2 amps and increase to 1.5 amps at 80 percent, that is a clear sign that the DL is not actually directly controlling the charging . Most likely the DL just supplies a (supposedly) steady 5V and there is another circuit which is actually in the iPad, most likely also using another buck switcher to supply the battery with a constant current until the voltage comes up to ~4.1V. As the battery voltage rises from say 3.3V (fairly well discharged) to 4.1V (mostly charged), the charge current to the battery stays the same, (probably around 1.7A, going by your numbers), but the current draw from the 5V voltage source on the Mackie increases. The actual charging buck switcher in the iPad has to pull more increasingly more current from the 5 volt rail provided by Mackie, in order to deliver a constant current source to the battery.
So - if you saw the current increase to the iPad as the battery state of charge increased, that is a sure sign that the actual battery charging regulator is on the iPad, not on the DL. No LIon charger increases the current during the CC phase. The DL must just supplying a dumb 5V rail to the actual charging circuit - which may be a good thing for our purposes, because a dumb 5v rail is usually simpler and easier to clean up than a variable voltage regulating circuit.
Now, once the battery voltage comes up to about ~4.1V, the charging circuit will switch from fast CC mode to constant voltage (CV) mode. During CV mode, the charging circuit will hold the battery voltage to ~4.1v and the current will decrease from ~1.7A down to almost nothing. The CV charge stage is terminated when the current reaches a small value, usually something like 0.02C, or in this case about 40mA. Beyond this point, the charger will stop, monitor the voltage, and if it drops below a certain threshold (usually 0.05 below the CV cutoff), a small "trickle" charge is periodically supplied to keep it up to full charge.
That is how LIon battery charging works in a nutshell. Keep in mind that you are never going to see a current as low as 40ma coming from the DL, because in addition to keeping the battery topped off, the 5v rail is also what is powering the iPad. So final current will be much higher than 40ma.
What is interesting in your case is that you reported a sudden drop in charging current, from 1.5 amps to 0.5 amps to the iPad. That is not how the CV stage of LIon charging works. Typically, you will see the current reduced in the shape of an exponential curve, from CC current (~1.7A to the battery, ~1.5A from Mackie's 5V supply) all the way down to almost nothing, over a period of about 90 minutes. Not a step decrease like this.
When you said that the unit "failed" at 80 percent when the current dropped to 0.5A, are you saying that Master Fader lost synch at this point? If so, then it seems to me that the failure could be one of three things:
a) If the 5V regulator circuit on the DL is not adequately sized, as the charging circuit on the iPad pulls more current, the rail could be getting hogged down, which will make it increasingly noisy. As voltage on the battery comes up, the current required from the 5V rail in order to keep the battery in constant current mode goes up, and the noise increases until it starts affecting data signals and something hiccups. This is actually the best case scenario because it ought to be relatively easy to fix a simple 5v regulator circuit
or (more problematic)
b) the fact that the current dropped right to 0.5A when the failure occurred is suspicious to me. Because that is the max current supplied under the normal USB spec. I havent delved too much into the apple charging particulars, but my limited understanding is that some USB signaling is involved which allows the iPad or iPhone to tell the wall wart "give me all the juice you can" instead of the normal 500mA. The fact that the current dropped right to 500mA when things went south is a troubling sign to me that Mackie might have used the same chip that Apple uses in their wall warts to supply the 5v rail to the iPad. And when the MF synch dropped, the USB signal from the charging circuit also dropped and the supply circuit on the DL went into dumb "I'm just a standard USB charger" mode. There really should be no need for Mackie to have used the apple chip, because there is no chance that anything other than an Apple product will be using the DL's USB circuit, hence no need to support the published USB spec, just give the device as much current as its onboard charger wants. However Apple is pretty good about bamboozling manufacturers into buying chips that they dont really need. In all likelihood, the apple chip can be removed and replaced with a simple 5v 2A regulated supply without too much fuss, but it wont be a pretty mod, and definitely not a drop in parts replacement or an easy factory BOM change like (a) is likely to be. This might also explain why Mackie hasnt fixed this yet - because its not really their problem. Under normal wall wart conditions, apple probably doesnt care if their regulated 5v is noisy as hell, but put it in a cable with a lot of other control signals and bad things can happen. Still this is fixable in the field.
or (worst case)
(c) Both MF and the iPad's charging circuit are both using the same USB data lines for control, with at least two devices on the DL (Apple's 5V supply chip, DL's control interface) and a single host controller, with multiple application interfaces (charge circuit control FW and MF) on the iPad. When the iPad battery needs more than 500mA of current, both the charge circuit and MF are using the same USB lines. In theory there is nothing wrong with this. The USB protocol supports multiple device and host endpoints on the same bus, but in practice, this can get tricky. I've worked on host mode USB drivers before and they can be a real bitch to get to be 100 percent reliable even when interfacing to known devices. In the case of Apple, there probably are not too many instances where an iPad will be both charging and doing some other critical control data over same USB lines at the same time, so Apple may not have spent a lot of time beating on their USB host controller firmware. In that case, there may have been a data collision in the USB host driver, which would have occurred on the iPad and it could take out both MF and the charging circuit. That would explain why MF drops out and the current drops to 0.5A at the same time (assuming that is what you saw). That would also explain why Mackie is having such a hard time fixing this. Apple is probably telling them that they are not using the USB host driver correctly, but the truth may be that Apples USB host driver has holes in it.
Fortunately for everyone having these problem, in the short term, no matter what the cause (assuming one of the above), the workaround is the same. Keep your iPad fully charged, so that it doesnt need to draw more than 500mA from the DL to charge its battery. Or, if your iPad's battery is old and doesnt hold a charge well, get a cable and cut pin 23
