Here are some measurements of the charge cycle for an 18 mo old Gen 3 iPad, with the DL1608, and Apple 10W and 12W chargers
To measure current from the DL, I popped the plastic housing off my old Patazon 3ft 30 pin cable, cut back about an inch of the jacket in order to separate out the red usb power wire (pin 23) and measured it with a current clamp meter (accurate to 1% and calibrated a few months ago). Voltage readings were from Pin 23 to 15.
http://imgur.com/DQ6fL70Under no load, the DL puts out 5.01v
With the iPad battery completely discharged, so that the iPad is unable to boot, the DL puts out 0.48A (dumb USB mode) and voltage drops a bit to 4.94v. Interestingly both the 10W and 12W apple chargers put out about 1A in this mode, and have a lot more of a voltage drop (~4.2v)
When the battery gets to 2 percent charge (so that the iPad is just barely able to boot), the current from the DL goes to 2.01A and the voltage drops to 4.69v. I put a scope on pin 23, expecting to see a sawtooth waveform on top of the rail, with a spike on the rising edge, but it actually looked pretty clean, with a solid 4.7v rail (I'm still not sure I believe this though, and might try to borrow a better scope for further testing).
At this point, the apple 10W charger was putting out 1.03A @4.21v and the 12W was putting out 1.05A at 4.24v. This is the most current that they put out during the whole charge cycle, and is only about half of what they should be able to produce. In order to measure these, I had to plug them through the 3 ft Patazon cable, so the total charge cable length was about 6 feet. That could have caused them to cut their current back, although the iPad did display "charging". If I can find an extra iPad charging cable, I might open one up like I did the Patazon and see how these two chargers work when there is only 3 feet of cable
The DL continued to put out about 2 amps until the battery was somewhere between 50 and 65 percent (I wasnt watching it closely enough to see exactly where it went into CV mode). By 65 percent, the current had cut back to 1.85A and it slowly dropped until it got to 1.50A at 95 percent - which is where things started to get a little strange.
At 95 percent, I was still seeing dropouts with MF, which was unusual. Then, while playing with MF, the SOC went straight from 95 to 100 percent in a blink. Next, when unplugging and re-plugging the cable to get my synch back, the current from the DL dropped from 1.50A to 0.89A, at which point I had a solid connection. Unplugging and re-plugging the 30 pin cable, sometimes I would get ~1.5A, in which case MF could not keep the synch and sometimes I would get ~0.90A and have a solid connection.
What is probably going on here is that when the iPad first displays 100%, it is not really at 100%. This has been documented elsewhere. In reality, the battery is still halfway down the CV curve. It turns out that 0.89A is the current draw for my iPad running MF when the battery is completely charged. So at this point, the iPad's charger is having trouble deciding what to do. Plug the cable in once and the iPad thinks it still should be charging and pulls 1.5A. Plug it in again and it thinks the charging is done and just pulls what it needs to run the iPad. All the time, the battery display is 100%.
By this point it was getting late, so I disconnected the iPad from the charging cable and left it sitting overnight. The next morning it was at 97 percent and when I ran MF, it almost immediately dropped to 93 percent. That was not too surprising since it didnt really charge to a full 100 percent the night before (even though the display said it did). What was very strange though is that now when I plugged the iPad into the DL, it was only drawing 16mA, which is much less than the current needed to power the unit. The 10W and 12W apple chargers were only supplying 7mA and 8mA respectively, so this appeared to be an iPad issue, not a DL issue. During this time MF would not synch at all, even though I would normally have no problems at 93 percent. Then, to make matters more confusing, the battery indicator ticked up to 94 percent. It was impossible that the battery could actually be charging because the iPad needs ~0.8A just to stay on and the DL was only putting out 16mA. I then played some music through the iPad speakers, just to use more power and let it go for about an hour. During this time, the battery level crept up to 100 percent again, even though the most the iPad ever pulled was 17mA. Over the course of an hour with the iPad's display on and music playing, the actual battery charge would have dropped by about 3-4Whr - even though the indicator said it was charging! So clearly the iPad charging logic has some issues. At the end of the hour, I checked the iPad draw from the Apple chargers again, and they were still only putting out 7-8mA. I powered the iPad down and back up again and same thing. Unit displays 100 percent, and only draws a few mA from the DL or either Apple charger. This looks to me like some sort of bug in the iPad's charging circuit, which happens right around the time that the battery display goes to the false 100 percent charge (but battery is still in the middle of the CV curve). In retrospect, I should have tried docking the iPad to see if it had any effect. I might try this again to see if I can replicate it. I suspect that when the iPad gets into this mode - right at the onset of the "false 100 percent" charge, the iPad can lock up in a state where it wont synch to the DL and wont take a charge. This appears to be the iPad's issue, not the DLs. Running into this at a gig would really, really suck.
Next, I powered down the iPad and left it for an hour. When I powered it back on, the battery display was still at 100 percent, but the DL was putting out 1.5A again, so at least the iPad realized that it was supposed to be in CV mode and was drawing a reasonable current. I let the iPad go to sleep mode, at which point the charge current went to 0.87A and tapered down to 0.22A over 90 minutes. After 3hrs, it reached what seems to be a steady 0.11A.
At this point, the DL and iPad went back to being their old normal selves again. Power draw from the DL was about 0.8A to 1A while MF was running and MF had no problems keeping synch with other programs running. However playing around with the setup some more, I realized that "normal" is not quite what I thought it was.
It turns out that keeping MF from dropping out is not really so much of having the battery fully charged, but keeping the current draw from the DL below about 1.2A (for which a full battery is needed). If I increased the brightness of the iPad display to about 3/4, I would get periodic drop outs, and at full bright (about 1.5A), it wont hold synch at all. No problems at default and lower brightness settings. Even my 6 ft Cable Jive cable (normally rock solid at full charge) drops out after about 15 min at full brightness on a fully charged battery. The 2ft Cable Jive has no problems at full bright (or any other condition that I've run into so far).
That's about all I have time for this week and still a lot of open questions, but I figured I'd throw these results up to see if anyone else has some insight as to what might be going on.
TLDR: Looks to me like the problem(s) might be with power distribution on the iPad, but further testing is needed. In the meantime, if you are having disconnect problems, make sure that the iPad battery is fully charged (not just displaying 100 percent - give it a good 2 hours of charge time after it gets to 100 percent). If you are still having problems and dont want to hack a 30 pin cable, try reducing your display brightness and see if that helps
