Ooit, als Apple-hater, een USB lader gemaakt voor nichtje met een iPad. Onderstaand onderzoek toen uitgevoerd.
(1) het was een poos geleden, dus geen idee of o.a. Apple iets gewijzigd heeft sinds toen. Quick Charge met hogere spanningen was toen nog helemaal niet aan de orde.
(2) het is in het Engels, hopelijk geen probleem. Je kunt skippen naar "Conclusions" voor de weerstandswaardes (of weerstandsdelers eigenlijk). Dit gaat uit van een Apple iPad 2 en vergelijkt de toegepaste oplossingen soms met andere merken.
Some more testing
After reading everything I could find and getting conflicting information, I mailed TimothyB about someone's suggestion the drawn current changes with supplyvoltage. He couldn't confirm this, so as a last resort, I decided to go and test for myself. Quite a feat for someone without any Apple products...
I'll first post the claim found on the net (possibly this thread) and then my measured result. Note that I used a switching powersupply which I used to read the current of off. The voltage was measured IN THE USB-plug, about 15cm/6" before the actual plug going into the iPad3. That rules out any voltage drop upto this point. I used 20k multiturn potmeters to adjust the voltages on D- and D+ (slider on D-, 20k between 5V and gnd, same with a second potmeter and D+). Please also note these voltages were set before loading the 5V voltage and were not adjusted during the test.
Claim: anything below 300k between D- and D+ will work.
Used nothing between D- and D+ (infinite resistance): iPad charged with 90mA.
Used 100k, the iPad3 charged with 1.36A
Used 10k, 5k, 2k and 0ohm: each time the iPad3 charged with 1.36A
Supply voltage did not make any difference (4.5-5.2V).
Claim: both D- and D+ at 2.66V is a good solution.
At 4.5V, the iPad charged with 1.86A
At 4.6V, the iPad charged with 2.09A
Every next increment of 100mV resulted each time in a charging current of 2.16 to 2.19A (fluctuated slightly, might e.g. be measurement inaccuracy of currentmeter, don't read anything into this).
Claim: 2.75V on D- and 2.00V on D+ results in 900mA charging current.
I used 2.725 and 2.000V and found 0.90A across 4.7-5.2V supply voltage.
Claim: 2.75 on D+ and 2.00V on D- results in 2A charging current.
I used 2.725 on D+ and 2.023 on D- and found 1.90A chargingcurrent from 4.8V to 5.2V.
- Using a dead short between the datapins results in a worthwhile chargingcurrent of 1.36A. This may not be enough to keep it topped up during use, BUT it does make the charger compatible with HTC, Samsung and the rest of the world. One chargingtrick for all!
- Using 2.66V on both datapins results in the highest chargingcurrent. More than the (iirc) claimed 2.1A @ 5.1V of the 12W charger.
- Using 2.75 and 2.00V on the datapins results in 0.9 or 1.9A chargingcurrent, depending on which way you switch it. This opens up the possibility to make a simple switch to change between the two charge modes (e.g. on a solar-charger, I would need to draw this out, but I think a DPDT switch should do the trick).
- the claim the drawn current depends on the supplyvoltage was limited: they (link I cannot post: voltaicsystems) claim 1A@4.5V and 2A@5V. I saw 1.86A@4.5V and 2.2A@5V (different charging mode, but just to show my variation was not as extreme as theirs).
Hope this helps out!
Posted here as this discussion is one of the best out there.