Should I drop the NiMH aspects here for the two of you to continue along the lines of Li-ion development/howto?
Not at all. The Li-Ion dialog was itself a bit of a hijack I merely posted upon request.
However we do know from the two wire EEPROM implementation that I2C runs though it! So I'm wondering just what that PIC chiplet might be doing over that nifty little communications setup? Why EEPROM? Was serial EEPROM the only readily available (overkill) component suitable to the identification task at hand or could the Power Manager be keeping notes on accumulating charge cycles of each individual battery pack?
Very interesting certainly. Would love to scope it and do a little bit banging. If it can manipulate the power manager's behavior, that could open all sorts of possibilities, including possibly overriding the 11v cutoff that necessitates an internal boost converter in my Li-Ion packs.
You don't perhaps happen to have any info on what differentiates the 3 types of batteries do you? I haven't been able to find much of anything in the various Duo-related Developer's Notes or Service Source docs, but the way they distributed info throughout those publications is infuriatingly inconsistent. You really have to consume all of them to get a complete picture. There's nuggets of info sprinkled throughout the "Duo System" and platform specific 270c and 280c dev notes, but I haven't found much of
anything about the batteries themselves, other than that using type-IIIs on older OSs requires a system extension for them to charge properly. Hmm...
WAGs would be that BTI may have managed to: break into the power manager's ROM to source battery IDs reserved for incremental future development, figured out how to patch the Power Manager ROM for such or to jigger settings on the fly for pulse charging to higher than provided capacity? Dunno at all really and haven't much of a clue.
I frankly wonder if, instead of some sort of I2C wizardry and power manager manipulation, BTI went exactly the opposite direction. What if these hypothetical charging algorithms, likely introduced to combat memory effects and with the unintended consequence of thwarting attempts to use higher mAh cells, were only introduced in type II or type III? BTI may just be using earlier type S/Ns that only perform voltage-referenced charging? As for the overkill PIC, who knows? Maybe that's the best way they could figure to reverse engineer and re-create the Apple-proprietary ID chip?
What I do know for certain, is that my 3500mAh Li-Ion pack based on a cannibalized type-II pack charges all the way up to full capacity using an M7783 charger just fine. It also still doesn't make sense to me that the power manager could somehow keep track of the pack's overall capacity, and only impart that much energy. What if I swapped the pack into a different Duo and discharged it halfway? It's
gotta be voltage referenced.
I wonder if the cell chemistries differ? What if a type-III used an at-the-time new, higher capacity, but slightly lower voltage cell or something? Using modern, conventional NiMH, cells (higher capacity or otherwise) would then result in them being under-charged? There's also the possibility that folks are buying falsely advertised cells and not getting the fictional output they expect. I'd estimate that
half the cells I've ever purchased on eBay, AliExpress, etc., were mis-labelled with some hilariously overrated.
Anyway, here's a bit of interesting but still agonizingly unspecific info from various dev notes:
Duo System:
- The charging current is programmable, with an external voltage supplied by the Power Manager, which also controls the charging functions, by monitoring battery voltage and battery temperature.
270c:
- The latest release of the Power Manager code, which resides in the PowerBook Duo 270c ROM, allows you to modify the system time without impacting the battery-charging algorithm. Firmware features added for the PowerBook Duo 270c computer also support a sleep LED, version dependent code, 4/5A Panasonic battery support, and battery conditioning.