Peter Bradstock, who started all those rumors about the "authentication" chip and then "dynamic pin assignment", which have since made it all the way around the interwebs and even CNN has "reported" on, also sells custom headphone cable kits for $17,000.00. I'm not sure I would trust any of what he has to say about the engineering behind Apple's Lightning interface.
I kind of agree with you that this chip being for pin reassignment doesn't add up. First of all, from just looking at the ribbon cables in the iFixit teardown, you can see that there are two sets of traces going to the female Lightning connector in the iPhone 5 that look like differential signaling pairs. I just double checked my Lightning cable against Mr. Bradstock's pinout, and came up with similar results to his, although his pin numbering seems a bit wonky. If we number the pins on the female jack from left to right, like this:
Shield 1 2 3 4 5 6 7 8 Shield
And call "top" side of the male Lightning connector "Side A" (the one I believe the chips to be on, although I didn't actually cut mine open to find out.) Then we arrive at a pinout that looks like this:
Code:
Lightning Lightning
Plug Side A Plug Side B USB Type A
Shield --------- Shield ---- Shield (tied to GND)
1 ------------- 1 ------------ 4 GND
2 ------------- 7 ------------ 3 D+
3 ------------- 6 ------------ 2 D-
4 ------------- 8
5 ------------- 5 (1 VBUS)?
6 ------------- 3
7 ------------- 2
8 ------------- 4
As you can see, there is no direct path for the USB VBUS, but there sure is for D+, D- and GND, so the cable does not reassign its own pins. The SoC in the iPhone 5 would appear to switch the signaling pair being used depending on the cable orientation. It would also appear to be outputting a completely normal USB 2.0 signal; no muxing/demuxing appears to be done by the chips in the cable.
I would therefore presume that the chips have more to do with voltage regulation and/or digital negotiation of bus power. I suppose it is possible that the larger chip also communicates with the SoC via an AUX channel or C-wire to request a USB 2.0 signal rather than a video signal, or acts as a sort of trusted platform module to lock out unauthorized aftermarket accessories, but I have trouble believing that the latter is the primary purpose.
