Disclosure: I work at Amazon, but nowhere near Lab126 where these bad boys get engineered. I’m acting on my own accord using technology that anyone can have themselves. I have no internal information about these and am doing it out of my own curiosity. (Yes, that’s a leadership principal, so there!)
A few bits of 28 AWG wire, some tape, and I made a very crude breakout system for the Kindle. I had read elsewhere that the general pinout is like this:
The grounds are isolated. I’m guessing that pin 1 is the reference for the I2C bus and pin 5 is the battery ground.
This led to me hooking up the system to the scope:
So, to note here I’m probing yellow as pin 2, blue as pin 3, and battery as pink (though I turned it off on the display since it was just noisy). All of these are referenced to pin 1.
Sure enough! Pin 3 is clock while 2 is data!
There’s a crap-ton of pickup on the signal. With the lights turned on it was even worse. It was good enough to get a decode though.
I’ve not had a chance to dive into this, but here’s a raw dump of the data stream:
|Bus Definition: I2C|
|Time||Repeat Start||Direction||Address||Data||Missing Ack|
|4.25E-03||X||Read||1B||25 55 04||X|
Lots of chatter on address 0x55. I’ve not had a chance to guess what might be there. But there’s certainly a lot of activity going on there.
When I get back from my vacation I’ll set up a trace to see if anything other than 0x55 gets any action.