HDMI adapter opened
I opened my Nook HD+, and there is a test point near the dock connector labeled usb-id. If this TP is shorted to GND with a slave USB device attached thru a USB-A to A gender bender, the Nook acts as host as expected per USB-OTG standard. This TP maps to pin 24. Unfortunately, this isn't one of the pins to a Nook HD+ USB cable.
The Nook HD+'s motherboard is labeled EXCELLENTLY. I've gotten most of the pins I can get easily. I think the remaining pins are for HDMI (which makes sense seeing B&N makes an HDMI adapter cable...). The HDMI adapter cable will probably get the remaining pins.
Here's the pinout I've got so far:
Pin - Signal
--------------------------------------
1 - USB VBUS (+5VDC)
2 - USB VBUS (+5VDC)
3 - HDMI 5V
4 - LINEOUT-DTC
5 - USB D-
6 - USB D+
7 -
8 -
9 -
10- GND
11-
12-
13- GND
14-
15-
16- GND
17-
18-
19- GND
20-
21-
22-
23-
24- USB-ID
25- LINE-L (Audio)
26- GND
27- LINE-R (Audio)
28- GND
29- GND
30- GND
Using a chisel with a fine edge inserted along the left rear of the plastic case, I pried open a Nook HDTV adapter, with only very minor marring of the plastic package (it was held together with weak glue). I discovered a couple things:
- It isn't just wires; there's quite a bit of logic in there. There's a "P13HDMI 101-BZHE Y1246GG" 42-pin chip, along with at least one voltage regulator, several transistors, and at least a dozen each of associated diodes, capacitors, and resistors.
- Pins 26 and 28 on the Nook edge are not grounded in the package, and are presumably shields for the accompanying audio LINE-L/R pins.
- Pin 24 (USB-ID) on the Nook edge is not grounded or floating, but goes through a (measured 0.0 ohm???) resistor (R17-??, just below transistor Q5), to transistor Q5 and on to other circuitry. For pin 24 to be USB-ID (as commonly thought to be "active-low" in the cable for USB host mode) makes no sense (see related thread http://forum.xda-developers.com/showthread.php?p=56896393), unless pin 24 is really the NEGATION of USB-ID. Using USBHostSwitcher.apk to place the Nook in USB host mode (with the HDTV adapter inserted) causes pin 24 to go from TTL-low (.92v) to TTL-high (3.3v), and also causes pin 1 (USB VBUS) to be powered to 5v. Edit: However, USB host mode still doesn't work through the HDTV adapter (see edit comments below).
Here is an image of the top of the adapter, with the Nook-edge connector down, and the pins on it numbered left to right 29, 27, ..., 3, 1. The cable-edge connector is up, and the pins on it are numbered left to right 30, 29, ..., 3, 2, 1 (high-res image
here):
Here is an image of the bottom of the adapter, with the Nook-edge connector down, and the pins on it numbered left to right 2, 4, ..., 28, 30 (high-res image
here):
Edit #1: After further examination of the images (see link above) and measuring voltages, it's clear that when in USB host mode, the data pins (5 & 6) and ground (29 & 30) go straight through from the Nook edge of the adapter to the cable edge. The Nook also provides USB power on pins 1 & 2 (USB VBUS). However, that power never gets to the corresponding pins 1 & 2 on the cable connector edge of the adapter, because transistor Q4 stands in the way. Transistors Q6, Q7, and Q9 are used to switch transistor Q4 on or off, and I think pin 23 (Nook edge) plays a role in that.
Edit #2: What that last sentences means, I think, is that USB host mode with the HDTV/HDMI adapter may work without modification, if the Nook USB cable is connected to a
powered hub. Verygreen's CM11
and USBHostSwitcher.apk will probably be needed.
Edit #3: Verygreen's CM11 (at least the 8/17 SDcard-bootable variant) includes (as a notification tile) the USBHostSwitcher.apk functionality.