View source for Debug Board v2

From Openmoko

Jump to: navigation, search

You do not have permission to edit this page, for the following reasons:

  • The action you have requested is limited to users in the group: Administrators.
  • You must confirm your email address before editing pages. Please set and validate your email address through your user preferences.

You can view and copy the source of this page:

Templates used on this page:

Return to Debug Board v2.

Personal tools

The Debug Board page provides information on the architecture of the second version of the Neo1973 debug board.

Architecture Diagram
Annotated PCB Photograph


The debug board provides these key components.


The Schematics of this board can be found at


This bus-powered hub

  • attaches to the laptop using its upstream port
  • attaches to the FT2232D for JTAG and serial console on downstream port 1
  • attaches to a free USB-A connector using downstream port 2
  • attaches to the phone using downstream port 3

Using the hub, you can have access to the phone, JTAG and serial simultaneously, through a single USB cable.

You can actually even charge the phone (100mA slow charge) using that downstream port.

For the USB hub we use the TUSB2046B chip.

JTAG using FT2232

Basically, we integrate a USB-JTAG adaptor similar to the [Amontec JTAGkey-tiny]. The actual reference design that we used (Thanks to Joern!) can be found at

This provides us full JTAG debugging, at about 150times the speed of the wiggler

The actual JTAG port is connected with

  • the phone (via debug flex cable)
  • A separate 20-pin header using standard ARM-JTAG pinout
    • this allows the user to use this device as JTAG adaptor even for other arm-based devices

Serial Port

Instead of replicating a true RS232 port, we wanted to use a USB serial converter chip, such as the FT232 or the PL2303.

As it turns out, we can even use the second port of the FT2232 simultaneously with the JTAG. So no extra FT232 or PL2303 is required.

Tri-State Serial Port

The serial port needs a tri-state driver, since it is both connected to the GSM modem AND to the debug board. We drive the tri-state driver by an inverted GSM_EN signal (pin 7 on FPC connector).

Usage Instructions



Your Debug Board contains a small serial EEPROM which should be flashed correctly during production. However, some boards have received incorrect/incomplete programming during production and thus you should verify this.

If the board shows up as USB ID 1457:5118, then everything is correct.

Only if your board shows up as USB ID 0403:6010, then you will need to flash the board!

Flashing with Linux
NOTE: FIC has done that for you, in case you have officially been supplied with the board. However, if you are one of the early adopters, the configuration and USB vendorID / productID might have not yet been set correctly.

In order to do so, you can use the ftdi_eeprom program from Unfortunately, the latest version (0.2) doesn't yet contain support for our FT2232D, so you need the patch from

You may also need to apply the following patch to libftdi: (See for additional details.)

Furthermore, you will need the EEPROM config:

Once you have compiled ftdi_eeprom, you can run

ftdi_eeprom --flash-eeprom neo1973_debug_board_v2.ftdi
WARNING: Make sure you don't have any other FTDI FT232 / FT2232 based devices attached to the USB while running ftdi_eeprom. It might be wise to disconnect everything but the debug board

You should get something like the following printout if everything was successful:

FTDI eeprom generator v0.2
(c) Intra2net AG <>
FTDI init: 0
Unable to find FTDI devices under given vendor/product id: 0x1457/0x5118
Retrying with default FTDI id.
Used eeprom space: 102 bytes
FTDI write eeprom: 0
Writing to file: neo1973_debug_board_v2.eeprom
FTDI close: 0

Once the flashing has finished, just disconnect and reconnect, and you should see

$ lsusb -v -d 0x1457:                                                                                                                                                                                               
Bus 005 Device 009: ID 1457:5118
Device Descriptor:
  bLength                18
  bDescriptorType         1
  bcdUSB               2.00
  bDeviceClass            0 (Defined at Interface level)
  bDeviceSubClass         0 
  bDeviceProtocol         0 
  bMaxPacketSize0         8
  idVendor           0x1457 
  idProduct          0x5118 
  bcdDevice            5.00
  iManufacturer           1 OpenMoko
  iProduct                2 Debug Board for Neo1973
  iSerial                 0 
  bNumConfigurations      1
Flashing with Windows
NOTE: FIC has done that for you, in case you have officially been supplied with the board. However, if you are one of the early adopters, the configuration and USB vendorID / productID might have not yet been set correctly.

In order to do so, you can use the FTDI Mprog program from

You will need



Please make sure you have libftdi-0.8 or later. Earlier versions are known to cause problems

Furthermore, you will need to do the following:

ftdi_sio module option
rmmod ftdi_sio
modprobe ftdi_sio vendor=0x1457 product=0x5118

or the equivalent in your modules.conf

udev rule

Please install the rules from


Please use the driver from

Hardware connection


  • J1 - is a regular 20 pin ARM JTAG at 3.3V level with 5V tolerance
  • J10 - is a 3.3V serial connected to portB of the FTDI2232D
pin name
1   TXD
2   RXD
3   RTS
4   CTS
5   DSR
6   DTR
7   DCD
8   RI
9   GND
10  VCC (3.3V)

Actually using it

On Linux

Serial Port

The Linux kernel of your host system will create a virtual serial device called /dev/ttyUSBx where 'x' is a sequentially assigned number. If you don't have any other USB serial converters attached to your machine, the device name will be /dev/ttyUSB0.

You can use your favourite terminal emualtor (minicom, cu, zc, ...) just like for any other/real serial port. You may use gdb as well (eg. target remote /dev/ttya).

bdb - My experience with the debug board shipped with GTA01B (v3??) was that the serial port eventually came out as /dev/ttyUSB1. I assume ttyUSB0 was the jtag. I also couldn't get the serial to work at all until after I had fired up openocd to take over the jtag. (Which disappears ttyUSB0) That could have been coincidence of course... I had also removed all flow control on the serial port settings - maybe that was the key. PS - 115200 baud.


Once you have installed libftdi >= 0.8 and configured OpenOCD with the correct openocd.cfg from OpenOCD#openocd.cfg, it should work just fine.

Please see OpenOCD#OpenOCD_and_Debug_Board for some more information.


We previously had Neo1973 Debug Board v1 (inherited from some weird engineers who must have seen Brazil too often). Version 1 was never shipped to phase-0 or phase-1, and not sold to anyone.

Changes from v1 to v2

  • get rid of ethernet
    • we don't need it, and
    • we especially don't want a 40pin parallel 66MHz bus going between two pcb's
  • get rid of 7-segment LED displays
    • not really needed. We have a serial port
    • could be replaced by one or two GPIO LEDs
  • get rid of built-in wiggler
    • nobody has a parallel port on the laptop these days
  • get rid of li-ion battery (including charger)
    • the device can be fully usb powered by the laptop