New Life in Openmoko Phones

Rask Ingemann Lambertsen rask at
Wed May 20 02:11:44 CEST 2009

On Tue, May 19, 2009 at 10:52:22AM -0400, Stefan Monnier wrote:

> That still doesn't explain why removing one of the two accelerometers is
> a good idea.  What is the benefit?  Why not remove them both?
> Is it that all the programs that use the accelerometers (as of now) only
> use one of the two?  Is it that having two accelerometers introduces
> layout difficulties?  Is it that there aren't enough interrupt lines on
> the SoC to properly support the two accelerometers? ...

   I don't think so. We already have both EINT8/GPG0 and EINT16/GPG8
reserved for the second accelerometer, but not connected it to EINT16/GPG8
(R1547 = NC). Last time I counted[1], there wasn't a shortage of interrupt
or GPIO pins.


   If we're really going to mess with the accelerometers, why not move them
off the SPI1 bus and onto GPIO pins? We're currently using the bitbanging
GPIO_SPI driver anyway. That way, we could keep a GTA01/GTA02 compatible
debug connector (because WLAN could use SPI1 instead of SPI0).

> I actually have the same question for the audio-amp: why remove it?
> But that one is a bit more complicated, because I'm not sure what is
> this "audio-amp" anyway (is it the thing that drives the
> headphone plug?)

   It drives either the headphone speakers or the bottom handset speaker,
depending on the presence of the headphone plug. It's the LM4853 (U4101 on
page 7 of the schematics). IIUC, the GTA03 was going to drive both from the
WM8753L sound chip directly as suggested in the WM8753L datasheet. I wanted
to compare the output power of the two, but I can't find the exact LM4853
variant we're using.

Rask Ingemann Lambertsen
Danish law requires addresses in e-mail to be logged and stored for a year

More information about the community mailing list