Re: [Kernel-discuss] Re: ads7846ts_ssp elaborated, was: Re: new patch

From: pHilipp Zabel <philipp.zabel_at_gmail.com>
Date: Sat, 10 Feb 2007 00:42:34 +0100

On 2/10/07, Pierre GAUFILLET <pierre.gaufillet_at_magic.fr> wrote:
>
> Well, I tested quite a few SSP configurations for TSC2046...
>
> The current ads7846_adc_ssp.c does not work for Axim X50 family.
> I had to modify ads7846_init_chip() to get something, in the following way
> :
> SSCR0_P(port) = 0;
> pxa_set_cken(CKEN23_SSP1, 1);
> SSCR0_P(port) |= SSCR0_DataSize(12); /* 12 bits */
> SSCR0_P(port) |= SSCR0_Motorola;
> SSCR0_P(port) |= 0x11 << 8; /* ~100 KHz */

Btw, this is the same on pxa25x and pxa27x?
The hx4700 (24bit) driver used SSCR0_Motorola and (0x6 << 8).

> SSCR1_P(port) = 0;
> SSCR0_P(port) |= SSCR0_SSE;

What is the difference between Motorola SPI and National
Semiconductor's Microwire mode?

> Even with these modifications, all I get is x=2048, y=2048...

I tried this, too, and I get the same. (Communication with
SSCR0_National works here, though).

> Communications on 24 bits as in my previous patch seems the only stable
> configuration.
> I do not understand why : there is no special constraint at this level in
> the TSC 2046 datasheet...
> Nervertheless, I then get linear measurements in the [5000, 30000]x[5000,
> 30000] range, that seem useable.

With the driver as-is I sometimes get oscillating Y measurements.
Trying to get a result by averaging causes some noise in the Y signal.

regards
Philipp
Received on Fri Feb 09 2007 - 18:42:42 EST

This archive was generated by hypermail 2.2.0 : Fri Feb 09 2007 - 18:42:58 EST