Hello Pierre,
> this patch takes into account the new touchscreen API for aximx50.
Sorry for this confusion, I eventually updated all ads7846ts-ssp
client code myself, there was just delay while we settled some
details. By now, that code is in CVS and builds, please update and
rediff, and send new patch if something is missing. One thing I can
spot now however:
Sunday, February 18, 2007, 12:23:04 PM, you wrote:
> diff -rNu -x CVS -x .project -x .cdtproject -x .settings -x debian
> kernel26_vanilla/arch/arm/mach-pxa/aximx50/aximx50_ts.c
> kernel26/arch/arm/mach-pxa/aximx50/aximx50_ts.c
> --- kernel26_vanilla/arch/arm/mach-pxa/aximx50/aximx50_ts.c
> 2007-02-17 21:10:39.834559729 +0100
> +++ kernel26/arch/arm/mach-pxa/aximx50/aximx50_ts.c 2007-02-18
> 09:49:30.717482213 +0100
> @@ -30,6 +30,7 @@
> struct ads7846_ssp_platform_data aximx50_ts_ssp_params = {
> .port = 1,
> .pd_bits = 1,
> + .freq = 100000,
> };
> static struct platform_device ads7846_ssp = {
> .name = "ads7846-ssp",
Are you sure you want 100KHz here? Your previous patch, with
SSCR0_SerClkDiv(7) corresponds to 13MHz / 7 = 1.85MHz on PXA27x. What
I found (well, confirmed, Philipp Zabel had that idea) is that
SSP frequency influences TS behavior, and with our current debouncing
algorithm, it needs adjustment for specific model. For example, h4000
works well at 100KHz and hx4700 at 720KHz. We'll need to make
debouncing params adjustable, and after that, maybe algo too. In the
meantime, please test few freq values and tell what works the best.
--
Best regards,
Paul mailto:[EMAIL PROTECTED]
--
_______________________________________________
Aximx50-port mailing list
[email protected]
https://www.handhelds.org/mailman/listinfo/aximx50-port
