[M800] Touchscreen filters are not working

Joerg Reisenweber joerg at openmoko.org
Tue Apr 14 00:06:58 CEST 2009


Am Mo  13. April 2009 schrieb morphis:
> On 09.04.2009 06:25, Nelson Castillo wrote:
> > On Wed, Apr 8, 2009 at 11:16 AM, Nelson Castillo
> > <arhuaco at freaks-unidos.net> wrote:
> > (cut)
> >>> Sorry I forgot to add the link. Here it is:
> >>> 
http://git.gnufiish.org/?p=gnufiish.git;a=blob;f=arch/arm/mach-s3c2440/mach-glofiish.c;h=bdf4d1e2d235a47ddc0c282d33fd7df1fa94036a;hb=HEAD
> > 
> > BTW, I'm working on this code now and I had trouble with "delay". Thus
> > I checked yours to see what you were using.
> > 
> > 376         .delay = 10000,
> > 377         .presc = 50000000 / 1000000, /* 50 MHz PCLK / 1MHz */
> > 378         //.presc = 0xff, /* slow as we can go */
> > 
> > You might want to use the original .presc until you get the TS to work
> > the first time... I tested with the values you have and the TS
> > wouldn't work for me.
> 
> Ok, I merged with current andy-tracking to get a current ts filter chain
> but it is still the result. I produced a little example for the current
> problem with touch_test.py. I only pressed my finger on one point of the
> screen. The app reports:
> Move: (352, 242)
> Down: (352, 242)
> Move: (380, 253)
> Move: (394, 257)
> Move: (409, 268)
> Move: (420, 284)
> Move: (426, 291)
> Move: (429, 295)
> Move: (430, 298)
> Move: (433, 301)
> Move: (433, 302)
> Move: (434, 302)
> Move: (434, 303)
> Move: (433, 303)
> Move: (434, 303)
> Move: (433, 303)
> Move: (436, 304)
> Move: (435, 304)
> Move: (434, 304)
> Move: (433, 304)
> Move: (430, 304)
> Move: (431, 304)
> Move: (434, 304)
> Move: (433, 304)
> Move: (434, 304)
> Move: (435, 303)
> Move: (436, 303)
> Move: (434, 300)
> Move: (430, 292)
> Up: (430, 292)
> 
> So the cursor moves up to top left corner while I am press one point of
> the screen. Any hints which calibration of the filters is wrong?
> 
> 
As your finger doesn't create a point of contact on ts, but rather a quite 
large area, this dump looks quite smooth and reasonable.
Before complaining about performance of a particular subsystem, it's mandatory 
to learn basics about it's operation principles.

/j

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part.
Url : http://lists.openmoko.org/pipermail/openmoko-kernel/attachments/20090413/697d2e0b/attachment.pgp 


More information about the openmoko-kernel mailing list