[beagleboard] 4dcape-70t lcd and jitter issues

2014-02-07 Thread Alex Gal
Hello,

I tried the latest kernel from git://github.com/beagleboard/kernel.git
that contains Mikka's patch to fix the jitter on lcd7 and 4dcape-70t
and I noticed the following:

When I start the system, once I touch the screen, hexdump
/dev/input/event1 starts dumping data without stopping.

In tslib (ts_test), the cursor "goes" to the bottom left corner.

As soon as I issue an
echo1 > /sys/class/graphics/fb0/blank
the data from /dev/input/event1 stops.

This behavior persists until the screen blanks on its own (tilcdc
4830e000.fb: timeout waiting for framedone)

After that, the screen becomes usable however there is still jitter.
There seems to be an event generated at the bottom left corner.

The screen was calibrated with ts_calibrate.

I have tested the latest debian image from RCN and the same issue is present.

In this image, the screen was calibrated when X started.

Can other people confirm this ? So far I tried this with two
4DCape-70T. LCD7 from CircuitCo does not have this issue.

-- 
For more options, visit http://beagleboard.org/discuss
--- 
You received this message because you are subscribed to the Google Groups 
"BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to beagleboard+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


[beagleboard] Re: 4dcape-70t lcd and jitter issues

2014-02-07 Thread Alex Gal
Hello all,

I forgot to mention that adding a udelay of 5 microseconds seems to
make the issue go away:


diff --git a/drivers/input/touchscreen/ti_am335x_tsc.c
b/drivers/input/touchscreen/ti_am335x_tsc.c
index ed83043..e1c0499 100644
--- a/drivers/input/touchscreen/ti_am335x_tsc.c
+++ b/drivers/input/touchscreen/ti_am335x_tsc.c
@@ -227,6 +227,7 @@ static void titsc_read_coordinates(struct titsc *ts_dev,
nY=0;
for (i = 0; i < fifocount; i++) {
read = titsc_readl(ts_dev, REG_FIFO0);
+   udelay(5);

channel = (read & 0xf) >> 16;
read &= 0xfff;


So this bug seems to be related perhaps to synchronization issues, I
do not know. I noticed that the bug went away when I added printk
statements. Unfortunately, that does not help me with the debugging.

If anybody has any ideas, let me know.

Cheers,

Alex

-- 
For more options, visit http://beagleboard.org/discuss
--- 
You received this message because you are subscribed to the Google Groups 
"BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to beagleboard+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


Re: [beagleboard] Re: 4dcape-70t lcd and jitter issues

2014-02-28 Thread Alex Gal
Hi Daniel,

> Alex, can you help me, I have the exact same issue.  Could you upload your
> image that fixes the issue with the 5 micro seconds delay.  Are you running
> debian or Ubuntu?  I am not an expert on rebuilding the kernel to change the
> ti_am335x_tsc.c.

I use debian and angstrom.

I modified RCN's kernel with a 15 microsecond hack.

I'll email you directly the contents of deploy:

3.8.13-bone40.1-dtbs.tar.gz
3.8.13-bone40.1-modules.tar.gz
config-3.8.13-bone40.1
3.8.13-bone40.1-firmware.tar.gz
3.8.13-bone40.1.zImage

Cheers,

Alex

-- 
For more options, visit http://beagleboard.org/discuss
--- 
You received this message because you are subscribed to the Google Groups 
"BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to beagleboard+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


Re: [beagleboard] Re: 4dcape-70t lcd and jitter issues

2014-03-04 Thread Alex Gal
Hi Dan,

Glad it worked.

It is still a hack and there is most probably a better solution. I
still get some random clicks in the bottom left corner from time to
time.

Cheers,

Alex



On Mon, Mar 3, 2014 at 2:39 PM,   wrote:
> Alex, it appears your files are working.  I can't thank you enough for the
> patch.  I simply replaced all the files by mounting a previous SD card image
> and copying all the files to the appropriate locations.  Your code fixes the
> issue.  Again thanks thanks thanks.
>
> Dan
>
>
> On Sunday, March 2, 2014 12:27:53 PM UTC-5, daniel...@gmail.com wrote:
>>
>> Any chance you could upload a full image file of debian with your fix to a
>> site like drop box or something?  One that I could use Win32DiskImager to
>> burn to a 4 gig micro sd card?
>>
>> I would really appreciate it.
>>
>> Thanks.
>>
>> Dan
>>
>> On Friday, February 28, 2014 11:44:19 AM UTC-5, Alex Gal wrote:
>>>
>>> Hi Daniel,
>>>
>>> > Alex, can you help me, I have the exact same issue.  Could you upload
>>> > your
>>> > image that fixes the issue with the 5 micro seconds delay.  Are you
>>> > running
>>> > debian or Ubuntu?  I am not an expert on rebuilding the kernel to
>>> > change the
>>> > ti_am335x_tsc.c.
>>>
>>> I use debian and angstrom.
>>>
>>> I modified RCN's kernel with a 15 microsecond hack.
>>>
>>> I'll email you directly the contents of deploy:
>>>
>>> 3.8.13-bone40.1-dtbs.tar.gz
>>> 3.8.13-bone40.1-modules.tar.gz
>>> config-3.8.13-bone40.1
>>> 3.8.13-bone40.1-firmware.tar.gz
>>> 3.8.13-bone40.1.zImage
>>>
>>> Cheers,
>>>
>>> Alex
>
> --
> For more options, visit http://beagleboard.org/discuss
> ---
> You received this message because you are subscribed to the Google Groups
> "BeagleBoard" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to beagleboard+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.

-- 
For more options, visit http://beagleboard.org/discuss
--- 
You received this message because you are subscribed to the Google Groups 
"BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to beagleboard+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


Re: [beagleboard] Re: Android running on BBB with Linux 3.8

2014-03-16 Thread Alex Gal
Hi,

Check out this thread:

http://beagleboard.org/Community/Forums/?place=msg%2Fbeagleboard%2FnwS91gX6dtk%2FKzn5LeytsCUJ

I had the same issue with the 4DCAPE-70T.

I replaced the delay to 15 ms and the issue went away but it still
feels like a hack.


On Sun, Mar 16, 2014 at 4:11 PM, paccer  wrote:
> I've been looking into the issue some more (sorry for the delay), and I've
> still not been successful on getting the touch screen to work with 3.8.
> I am getting touch interrupts, but i don't appear to be able to calibrate
> myself out of the problem.. (?)
>
> After touching the screen the first time, events just keep scrolling
> indefinitely..
> it appears that calibration is (rather) correct, but whenever i "release",
> its behaving like i'm still touching a part of the screen..
>
> I recorded a short video of the issue, in hope that someone might be able to
> help me identify what's going on, and how to fix it.
> https://www.youtube.com/watch?v=we--uPXqye8
>
> Thanks
>
>
> On Thursday, February 27, 2014 8:01:46 AM UTC+1, Vishveshwar wrote:
>>
>>
>> There are two things w.r.t touchscreen:
>> - The touch operation itself - is the driver loaded? are you getting touch
>> interrupts?
>> - Calibration - mapping the usable area so that the touch interacts with
>> the correct display element (icon etc)
>>
>> Some information is at the AOSP page:
>> http://source.android.com/devices/tech/input/touch-devices.html
>>
>> FYI, In the TI DevKit kernel (3.2) the touch calibration data is part of
>> the platform data for the touch driver.
>> See my post in android-porting for more information about this:
>>
>> https://groups.google.com/d/msg/android-porting/CpP4xCk0U_g/DaCoAmqj6FwJ
>>
> --
> For more options, visit http://beagleboard.org/discuss
> ---
> You received this message because you are subscribed to the Google Groups
> "BeagleBoard" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to beagleboard+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

-- 
For more options, visit http://beagleboard.org/discuss
--- 
You received this message because you are subscribed to the Google Groups 
"BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to beagleboard+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [beagleboard] 4DCAPE-70T display blocks all analog inputs

2014-08-12 Thread Alex Gal
Hi,

On 12 August 2014 09:37,   wrote:
> Has anyone figured out how to access the analog inputs 4, 5, 6 that are
> supposedly unused?

I am using the 4DCAPE-70T.

What do you have in /sys/devices/ocp.3/44e0d000.tscadc/tiadc/iio:device0  ?

I see:
in_voltage4_raw
in_voltage5_raw
in_voltage6_raw
in_voltage7_raw

Everytime I cat these files, their values change:

debian@hmi2:/sys/devices/ocp.3/44e0d000.tscadc/tiadc/iio:device0$ cat
in_voltage*
531
1472
2069
3812
debian@hmi2:/sys/devices/ocp.3/44e0d000.tscadc/tiadc/iio:device0$ cat
in_voltage*
448
1289
2067
3824


Cheers,

Alex

-- 
For more options, visit http://beagleboard.org/discuss
--- 
You received this message because you are subscribed to the Google Groups 
"BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to beagleboard+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [beagleboard] BBB UART2 Write Issue (C/C++)

2014-04-17 Thread Alex Gal
Hi Michael,

> I'm currently using the BBB to control a rover robot for a project at my
> university. I'm currently having an issue using the UART file /dev/ttyOx to
> write an array of characters to a smart servo using c/c++.
> More specifically, I've opened the file using
> fopen("/dev/ttyO2", "wb")
>
> which should open it in binary mode. I've also tried the c++ equivalent with
> ofstream file("/dev/ttyO2", ios::out|ios::binary)
>
> For the most part these have worked fine. I have been using the UART tool
> for the PicKit 2 to verify what I've been receiving. But, despite having
> opened the file in binary mode, if I try to write 0x0A it will still append
> 0x0D to form CRLF, or newline.
>
> For example, if I create a character array:
> char c[] = {0xFF,0xFF,0x0A,0x01,..}
>
> This is seen on the UART tool:
>  FF FF 0D 0A 01 ...
>
> If anyone knows a fix, or can offer any insight (should I be approaching it
> incorrectly), I would greatly appreciate it.

I believe you have to properly configure your serial port (baud rate,
line discipline, etc).

See this wiki page:

http://en.wikibooks.org/wiki/Serial_Programming/termios

-- 
For more options, visit http://beagleboard.org/discuss
--- 
You received this message because you are subscribed to the Google Groups 
"BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to beagleboard+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.