5500 Error - Will No Longer Calibrate?

[Cy1clown]

I checked the version with cat /proc/deviceinfo/revision and I got 3.1
The 2.78 is from the bottom of the screee when it boots up just before I hit / to go into the console mode.

I am using the AC adapter when I try all these items.

I have reflashed the Sharp ROM 3.1 from CF card and that has not helped.

The console function seems to be working fine.  I just can't get past the calibration screen on boot up.

Could this be a hardware problem?  It does recognize that I have tapped the screen, it clicks and moves to the next position.

Thanks
Jon

[dougeeebear]

I've noticed on my 5500 that I have to hit the calibration crosshairs right on target, or it just keeps looping around.
I think it's due to the low screen resolution.

[Cy1clown]

dougeebear

Even when I am very careful about hitting the mark on the crosshairs it keeps cycling.

Thanks for the suggestion though.

Anybody have other ideas on how to get past the calibration screen?

Jon

[sdjf]

I've noticed on my 5500 that I have to hit the calibration crosshairs right on target, or it just keeps going around.
I think it's due to the low screen resolution.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=153619\"][{POST_SNAPBACK}][/a][/div]

Great thought.  Now that I think about it, it has you tap the positions, and then sorta rechecks their accuracy.

If he hits the exact same spots each run through, then maybe the screen is scratched at those spots and he could try purposefully tapping at, say, a mm or two to the side and see what happens.

sdjf

[dougeeebear]

Ok, I just reset my 5500 and it went through calibration process with no problem.
At this point, I have no idea what could be causing yours to loop.

[Cy1clown]

Unfortunately it doesn't matter how I tap the screen ti still cycles.

I have tried exactly on target
I have tried various inaccuracies, both smaller than and larger than the target square.

Still no luck.

Thanks for all the suggestions

Don't know what to try next maybe reflash with OZ.
I have tried this but it won't flash can anyone walk me through the "exact" steps required from the download on?

Jon

[quickening]

I have 3.10 too.   Pointer calibration is a test qtopia runs when it first comes up.  It should be creating a file /etc/pointercal.  Mine looks like this:

21596 0 173612 0 14745 978165 65536

You can log what qtopia is doing by uncommenting the line in /opt/QtPalmtop/qpe.sh

It's my guess that filesystem errors (bad ram) are forcing root mount into read-only.  Try creating a file.

[Cy1clown]

quickening

Your suggestion of a bad ram spot makes sense.
I am not very good at linux can you tell me step by step how I would create a file as you suggest.
I can now get into the linux console just need to know the speciifc commands to do what you are suggesting.

Thanks for the help
Jon

[quickening]

quickening

Your suggestion of a bad ram spot makes sense.
I am not very good at linux can you tell me step by step how I would create a file as you suggest.
I can now get into the linux console just need to know the speciifc commands to do what you are suggesting.

[div align=\"right\"][a href=\"index.php?act=findpost&pid=154106\"][{POST_SNAPBACK}][/a][/div]

from the console,

cd /etc
ls -l pointercal

If it's not there, do
echo "21596 0 173612 0 14745 978165 65536" > pointercal

If the filesystem is read-only , you will get an error message.

do "ls -l pointercal" again, to verify it updated.  exit from console, and it should start up.

[sdjf]

Jon,

to understand the ls command output better, look at this:
http://www.scism.sbu.ac.uk/law/UnixStuff/ls.html

My 2.38 shows read and write permission for root for that file.
# ls -l /etc/pointercal
-rw-rw-rw-    1 root     root            35 Feb  9 23:39 /etc/pointercal

[Cy1clown]

quickening

I did thisa part of your seggestion and got nothing

from the console,

cd /etc
ls -l pointercal

I then performed the second part.

If it's not there, do
echo "21596 0 173612 0 14745 978165 65536" > pointercal

If the filesystem is read-only , you will get an error message.

do "ls -l pointercal" again, to verify it updated. exit from console, and it should start up.

And when I exoted the console it started up without the calibration screen.
I can now get to the normal qtopia stuff but obviously the touch screen is way off.
If I go into calibrate it will just cycle through the points like it has been doing.  I can cancel out of that and get back to the menues again.

Are those numbers in the command locations on the screen?
If I fiddle with them do you think I can get the touch screen working?

Thanks for the help so far, I hope I can get this resolved I wan't to use my Zaurus at work.

Jon

[Cy1clown]

I just thought of something.  I do have a backup of the system when it was working properly.  If I restore to this will it reset my calibration to the settings it had back then?

Jon

[Cy1clown]

I restored from my back up and the restore worked fine but my calibration is still messed up.

Anyone have any Ideas on this?

Jon

[Cy1clown]

sdjf

what are the numbers in your pointercal file.

I have tried the ones that quickening gave plus one other I found in this forum and they are not close to what mine needs to work right.

Jon

[quickening]

I restored from my back up and the restore worked fine but my calibration is still messed up.
[div align=\"right\"][{POST_SNAPBACK}][/a][/div]

That's the puzzler - I would expect the hardware to stay constant and your pointercal values not to change.  However, I looked at my last 3 pointercal's:

21596 0 173612 0 14745 978165 65536
20830 0 674620 0 14951 919510 65536
21541 0 328964 0 15182 766702 65536

I just ran re-calibrate and got this:
21596 0 346380 0 14857 926090 65536

and you can see some of the numbers change quite a lot.  There are some clues here about what they represent:

[a href=\"http://lists.arm.linux.org.uk/pipermail/linux-arm/2005-May/009880.html]http://lists.arm.linux.org.uk/pipermail/li...May/009880.html[/url]

Appears to be very sensitive and will keep prompting you to reenter if result is inconsistent.