Re: 5" touchscreen issues
What are the libraries and teensy software version that were known to work? I¡¯m suspicious of the V1.52 teensy libraries since the problem occurs using the recommended RA8875 library and latest JackAl s/w. Perhaps brute force but maybe a straight up comparison of what worked vs. current is a good starting point?
Joe
|
Re: 5" touchscreen issues
I am using the JackAl RA8875 library with the JackAl.
I have another Teensy 3.6 on a breadboard that I connected to the LCD via the protection board instead of the JackAl which I was trying some of the example sketches with the other library to try an a-b comparison.
|
Re: 5" touchscreen issues
Did you download our version of the display library?
Jack, W8TEE
On Monday, May 25, 2020, 8:06:32 PM EDT, jason kitchens <jasonmkitchens@...> wrote:
I would say that the issues are not serious enough for you to put yourselves in peril! HiHi!
I've been working on installing the uBitx and JackAl board into the case before doing much more digging into the issue.
I did try an example sketch from sumotoy's RA8875 library with a teensy 3.6 on a breadboard with no luck at all.
I have preliminary mounts for the boards working, so I hope to dig into testing the display and related software over the coming days.
|
Re: 5" touchscreen issues
I would say that the issues are not serious enough for you to put yourselves in peril! HiHi!
I've been working on installing the uBitx and JackAl board into the case before doing much more digging into the issue.
I did try an example sketch from sumotoy's RA8875 library with a teensy 3.6 on a breadboard with no luck at all.
I have preliminary mounts for the boards working, so I hope to dig into testing the display and related software over the coming days.
|
Re: Video of my touch screen issues
Had the same issue...I abandoned the project and sold all..
|
Re: 5" touchscreen issues
Al and I are going to defy the Stay-In-Your-House-Or-Else edict and try to get together after the holiday to see if we can provide a fix to your problem.
Jack, W8TEE
On Saturday, May 23, 2020, 4:13:51 PM EDT, jjpurdum via groups.io <jjpurdum@...> wrote:
On Saturday, May 23, 2020, 2:06:03 PM EDT, Joseph Giallo II <jgiallo@...> wrote:
In looking at my trace outputs it seems that my tx values are in the 400's and the ranges in the GetTouchCoordinates() code goes up to 750. Are the scale factors in the code set to a 5" display or a 7" display??
|
Re: 5" touchscreen issues
On Saturday, May 23, 2020, 2:06:03 PM EDT, Joseph Giallo II <jgiallo@...> wrote:
In looking at my trace outputs it seems that my tx values are in the 400's and the ranges in the GetTouchCoordinates() code goes up to 750. Are the scale factors in the code set to a 5" display or a 7" display??
|
Re: 5" touchscreen issues
In looking at my trace outputs it seems that my tx values are in the 400's and the ranges in the GetTouchCoordinates() code goes up to 750. Are the scale factors in the code set to a 5" display or a 7" display??
|
Re: 5" touchscreen issues
No joy on the Jacks 7 reference (and variations thereof) other than in the JackAl.h file and appears commented out - no other instances found. In comparing the active vs. inactive code in DisplayCalibration.cpp we see an apparent difference in variable defs (long vs. int). The active code in GetTouchCoordinates() is:
calty = map( (long) yFinal, (long) yMapLowAve, (long) yMapHighAve, (long) 50, (long) 430);? ? ? ?// Jack
?
? range = map((long)calty, 0L, 479L, 1L, 4L);
? switch (range) {
? ? case 1:
? ? ? caltx = map((long) xFinal, (long) xMapLowAve1, (long) xMapHighAve, (long) 50, (long) 750); // Map from ADC values to Screen? ? ?// Jack
? ? ? break;
? ? case 2:
? ? ? caltx = map((long) xFinal, (long) xMapLowAve2, (long) xMapHighAve, (long) 50, (long) 750); // Map from ADC values to Screen? ? // Jack
? ? ? break;
? ? case 3:
? ? ? caltx = map((long) xFinal, (long) xMapLowAve3, (long) xMapHighAve, (long) 50, (long) 750); // Map from ADC values to Screen? ? // Jack
? ? ? break;
? ? case 4:
? ? ? caltx = map((long) xFinal, (long) xMapLowAve4, (long) xMapHighAve, (long) 50, (long) 750); // Map from ADC values to Screen? ? // Jack
? ? ? break;
? }
? whichButton = myButton.WhichButton(caltx, calty);
}
And the inactive code in GetTouchCoordinates2() is:
?/* calty = map(yFinal, yMapLowAve, yMapHighAve, 50, 430);? ? ? ?// Jack
?
? range = map(int(calty), 0, 479, 1, 4);
? switch (range) {
? ? case 1:
? ? ? caltx = map(xFinal, xMapLowAve1, xMapHighAve, 50, 750); // Map from ADC values to Screen? ? ?// Jack
? ? ? break;
? ? case 2:
? ? ? caltx = map(xFinal, xMapLowAve2, xMapHighAve, 50, 750); // Map from ADC values to Screen? ? // Jack
? ? ? break;
? ? case 3:
? ? ? caltx = map(xFinal, xMapLowAve3, xMapHighAve, 50, 750); // Map from ADC values to Screen? ? // Jack
? ? ? break;
? ? case 4:
? ? ? caltx = map(xFinal, xMapLowAve4, xMapHighAve, 50, 750); // Map from ADC values to Screen? ? // Jack
? ? ? break;
? }*/
? whichButton = myButton.WhichButton(caltx, calty);
}
|
Re: Video of my touch screen issues
Very much the same as what I seen with the 7¡± only the performance was a bit worse. ?
toggle quoted message
Show quoted text
On May 23, 2020, at 1:08 PM, Joseph Giallo II <Jgiallo@...> wrote:
?I looked at the video - virtually identical to what I'm seeing.
Joe
|
Re: 5" touchscreen issues
As I recall, there are some sections in the code that have "Jacks 7" in them. Do a global search on that and see if they have any clues. (I can't get to my machine right now.)
Jack, W8TEE
On Saturday, May 23, 2020, 1:00:48 PM EDT, Joseph Giallo II <jgiallo@...> wrote:
I went back to ground zero following the library assignments in Jack's color coded post. It looks like calibration runs as intended, however, for some reason touches are not properly mapped to button locations. Looks like everything is shifted right about one button more or less. Sometimes this results in touch coordinates that can't be interpreted so it looks like the touch screen is unresponsive but the serial monitor shows touches are being sensed. When I look at the display calibration code (DisplayCalibration.cpp) there is an entire section of GetTouchCoordinates2() that is commented out and this is not the case in GetTouchCoordates() which runs initially. Playing with uncommenting this section doesn't change the behavior unfortunately. There are function calls at the end of each case statement, i.e.?whichButton = myButton.WhichButton(caltx, calty) that would seem to identify button location. Maybe these are not working as they once did??
73,
Joe
AF4XS
|
Re: Video of my touch screen issues
I looked at the video - virtually identical to what I'm seeing.
Joe
|
Video of my touch screen issues
Like Joe some of my touch points seem to be off even after a calibration.
I make a few minute video showing the behavior in case it's helpful.
Has anyone taken a video of their JackAl for comparison either 100% working or not?
Here's a link -?
I have it unlisted at the moment but anyone with the link should be able to view.
|
Re: 5" touchscreen issues
I went back to ground zero following the library assignments in Jack's color coded post. It looks like calibration runs as intended, however, for some reason touches are not properly mapped to button locations. Looks like everything is shifted right about one button more or less. Sometimes this results in touch coordinates that can't be interpreted so it looks like the touch screen is unresponsive but the serial monitor shows touches are being sensed. When I look at the display calibration code (DisplayCalibration.cpp) there is an entire section of GetTouchCoordinates2() that is commented out and this is not the case in GetTouchCoordates() which runs initially. Playing with uncommenting this section doesn't change the behavior unfortunately. There are function calls at the end of each case statement, i.e.?whichButton = myButton.WhichButton(caltx, calty) that would seem to identify button location. Maybe these are not working as they once did??
73,
Joe
AF4XS
|
Re: 5" touchscreen issues
Let me talk with Al first...
Jack, W8TEE
On Saturday, May 23, 2020, 12:33:41 PM EDT, James Zdunic <james_zdunic@...> wrote:
Jack,
I¡¯m local and have a 7¡± display that I¡¯d be willing to let you guys use for revisiting this. ?Let me know if you¡¯re interested.
On May 23, 2020, at 9:07 AM, jjpurdum via groups.io <jjpurdum@...> wrote:
?
Al and I talked about this yesterday, and we (vaguely--we're old) remember the coordinate system was "reversed" on the 7" system. If you can, put in some Serial.print()
statements in the calibration routine and see if they look that way. Meanwhile, we're trying to dig out a 7" display on this end.
Jack, W8TEE
On Saturday, May 23, 2020, 12:07:00 AM EDT, Joseph Giallo II <jgiallo@...> wrote:
Yup, they make sense and are in fact being stored. Now trying to see if the issue is in the RA8875 library. Curiously using the sumotoy 8875 library referenced in some JackAl literature produces a compile error: "class RA8875 has no member named touch".
Interesting since the problem is not recognizing touches. The RA8875 library off of the groups.io JackAl site does compile without the error but exhibits the original problem I described. Seems like the issue is somewhere in the RA8875, URTouch, or UTFT libraries....investigative
suggestions welcome.
Joe
|
Re: 5" touchscreen issues
I have the same touch freeze issue on the 7¡± screen.
toggle quoted message
Show quoted text
On May 23, 2020, at 10:14 AM, Joseph Giallo II <Jgiallo@...> wrote:
?
[Edited Message Follows]
[Reason: note correction to last sentence - touch recognition seems to be freezing after initially working (albeit incorrectly)]
OK - will do. Note that I'm using a 5" display but latest behavior definitely indicates what is recognized is shifted from what is actually touched so some sort of coordinate reversing is a possibility. I have gone back to using the time, timerone libraries
referenced in your "libraries" document to see this latest behavior. It does seem like the some of the latest teensy libraries are somehow different from those I understand other users have previously shown to work. Also, after a few touches the screen stops
reacting, almost as if there is some sort of buffer overflow occuring where new touches aren't recognized on the screen and the debug trace doesn't register them either....
Joe
AF4XS
|
Re: 5" touchscreen issues
Jack,
I¡¯m local and have a 7¡± display that I¡¯d be willing to let you guys use for revisiting this. ?Let me know if you¡¯re interested.
toggle quoted message
Show quoted text
On May 23, 2020, at 9:07 AM, jjpurdum via groups.io <jjpurdum@...> wrote:
?
Al and I talked about this yesterday, and we (vaguely--we're old) remember the coordinate system was "reversed" on the 7" system. If you can, put in some Serial.print()
statements in the calibration routine and see if they look that way. Meanwhile, we're trying to dig out a 7" display on this end.
Jack, W8TEE
On Saturday, May 23, 2020, 12:07:00 AM EDT, Joseph Giallo II <jgiallo@...> wrote:
Yup, they make sense and are in fact being stored. Now trying to see if the issue is in the RA8875 library. Curiously using the sumotoy 8875 library referenced in some JackAl literature produces a compile error: "class RA8875 has no member named touch".
Interesting since the problem is not recognizing touches. The RA8875 library off of the groups.io JackAl site does compile without the error but exhibits the original problem I described. Seems like the issue is somewhere in the RA8875, URTouch, or UTFT libraries....investigative
suggestions welcome.
Joe
|
Re: 5" touchscreen issues
OK - will do. Note that I'm using a 5" display but latest behavior definitely indicates what is recognized is shifted from what is actually touched so some sort of coordinate reversing is a possibility. I have gone back to using the time, timerone libraries referenced in your "libraries" document to see this latest behavior. It does seem like the some of the latest teensy libraries are somehow different from those I understand other users have previously shown to work. Also, after a few touches the screen stops reacting, almost as if there is some sort of buffer overflow occuring where new touches aren't recognized on the screen and the debug trace doesn't register them either....
Joe
AF4XS
|
Re: 5" touchscreen issues
Al and I talked about this yesterday, and we (vaguely--we're old) remember the coordinate system was "reversed" on the 7" system. If you can, put in some Serial.print() statements in the calibration routine and see if they look that way. Meanwhile, we're trying to dig out a 7" display on this end.
Jack, W8TEE
On Saturday, May 23, 2020, 12:07:00 AM EDT, Joseph Giallo II <jgiallo@...> wrote:
Yup, they make sense and are in fact being stored. Now trying to see if the issue is in the RA8875 library. Curiously using the sumotoy 8875 library referenced in some JackAl literature produces a compile error: "class RA8875 has no member named touch". Interesting since the problem is not recognizing touches. The RA8875 library off of the groups.io JackAl site does compile without the error but exhibits the original problem I described. Seems like the issue is somewhere in the RA8875, URTouch, or UTFT libraries....investigative suggestions welcome.
Joe
|
Re: 5" touchscreen issues
Yup, they make sense and are in fact being stored. Now trying to see if the issue is in the RA8875 library. Curiously using the sumotoy 8875 library referenced in some JackAl literature produces a compile error: "class RA8875 has no member named touch". Interesting since the problem is not recognizing touches. The RA8875 library off of the groups.io JackAl site does compile without the error but exhibits the original problem I described. Seems like the issue is somewhere in the RA8875, URTouch, or UTFT libraries....investigative suggestions welcome.
Joe
|
jjpurdum