Toolbar is functional but invisible; Settings are not preserved over restart

[oloendithas]

Describe the bug
I have built your wonderful repo for my Aquila S2. Now everything is working, I'm experiencing several bugs:

1. Toolbar is functional but invisible;
2. Not any of settings are preserved over restart, except for the bed leveling mesh.

To Reproduce
Steps to reproduce the behavior:
I used the default branch Mriscoc-Fall. Maybe I should use the 2.1.3 tag?
These are configs I build with
Configuration.h
Configuration_adv.h

Should it be visible? May by I disabled some required thing?

Also applied couple of bugfixes:

1. Raise the nozzle before travel Y on abort. Otherwise it was just dragging the nozzle over the bed:

diff --git a/Marlin/src/lcd/e3v2/proui/dwin.cpp b/Marlin/src/lcd/e3v2/proui/dwin.cpp
index 9fc4e16..9da8f38 100644
--- a/Marlin/src/lcd/e3v2/proui/dwin.cpp
+++ b/Marlin/src/lcd/e3v2/proui/dwin.cpp
@@ -1660,8 +1660,8 @@ void DWIN_Print_Finished() {
 void DWIN_Print_Aborted() {
   TERN_(DEBUG_DWIN, SERIAL_ECHOLNPGM("DWIN_Print_Aborted"));
   #if ProUIex
-    char cmd[21] = "";
-    sprintf_P(cmd, PSTR("G0 F3000 Y%i\nM84"), PRO_data.Park_point.y);
+    char cmd[31] = "";
+    sprintf_P(cmd, PSTR("G0 F3000 Z%i\nG0 F3000 Y%i\nM84"), PRO_data.Park_point.z, PRO_data.Park_point.y);
     queue.inject(cmd);
   #endif
   DWIN_Print_Finished();

2. Max X/Y for mesh inset was misplaced:

diff --git a/Marlin/src/lcd/e3v2/proui/dwin.cpp b/Marlin/src/lcd/e3v2/proui/dwin.cpp
index 9fc4e16..7cf5b74 100644
--- a/Marlin/src/lcd/e3v2/proui/dwin.cpp
+++ b/Marlin/src/lcd/e3v2/proui/dwin.cpp
@@ -1777,9 +1777,9 @@ void DWIN_SetDataDefaults() {
     #if HAS_MESH
       PRO_data.grid_max_points = DEF_GRID_MAX_POINTS;
       PRO_data.mesh_min_x = DEF_MESH_MIN_X;
-      PRO_data.mesh_max_x = DEF_MESH_MAX_Y;
+      PRO_data.mesh_max_x = DEF_MESH_MAX_X;
       PRO_data.mesh_min_y = DEF_MESH_MIN_Y;
-      PRO_data.mesh_max_y = DEF_MESH_MAX_X;
+      PRO_data.mesh_max_y = DEF_MESH_MAX_Y;
     #endif
     #if HAS_BED_PROBE
       PRO_data.zprobefeedslow = DEF_Z_PROBE_FEEDRATE_SLOW;

diff --git a/Marlin/src/lcd/e3v2/proui/proui.h b/Marlin/src/lcd/e3v2/proui/proui.h
index 46ca6f4..cfe6cb1 100644
--- a/Marlin/src/lcd/e3v2/proui/proui.h
+++ b/Marlin/src/lcd/e3v2/proui/proui.h
@@ -65,7 +65,7 @@ constexpr int16_t DEF_Z_MAX_POS = Z_MAX_POS;
     #define MESH_MAX_X  X_BED_SIZE - (MESH_INSET)
   #endif
   #ifndef MESH_MAX_Y
-    #define MESH_MAX_Y  X_BED_SIZE - (MESH_INSET)
+    #define MESH_MAX_Y  Y_BED_SIZE - (MESH_INSET)
   #endif
   constexpr int16_t DEF_MESH_MIN_X = MESH_MIN_X;
   constexpr int16_t DEF_MESH_MAX_X = MESH_MAX_X;

I hope, this is gonna be useful

[classicrocker883]

hey @oloendithas. thank you very much for addressing and helping with any issues. it's much appreciated.

so as for the toolbar disabled, that happens when you change the authors name in the Configuration.h. I see that in the file u provided, that is the case. it should remain whatever it was originally. its not really fixable unless you know how to open and edit the lib files ending .a

as for the mesh inset not saving from restart, I've been trying to figure that one out.

This happens with a BL touch, it subtracts the offset from the max inset in order to make all points accessible by the probe. I believe changing the grid to either 5x5 or 3x3 may solve this, I'll have to check that though. but with 7x7 I find works great, it probes almost all the points except for the right most column but it estimates what those values will be anyway, and pretty accurately BTW.

I'll look at the code and see about updating the files soon.

thanks again.

[classicrocker883]

so for raising the nozzle, I see you added extra and it comes out to 31 characters

cmd[31] = "";
sprintf_P(cmd, PSTR("G0 F3000 Z%i\nG0 F3000 Y%i\nM84"), PRO_data.Park_point.z, PRO_data.Park_point.y);

as u notice the previous code said 21 but it was less than that. I found it was an issue changing it to the actual length. I think it's because of the 'spaces'. the code may need to read spaces as %20so for the new code that should be changed to a greater number than '31' the characters in the command. so it's the characters plus spaces. so it doesn't hurt for having even a higher number.

[oloendithas]

the code may need to read spaces as %20so for the new code that should be changed to a greater number than '31' the characters in the command

Yeap, it gives me the warning:
'sprintf' output between 28 and 38 bytes into a destination of size 31
but as I'm not a C++ pro (more like a C#), this was just a guess from my part and since it worked with my Y/Z values, I left it like this. Will try to change to 38 and see, if it will clear the warning. Thanks for the clue)

[oloendithas]

that happens when you change the authors name

That dependency was completely unexpected)) Restored the authors name and now toolbar is visible and even more, thee settings are preserved over the restart. Except toolbar buttons, but I'll hardcode that)

Regarding the mesh settings, I was using the manual mesh setup then. Now trying to connect the 3D Touch sensor, bu not succeeding yet. So now built with AUTO_BED_LEVELING_BILINEAR and GRID_MAX_POINTS_X 7.

[classicrocker883]

the 3d touch isn't working? but there's nothing actually wrong with it, just something in the code?

in Configuration_adv.h line ~910 you can try #define BLTOUCH_FORCE_SW_MODE
and BLTouch_Delay

don't enable 5v mode, im pretty sure that may break something. but u can also disable HS mode and see if that works.

[oloendithas]

the 3d touch isn't working? but there's nothing actually wrong with it, just something in the code?

Well, it's a cheap chinese clone, so anything can be wrong. By now it goes it's internal deploy test on start and nothing more. Also, if try to Home, all start moving with vibration and noise. Before I try to access the sensor, they move smoothly.
When I press "Deploy probe" in LCD menu, it raises Z a bit and nothing more. May be some pin number is wrong. Wiring looks correct. Also, I have no idea, what's the difference between last two pins of the BLTouch port (OUT/G) and the actual Z-MIN port. Tried them both, but I guess this will not have effect until IN will work and deploy the probe.

UPD: Just tested the sensor on Arduino and it seem to be working. Maybe some In pin number is not right for my N32 board?

VID_20221229_131116_edit.mp4

[classicrocker883]

ahh I see. do not plug it into the Z port. you must remove pins from the connector and put them in the 2 missing in the other plug.

in the firmware there may be a way to enable this so u don't have to do that. but for all intensive purposes u should have all wires on one plug.

a quick look at another connector with all 5 pins shows the black is on the outside, left most in picture.

[classicrocker883]

did I read you DID try that? then I would try the force_sw_mode and delay from previous comment. but what you can do is leave the plugs separate like they are in picture - two in the Z_min port, and in Configuration.h @ line 1284

uncomment

//#define Z_MIN_PROBE_USES_Z_MIN_ENDSTOP_PIN // Probe connected to BLTouch port

and comment

// Force the use of the probe for Z-axis homing
#define USE_PROBE_FOR_Z_HOMING

as an option you can uncomment

//#define Z_MIN_PROBE_PIN PB1

double check that PB1 is the correct pinout. I believe it is since I think I put it there.

[oloendithas]

Yes, I already tested by multimeter which pin goes to te PB0 of the controller, and it't the fith pin of the BLTouch port (OUT), tried to connect to it, but it doesn't help, because the problem for now is not working servo pin (IN), PB1, if I remember correctly. Did set NUM_SERVOS 1/3, tried to run M280 P1 S10/S90. The prode shaft is not moving. Running out of versions.
Did tried force_sw_mode and the delay. Havent trieb to tinker with #define Z_MIN_PROBE_PIN PB1. In my config it says 32, instead of PB1.
I have some assumptions regarding 5V, but in gives me a bunch of compilation errors and the first one is "cannot be used with Z_MIN_PROBE_USES_Z_MIN_ENDSTOP_PIN".
I've seen some guide suggesting to use sw_mode with 5v. Maybe I should try this way.
So, for now the problem is in not working servo IN.

[classicrocker883]

@oloendithas also, I updated the code with the fixes u suggested. unfortunately the Mesh Inset resets after printer restart. like it says for X_max 185 (or whatever you have for the probe offset), and say I change it to 210. it changes back to 185 after restart.

ive tried looking into this, but it could be out of reach because of the proui lib.a files which aren't editable. so if you can find a work around, maybe just making a new variable and definition or reference would fix it? but all I know it's doing a recursive link to some equation.

I'd look into proui.h, bedlevel_tools.h, and dwin.cpp.
specifically DEF_MESH_MAX_Y or MAX_MESH_INSET or probe.offset.x

[oloendithas]

As the mater of fact I don't understand, how things are stored into the EEPROM. I haven't studied it yet in the Arduino school of myself) The ones I can see from the M503 command, are stored well. But why, for instance, color settings are preserved only with correct author's name is a total magic for me. I can try to dig in when return after a week.

[classicrocker883]

there is just something the original author, Mriscoc, implemented as a sort of safeguard for those wanting to steal his work. he was reluctant at first to release the source code. so he left a good amount of details out and encrypted them in his special little library in the lib folder.

the toolbar is his making, and without giving him credit as the author, he made it so it disappears. In the tests ive done, when comparing Mriscoc ProUI and the Marlin code he based it off, something about changing the delays made a couple things not work. i think what i did was change the delay for startup and the cr touch i have wouldnt work. anyway, just becareful with changing the bl touch to 5v, and see if u can look up more about what yours might be, because i read the warnings pertaining to changing it from open drain to 5v, and it may damage something so do so at your own risk. good luck.

also, u can check out his repo here https://github.com/mriscoc/Ender3V2S1

mriscoc#509
others may have issues with using the bl touch and post about it.

[oloendithas]

I have to check first, is this 5v for servo or for Zmin. Because servo out is already giving 5v and the sensor is rated to 5v, but still servo IN is not working. While the Zmin gives 3.3v. I suspect that this 5v feature is for Zmin switch.
I certainly have no intention to steal anybody's work) But still I'd like to have Toolbar settings preserved over reboots, while now it's not. I can't even get, how the items are assigned, to hard code them. I found some item numbers, but changing then gives unpredictable effect.

[classicrocker883]

so by going into Control settings, under setup Toolbar, and changing any of the toolbar buttons to something else doesn't actually save after a restart? I am unable to reproduce this. It is working normally for me, that is I am able to change any of the toolbar buttons to something else and it remains after restart.

remember, that changing the Author in Configuration.h is not recommended. leave this as it originally was #define STRING_CONFIG_H_AUTHOR "Miguel A. Risco-Castillo (MRiscoC)"

in file Marlin/src/lcd/e3v2/proui/dwin_defines.h
under line 94

#if HAS_TOOLBAR
#define TBMaxOpt 5 // Amount of shortcuts on screen
#if HAS_BED_PROBE
#define DEF_TBOPT {9, 8, 0, 1, 4} // Default shorcuts for ALB/UBL
#else
#define DEF_TBOPT {7, 0, 6, 2, 8}; // Default shortcuts for MM
#endif
#endif

changing the numbers in the DEF_TBOPT array will hard code the toolbar on the main menu.

the number represent the order as listed in file toolbar_def.h

[oloendithas]

Working fine for you? That's weird. I wonder, what am I doing wrong.
I tried to tinker with the DEF_TBOPT and it changes the toolbar buttons unpredictably.

The good news! I've finally fixed the probe servo pin. Edited timers.h to have MF_TIMER_SERVO0 = 3, rather than 1.
I don't understand why this works for Ender3 V2, since STM32F103R has the same mapping here. By default PB0 is bound to Timer3 and in board.cpp it's on &timer3 {&gpiob, &timer3, &adc1, 0, 3, 8}, /* PB0 */

Why in timers.h it gets #define MF_TIMER_SERVO0 1 // SERVO0 or BLTOUCH and how this suppose to work, I don't get. I can see in datasheet, that it can be redefined to TIMER1, but where and when this should happen?
Anyway, deploy and stow now are working.
The new battle is to make Z stop when the probe's shaft is pushed in. By now it doesn't stop.

Author: oloendithas <oloendithas@gmail.com>  2023-01-07 12:53:51
Committer: oloendithas <oloendithas@gmail.com>  2023-01-07 12:53:51
Parent: 1ac9c445e2713a2725910c3889a3fd343205c0e6 (current MM config)
Branch: master
Follows: 2.1.3
Precedes: 

    fix probe servo pin for Voxelab N32

----------------------- Marlin/src/HAL/STM32F1/timers.h -----------------------
index 0cd807f..8a23186 100644
@@ -80,7 +80,7 @@ typedef uint16_t hal_timer_t;
   //#define MF_TIMER_TEMP       4  // 2->4, Timer 2 for Stepper Current PWM
 #endif
 
-#if MB(BTT_SKR_MINI_E3_V1_0, BTT_SKR_E3_DIP, BTT_SKR_MINI_E3_V1_2, MKS_ROBIN_LITE, MKS_ROBIN_E3D, MKS_ROBIN_E3)
+#if MB(BTT_SKR_MINI_E3_V1_0, BTT_SKR_E3_DIP, BTT_SKR_MINI_E3_V1_2, MKS_ROBIN_LITE, MKS_ROBIN_E3D, MKS_ROBIN_E3, VOXELAB_AQUILA)
   // SKR Mini E3 boards use PA8 as FAN_PIN, so TIMER 1 is used for Fan PWM.
   #ifdef STM32_HIGH_DENSITY
     #define MF_TIMER_SERVO0  8  // tone.cpp uses Timer 4