Web interface for building binaries

[terjeio]

I have started tinkering with a Web intreface, focus has been on generating the needed compiler symbols, not the UI layout.
A preview is available here, currently symbols are output to the browser console only.

I have set up a development machine with the latest Ubuntu release which has platformio and docker installed. Next steps will be to post the compiler symbols to this machine to get the binary generated. It has been a while since I worked with browser apps/backends so any input on how to handle this will be appreciated.

Some technical details:

The browser code is pure javascript, no framework used (yet?).
A "global" json file will be hosted in the core repository, this contains links to all the drivers and info about plugins etc.
Each driver repository linked to will have a json file containing information about the driver and the boards/controllers supported.
The file format of the json files are work in progress, but the current iteration seems promising considering the complexity of options available.

Finally: a ton of testing is required to to get this off the ground and performing smoothly. Help is needed as I do not own all the supported boards and thus I am not able to verify binaries for these.

[dresco]

Afraid I don't have any input on the web backend, but is a nice idea. Happy to test whatever I can..

[terjeio]

i have a question, ...

@customsolutions Either ask it here or start a new discussion.

[customsolutions]

is there a way to add this board to webbuilder? "https://github.com/paulusjacobus/SuperGerbil/blob/master/inc/cpu_map.h" I currently have that board in my K40, and have a nucleo on the way.... I would like to test what we discussed in the other thread, however $35 doesnt show up in his firmwear version or perhaps an option to use a custom boardfile.h?

[terjeio]

@customsolutions Which processor variant? I know it is a STM32F103, but you need a variant with at least 128K flash for grblHAL. If enough flash a board map compatible with grblHAL has to be made for it.

[terjeio]

@customsolutions Which board? The original 5 axis SuperGerbil, or one of the two axis MG boards? Or can the same map be used for all?

[customsolutions]

I used the mg2, and swapped in an stm32f103b. Then compiled with edited board file in stmcubeide.

[GrblGru]

This is a great idea ! :)
There are a lot of possible users of GrblHAL who are not interested in details of the controller software.
They just want to flash their controller with the latest working program in the easiest way possible.

It would be still helpful if to the individual settings tooltips were present, which indicate additional information.
For example, it is not clear to me which board I have to select if I do not use a board.

I have selected for the bin file of my Arduino DUE under driver "SAM3XE(Arduinoe Due)" and "as board "Ramps SMART".
All other settings I took over unchanged.

Then I successfully flashed the bin-file with bossac.

bossac.exe --port=COM6 -U false -e -w -v -b FileToBeFlashed.bin -R

Erase flash
Write 140224 bytes to flash
[==============================] 100% (548/548 pages)
Verify 140224 bytes of flash
[==============================] 100% (548/548 pages)
Verify successful
Set boot flash true
CPU reset.

However, after connecting a terminal, I do not get any feedback from the DUE. :(

[terjeio]

It would be still helpful if to the individual settings tooltips were present, which indicate additional information.

On my todo list, it is still early days - I am a bit surprised I have managed to proceed this far already.

For example, it is not clear to me which board I have to select if I do not use a board.

You will have to check the pin mappings for the board selected - perhaps I should add a link directly to the map file?

All other settings I took over unchanged.
However, after connecting a terminal, I do not get any feedback from the DUE. :(

Connection: USB is for native USB, Connection: (Virtual) UART is for the programming port. So perhaps you tried the wrong physical port?

[GrblGru]

I have created a bin file from the SAM3X8E driver with the Arduino IDE and flashed it with bossac into the DUE.
I have used the Programming USB. If I connect a terminal here or also my program the data transfer works.
I have never used the native USB. What am I doing wrong ?

[terjeio]

If you want to the sender to communicate with the controller via the Programming port you have to select (Virtual) UART for connection, if you select USB the you have to connect physically via the Native USB port.

This is for using the programming port:

For the Due I am unsure which one to present as default, the programming port is the slowest, USB will increase CPU load - not that that may matter much.

I am a bit unsure how to label this - a USB connector is USB for the user whatever is behind it. The DUE is a bit special as it has two USB connectors, only a couple of other board in my collection has that.

[GrblGru]

Yep, that works :)
Congratulations ! This is a super easy way to flash the controller.

Unfortunately with this my other error (soft reset) is still there :( But that has nothing to do with this.
Your "binary builder" is really a great tool !!!! Thanks a lot.

I have another STM32, which I try to 'make happy' with GrblHAL. I will report to you.

[GrblGru]

The good news for you:
The bin file for STM32F1xx also works.
I could make the "Blue pill" happy with 'Flash Loader'. :)

The bad news for me:
The error with the soft reset is also present here :(

[dresco]

Hi Terje, have a couple more questions - sorry ;)

For the H7, how would I disable the Ethernet section for boards that don't support it? Am assuming it needs to be present in the root caps/plugins to show up for the ones that do?

For Trinamic support, I added the motors plugin in the root caps, but is still disabled for all boards - so I must be missing something extra?

• The BTT board uses soft UART and soft SPI.
• The WeAct board supports soft UART, plus both hard & soft SPI (selected with a TRINAMIC_SOFT_SPI symbol).. Is it possible to cater for that at build time? (Appreciate I only included both as it's more of a development platform, the BTT board defines TRINAMIC_SOFT_SPI in the map file - which I think would be more usual).

As a future consideration, it would be nice to be able to verify the output from the board json and build template without having to push to master to test it (just to save a messy commit history).. Don't know if you are planning to release your builder code, or some other way to validate the configuration offline?

Thanks!

[terjeio]

As a future consideration, it would be nice to be able to verify the output from the board json and build template without having to push to master

Agree, but a perhaps bit tricky.
The driver json file is fetched from the link specified in the drivers json file that I currently have on my server but later will be fetched from the drivers repo. If you can host the driver json file somewhere I can temporarily change my drivers.json to fetch from there.
Some other possible alternatives:

• It may be possible to modifiy the html and javascript to use the file: protocol so that you can have the files locally but use my web server for building? My web server only need access to the .tpl file.

• If you have a linux machine then set up a complete environment (apache web server, nodejs and platformio) and run everything locally.

• I open up my server so that you can ftp the driver.json to it.

Don't know if you are planning to release your builder code, or some other way to validate the configuration offline?

All but the backend code is already available, the backend is only useful if setting up a local server.

[dresco]

If you have a linux machine then set up a complete environment (apache web server, nodejs and platformio) and run everything locally.

Setting up a local server for testing would work for me.. 👍

[terjeio]

You have one ready? Apache, nodejs and platformio installed?

I can prepare a zip for download with the needed bits - I have to recap the configuration first.

[dresco]

You have one ready? Apache, nodejs and platformio installed?

Yep, my machines here are all Linux.. Just thinking aloud, I could maybe have a crack at containerizing it also?

[terjeio]

Giving you access to my master repo (subversion) is the best option then?

Just thinking aloud, I could maybe have a crack at containerizing it also?

I am running the builder on a dedicated development machine, later on I may move it to a hosted server instance I subscribe to.
A container could be nice, more so if other driver developers comes along and want to verify their builds?

[dresco]

Giving you access to my master repo (subversion) is the best option then?

Or even just a zip to get started with, will make sure I can get it all set up & working on a clean image?

[MeKeCNC]

It gives compilation error when sdcard is active for F1xx driver

[terjeio]

Thanks for feedback. You may try again now.

[MeKeCNC]

OK, compile error fixed, thanks

[GrblGru]

I just used the binary builder again to flash a bin file for the DUE.
Can it be that in the current version is again the known error with the soft reset ?

[terjeio]

A build server update that has gone wrong, you can try again now.

[GrblGru]

Thanks, everything works fine again. :)

[rschell]

Nice work, used it on my Teensy4 with a T41U5XBB board. Works like a charm.

Feature request: save selected options between sessions.

[terjeio]

Thanks. Saving/loading of options is on my todo list. Saving is easy, loading likely not so...

[terjeio]

New version up, with save and load of board features. Only ligthly tested, please report any misbehaviour.
I think I will add the option of bundling the file with the generated firmware in a zip as well.

[NeoKurtex]

This is really helpful, thanks!

Unfortunately, I am getting no response after uploading to a skr mini e3 v2.0 via sd card.

I’m using the default configuration on the web interface to generate the firmware.bin file. The flashing operation seems to work and the board is recognized as a usb device, there is just no activity after making a serial connection.

[terjeio]

Download and try the binary I have made with Eclipse. Working?

[terjeio]

Edit: I found a windows device and tried putty. I get 'error:7'

I did not see this edit, error 7 means that settings where reset. Any response from ? or <CTRL>+X in puTTY?

[NeoKurtex]

No response from puTTY, even with the binary you provided. It works with Marlin flashed back onto the board, though.

I have three identical boards here, none of them like the binary. Since this seems like an issue with the configuration of my boards, rather than web interface, would it be best to move this discussion to STM32F1xx?

[terjeio]

Can you try a new build now - I believe I have found a bug. I do not have a board to test with so cannot verify.

[NeoKurtex]

Yes, that fixed it! Thanks so much!

[JeffHank1]

Thanks for all the time and effort you've invested thus far.

I have successfully programmed my STM32F401CCU6 with a .bin from the syn.io-engineering site.

I used STM32F4XX driver and the Blackpill STM32F401 selections. I specified 4 axis and laser. I actually want to use 5 axes so I tried the STM32F4XX driver with CNC3040 5 axes, Laser module, and Backlash compensation. This failed to build sadly. I don't know if this is just an untested combination. The cnc3040_map.h looks like what I want. The failure report starts with this and goes on for quite a long time:

Command failed: pio run -etest -cjob.ini
In file included from /home/terjeio/.platformio/packages/framework-stm32cubef4/Drivers/STM32F4xx_HAL_Driver/Inc/stm32f4xx_hal_rcc.h:33:0,
from /home/terjeio/.platformio/packages/framework-stm32cubef4/Drivers/STM32F4xx_HAL_Driver/Inc/stm32f4xx_hal_conf.h:281,
from /home/terjeio/.platformio/packages/framework-stm32cubef4/Drivers/STM32F4xx_HAL_Driver/Inc/stm32f4xx_hal.h:30,
from /home/terjeio/.platformio/packages/framework-stm32cubef4/Drivers/STM32F4xx_HAL_Driver/Src/stm32f4xx_hal.c:36:
/home/terjeio/.platformio/packages/framework-stm32cubef4/Drivers/STM32F4xx_HAL_Driver/Inc/stm32f4xx_hal_rcc_ex.h:400:2: error: conflicting types for 'RCC_PLLI2SInitTypeDef'
}RCC_PLLI2SInitTypeDef;
^~~~~~~~~~~~~~~~~~~~~
/home/terjeio/.platformio/packages/framework-stm32cubef4/Drivers/STM32F4xx_HAL_Driver/Inc/stm32f4xx_hal_rcc_ex.h:99:2: note: previous declaration of 'RCC_PLLI2SInitTypeDef' was here
}RCC_PLLI2SInitTypeDef;
^~~~~~~~~~~~~~~~~~~~~
/home/terjeio/.platformio/packages/framework-stm32cubef4/Drivers/STM32F4xx_HAL_Driver/Inc/stm32f4xx_hal_rcc_ex.h:419:2: error: conflicting types for 'RCC_PeriphCLKInitTypeDef'
}RCC_PeriphCLKInitTypeDef;
^~~~~~~~~~~~~~~~~~~~~~~~

The list ended with:
*** [.pio/build/test/FrameworkHALDriver/Src/stm32f4xx_hal.o] Error 1
*** [.pio/build/test/FrameworkHALDriver/Src/stm32f4xx_hal_adc.o] Error 1
========================== [FAILED] Took 2.80 seconds ==========================

Any suggestions? Is this just an impossible combination? Thanks for any help.

[terjeio]

Try again now - bad declarations for the board.
Testing all combinations will take a long time, and I do not have all the board variants, so I leave it to the community to help out and provide feedback - so thanks for yours.

[JeffHank1]

I used a USB to TTL serial adaptor wiring to pins 9 and 10 as well as 3.3V + Gnd. The ST Flash Loader Demonstrator worked perfectly it seems.

[JeffHank1]

I've since Hooked up the above Black Pill to a TB6600 stepper driver and Nema23 motor. XYZA all worked properly. I had to wire in a 3.3 to 5v bidirectional shifter to get enough signal to get the stepper driver working. Otherwise no problems.
I didn't notice the fix you reported above so I will test that out too.

[JeffHank1]

The Black Pill I've mentioned was reprogrammed with STM32F4XX driver with CNC3040 5 axes, Laser module, and Backlash compensation.
The bin assembled perfectly. I've since also tested it physically and all 5 axes work. I was not able to test Probe, Backlash, or Laser unfortunately. Grblgru has to catch up ha ha
Thanks so much for the prompt fix.

[JeffHank1]

I have also tried an Arduino Due board (SAM3x8E) driver with RAMPS 1.6 board selected. 5 axes and Backlash as options selected. The .bin file assembled properly. I used Bossa with the Windows GUI to upload it via the Programming port (next to the power connector). I then switched to the Native port and Device Manager reported this as AT91 USB to Serial Converter.

I selected that port to try with Grblgru and GrblPanel.

Grblgru reports "Missing GrblHAL Version Message @ COM18"

Grblpanel reports "System.ArgumentOutOfRangeException: StartIndex cannot be less than zero.
Parameter name: startIndex
at System.String.Remove(Int32 startIndex, Int32 count)
at GrblPanel.GrblGui.showGrblPositions(String data)
at GrblPanel.GrblIF.raiseAppSerialDataEvent(String data)
at GrblPanel.GrblIF.VB$StateMachine_40__client_ComReadData.MoveNext()
--- End of stack trace from previous location where exception was thrown ---
at System.Runtime.ExceptionServices.ExceptionDispatchInfo.Throw()"

I don't know if this problem is a missing line within the assembled .bin file, or if this is all from the software I am using.
I'll collaborate with Grblgru to try to resolve possible things in his software. No idea about GrblPpanel.

If you think this is something with the bin please let me know and I'll try it again.
Thanks again for setting this all up.

[JeffHank1]

HI again. I finally got ioSender to work. My antivirus programs really don't like it :(

I tested it initially on my working Black Pill as above and got a suitable status report in Console.

I attached it to the Arduino Due described above and ioSender reports it doesn't get any response from the controller so it exits. I can see the AT19 USB to Serial Converter in Device Manager but it won't connect. I tried rebuilding the Due + RAMPS 1.6 bin. It seemed to write and verify successfully via the Programming port, but still wouldn't connect to the Native (or Programming) port in ioSender.
I tried a bin build with Due plus Mega 2560 but that wouldn't build = maybe an impossible combination ;)

[terjeio]

I have just fixed the Mega2560 board - PWM was not implemented. I have generated and flashed with both Native USB and Programming Port variants, and can connect with ioSender. Could it be that you are not flashing with correct settings in Bossa? FYI the Native USB port is listed as Arduino Due in the device manager on my machine.

[JeffHank1]

Thank you for the suggestion. I wasn't checking Boot to Flash since I didn't know what that meant. I was able to connect with the Due via the Native port successfully and get the appropriate Grbl bootup message. It wouldn't connect via the Programming port but I don't think Due is supposed to from what I've read. This was the Due + RAMPS 1.6 bin combination.

Grblgru still doesn't see the Version message so I'll discuss that with him.

I'm wiring up a 'board to logic level voltage converter to stepper driver to motor' setup to test both the Black Pill and the Due in real life with 5 axes. If there are any problems I will report back. All the help is greatly appreciated.

[terjeio]

It wouldn't connect via the Programming port but I don't think Due is supposed to from what I've read.

I does for me, did you change the Connection setting to (Virtual) UART in the Web Builder for that build?

Grblgru still doesn't see the Version message so I'll discuss that with him.

This could be due to Native USB port not sending the welcome message on (re)connect, this is a known problem that has no easy workaround at the controller side.

[JeffHank1]

Re: connecting via the programming port: I didn't try changing the connection setting to UART - I don't know enough about it all. From what I've read some years ago I am supposed to communicate via the Native port for regular work. You mentioned above it is supposed to be faster anyway I think. I just need one port that works.

I tried a regular USB connection with ioSender via the Programming port and didn't get the Welcome message. I do get it via the Native port though. I'm fine with that pending a discussion with Grblgru.

Meanwhile I have a Back Pill to play with. I have noticed I can change some of the Grbl settings using a macro within Grblgru - the motor parameter related stuff - but changes to $10 for example won't persist after reboot. I will be trying to change a few settings via ioSender since you've designed that for GrblHal and these types of processors.

[MeKeCNC]

Hi @terjeio,
web-based compilation is a great idea. It will be a good leveling up for the grblHAL platform that the end user can compile for the board he will use, without having to deal with the source code. This will generate more interest in the grblHAL platform.
In this context, I have a suggestion. If a yaml file is created for each board, changes can be made and this file can be uploaded on the web page during compilation and it will be more personalized.
For example, the yaml file given in the link below
https://github.com/bdring/FluidNC/blob/main/example_configs/3axis_v4.yaml

and this is my custom yaml file for ESP32
config.txt

[terjeio]

If a yaml file is created for each board, changes can be made and this file can be uploaded on the web page during compilation and it will be more personalized.

Supporting this will be a major undertaking - way beyond what I am able to do single-handedly. For the ESP32 it is fairly simple as it has a small number of pins and most can be routed to the peripheral of your choice. This is not so for many of the MCUs grblHAL supports, eg. the STM32F756 MCU on the Nucleo board I have has nearly 100 pins and freedom to chose routing of peripheral signals is limited. To let end users chose which pins to assign to what functionality would require extensive documentation, sensible feedback from failed compilation, and likely a lot of support...

[MeKeCNC]

I understand you, you are right. I am aware that a team is needed for this type of initiative to be successful, maybe there will be volunteers in the future.

[JeffHank1]

Changing GRBL parameters. I have uploaded my usual grbl parameters for steppers motors, etc. The parameter for $10=1 seems to be pretty stubborn about not changing. Can you tell me which line in the App.config file saves this value. I prefer $10=0 to work in WPos rather than MPos. Thanks

[JeffHank1]

I have hooked my STM32 Black Pill up to a real machine and started moving stuff. I entered the Grbl settings I use for a Mega5X which I know to be correct for 3 linear and 2 rotary axes.
Results: X and Y move the correct distances, Z moves normally. The Rotary axes don't seem to move in degrees but as linear axes. As a result they barely move. This was tested within ioSender.
I rechecked by hooking everything up to my Mega5X and movements are normal.
I have edited the Grbl Config settings within ioSender but despite being called Rotary axes, A and B seem to be moving linearly.
It is not clear if these are uploaded to the STM32 though. I save the changes and back them up.

[terjeio]

The parameter for $10=1 seems to be pretty stubborn about not changing.

Which driver? When testing with STM32F4xx there is no problem with this. Be aware that $10 has many more options (bit flags) than legacy Grbl - you do not normally want to set $10 to 0 or 1. Use $$=10 to list options or ioSender to visualize options as checkboxes (in Settings: Grbl > General group).

Can you tell me which line in the App.config file saves this value.

App.config has no options for saving controller settings.

The Rotary axes don't seem to move in degrees but as linear axes.

Is this due to the setting unit still beeing step/mm in ioSender? It should be step/deg - I may change this later if/when I am able to come up with a generic way to handle settings interdependencies. When you define an axis as rotary the only change is that no conversion (to mm) is done when in inches mode (M20).

[JeffHank1]

Thanks for the information.

Re: Grbl settings. I have tried to set $10=5 with ioSender on the MDI. Alas $10 prefers to stay at $10=1. 5 - Work coordinate offset (32) I have also tried just checking Machine Coordinates in the General selection window for Grbl but no change either. I am only familiar with setting/changing legacy Grbl so perhaps I need to do some more reading about GrblHal.

Re: Rotary axes. They are set to step/mm just like the linear axes from what I see in ioSender but there doesn't seem to be any way to change that yet as you have indicated.

[terjeio]

Re: Rotary axes.

Just ignore the unit in the UI for rotary axes and enter steps/deg.

[distebia]

salve, se genero il firmware per esp32 e mks dlc32 non riesco a caricarlo nella scheda mentre se ne carico una copia generata a settembre scorso tutto fila liscio, nessun problema. La procedura che utilizzo è sempre la stessa. non so da cosa dipende ma non riesco a caricare il firmware generato oggi

[terjeio]

Hi, if I generate the firmware for ESP32 and MKS DLC32 I can't load it into the board while I load a copy generated last September everything goes smoothly, no problem. The procedure I use is always the same. I don't know what it's because but I can't load the firmware generated today

[terjeio]

Duplicate of this issue.

[crackinx]

using a BTT SKR pro V1.2 with external drivers, it seems it is wired weird where the normal endstop pins don't work with external drivers. Is there any way you can add an option for this where it will use different pins (maybe PF4, PF5, and PF6) for the endstops when external drivers are used??

[terjeio]

I do not have a board to test with, but I see no reason for why they should not work. Could it be that your switches are normally open and you have to invert the inputs?

[crackinx]

the board has lights for each endstop, the lights function normally but do not register as being triggered in software, Kind of a stab in the dark, but the external drivers are the only thing I can think of that would cause this to happen. My endstop switches do all work and are not inverted. I don't see anything in the board schematic that would cause this loss of function either. I guess I'll do some more testing. I would guess bad board, but it's all the endstops and only the endstops.

[terjeio]

I have verified the inputs (XYZ) with a Nucleo-144 board.
Using external drivers should not affect the inputs unless you still have Trinamic drivers mounted and they are interfering with the limit inputs (via the Diag line). If not either your controller is defective or the schematics is incorrect.

[crackinx]

Got it all figured out!! It was my Uart to RS485 bridge, I had it on the wrong rx/tx output (installed on rx3/tx3 instead of rx1/tx1) which caused the signal issue with the endstops for some reason. Thanks very much for the assist!! It's awesome to have an active developer willing to help people through issues! Keep up the great work!!

[crackinx]

I'm also having an issue trying to figure out what uart pins are used for VFD control, I see
#define SERIAL_PORT 1 // GPIOA: TX = 9, RX = 10 #define SERIAL1_PORT 32 // GPIOD: TX = 8, RX = 9 #define SERIAL2_PORT 6 // GPIOC: TX = 6, RX = 7
in the board file, but can not find anywhere which says which is used. Any help would be much appreciated!

[terjeio]

Use the $pins command to see which.

[Smolinc]

Hi I'm using BTT skr3 with H723. Probe pin checkbox enabled. But after flashing Probe not work.

[VER:1.1f.20240604:]
[OPT:VNMSL2,100,1024,3,0]
[AXS:3:XYZ]
[NEWOPT:ENUMS,RT+,HOME,NOPROBE,ES,SED]
[FIRMWARE:grblHAL]
[SIGNALS:HSE]
[NVS STORAGE:*FLASH]
[FREE MEMORY:212K]
[DRIVER:STM32H723]
[DRIVER VERSION:240522]
[BOARD:https://github.com/bigtreetech/SKR-3]
[AUX IO:1,0,0,0]
[PLUGIN:MODBUS v0.16]

[dresco]

Hi I'm using BTT skr3 with H723. Probe pin checkbox enabled. But after flashing Probe not work.

Hi, will take a look. Has also been mentioned in the H7 repo here...

[dresco]

@Smolinc
fyi - have just pushed a fix for the probe issue to the H7 repo.

[Smolinc]

Cool but I can try it only in next month. I let you know as soon as I try.

[SpiderSpartan]

I'm trying to generate firmware for the STM32Fxx, BTT SKR-2 STM32F407, but keep getting an error Build failed: 422. Here are the changes I made to default configs:

• Y-Axis: Auto-squared
• Disable Probe Input
• Trinamic drivers: TMC2209
• Enable SD Card

Command failed: pio run -etest -cjob.ini
Src/diskio.c:554:1: error: unknown type name 'DRESULT'; did you mean 'FRESULT'?
DRESULT disk_ioctl (
^~~~~~~
FRESULT
Src/diskio.c: In function 'disk_ioctl':
Src/diskio.c:560:5: error: unknown type name 'DRESULT'; did you mean 'FRESULT'?
DRESULT res;
^~~~~~~
FRESULT
Src/diskio.c:565:21: error: 'RES_PARERR' undeclared (first use in this function); did you mean 'AES_IT_ERR'?
if (drv) return RES_PARERR;
^~~~~~~~~~
AES_IT_ERR
Src/diskio.c:565:21: note: each undeclared identifier is reported only once for each function it appears in
Src/diskio.c:567:11: error: 'RES_ERROR' undeclared (first use in this function); did you mean 'URB_ERROR'?
res = RES_ERROR;
^~~~~~~~~
URB_ERROR
Src/diskio.c:569:17: error: 'CTRL_POWER' undeclared (first use in this function)
if (ctrl == CTRL_POWER) {
^~~~~~~~~~
Src/diskio.c:572:17: warning: implicit declaration of function 'chk_power' [-Wimplicit-function-declaration]
if (chk_power())
^~~~~~~~~
Src/diskio.c:573:17: warning: implicit declaration of function 'power_off' [-Wimplicit-function-declaration]
power_off(); /* Power off */
^~~~~~~~~
Src/diskio.c:574:19: error: 'RES_OK' undeclared (first use in this function); did you mean 'FR_OK'?
res = RES_OK;
^~~~~~
FR_OK
Src/diskio.c:577:13: warning: implicit declaration of function 'power_on' [-Wimplicit-function-declaration]
power_on(); /* Power on */
^~~~~~~~
Src/diskio.c:589:13: error: 'Stat' undeclared (first use in this function)
if (Stat & STA_NOINIT) return RES_NOTRDY;
^~~~
Src/diskio.c:589:20: error: 'STA_NOINIT' undeclared (first use in this function)
if (Stat & STA_NOINIT) return RES_NOTRDY;
^~~~~~~~~~
Src/diskio.c:589:39: error: 'RES_NOTRDY' undeclared (first use in this function); did you mean 'RES_OK'?
if (Stat & STA_NOINIT) return RES_NOTRDY;
^~~~~~~~~~
RES_OK
Src/diskio.c:591:9: warning: implicit declaration of function 'SELECT' [-Wimplicit-function-declaration]
SELECT(); /* CS = L */
^~~~~~
Src/diskio.c:595:14: error: 'GET_SECTOR_COUNT' undeclared (first use in this function); did you mean 'ATA_SECTOR_COUNT'?
case GET_SECTOR_COUNT : /* Get number of sectors on the disk (DWORD) */
^~~~~~~~~~~~~~~~
ATA_SECTOR_COUNT
Src/diskio.c:596:18: warning: implicit declaration of function 'send_cmd' [-Wimplicit-function-declaration]
if ((send_cmd(CMD9, 0) == 0) && rcvr_datablock(csd, 16)) {
^~~~~~~~
Src/diskio.c:596:27: error: 'CMD9' undeclared (first use in this function)
if ((send_cmd(CMD9, 0) == 0) && rcvr_datablock(csd, 16)) {
^~~~
Src/diskio.c:596:45: warning: implicit declaration of function 'rcvr_datablock' [-Wimplicit-function-declaration]
if ((send_cmd(CMD9, 0) == 0) && rcvr_datablock(csd, 16)) {
^~~~~~~~~~~~~~
Src/diskio.c:609:14: error: 'GET_SECTOR_SIZE' undeclared (first use in this function); did you mean 'CF_SECTOR_SIZE'?
case GET_SECTOR_SIZE : /* Get sectors on the disk (WORD) */
^~~~~~~~~~~~~~~
CF_SECTOR_SIZE
Src/diskio.c:614:14: error: 'CTRL_SYNC' undeclared (first use in this function); did you mean 'CARD_SDSC'?
case CTRL_SYNC : /* Make sure that data has been written */
^~~~~~~~~
CARD_SDSC
Src/diskio.c:615:17: warning: implicit declaration of function 'wait_ready'; did you mean 'f_read'? [-Wimplicit-function-declaration]
if (wait_ready() == 0xFF)
^~~~~~~~~~
f_read
Src/diskio.c:619:14: error: 'MMC_GET_CSD' undeclared (first use in this function); did you mean 'RCC_IT_CSS'?
case MMC_GET_CSD : /* Receive CSD as a data block (16 bytes) */
^~~~~~~~~~~
RCC_IT_CSS
Src/diskio.c:625:14: error: 'MMC_GET_CID' undeclared (first use in this function); did you mean 'MMC_GET_CSD'?
case MMC_GET_CID : /* Receive CID as a data block (16 bytes) */
^~~~~~~~~~~
MMC_GET_CSD
Src/diskio.c:626:26: error: 'CMD10' undeclared (first use in this function); did you mean 'CMD9'?
if (send_cmd(CMD10, 0) == 0 /* READ_CID */
^~~~~
CMD9
Src/diskio.c:631:14: error: 'MMC_GET_OCR' undeclared (first use in this function); did you mean 'MMC_GET_CID'?
case MMC_GET_OCR : /* Receive OCR as an R3 resp (4 bytes) */
^~~~~~~~~~~
MMC_GET_CID
Src/diskio.c:632:26: error: 'CMD58' undeclared (first use in this function); did you mean 'CMD10'?
if (send_cmd(CMD58, 0) == 0) { /* READ_OCR */
^~~~~
CMD10
Src/diskio.c:634:30: warning: implicit declaration of function 'rcvr_spi' [-Wimplicit-function-declaration]
*ptr++ = rcvr_spi();
^~~~~~~~
Src/diskio.c:647:9: warning: implicit declaration of function 'DESELECT' [-Wimplicit-function-declaration]
DESELECT(); /* CS = H */
^~~~~~~~
*** [.pio/build/test/src/diskio.o] Error 1
========================= [FAILED] Took 26.91 seconds =========================

[terjeio]

Fixed now, try again.

[SpiderSpartan]

Thanks! It's working now. Only other thought I had was , it would be nice if I could download the config files this generates so I could make further tweaks, ie: change pins for the laser.

[jaymis]

We're having a little project on the PrintNC discord server (invite link if anyone wants to join) at the moment to try to get some old GRBLHAL2000 boards up to date with Probe Protection and other more recent grblHAL innovations.

Terje, I've had a bit of a search around and couldn't find anything mentioned: Is it possible to use a customised board map file on the Web Builder?

The reason we're asking is that the GRBLHAL2000 board doesn't have any spare AUX input pins, but not many people are using a B axis, so we could use the B limit input as a probe protect AUX pin.

[tcurdt]

It would be super nice to also have a board map for the FlexiHAL board. ❤️

[jaymis]

I believe the FlexiHAL isn't in the web builder because there's is some custom code/plugins (I forget what precisely) that aren't currently compatible with the main branch of grblHAL.

[terjeio]

Is it possible to use a customised board map file on the Web Builder?

No, and I do not plan to add such a capability.
The current board map could be modified, or an additional board map could be added though.

[jaymis]

Thanks Terje. Obviously not knowing how the web builder works, I didn't know if adding custom board maps would be a big and complex task.

An additional board map I think could help quite a lot. I don't know how many people are using 4-5 axis machines, so repurposing the GRBLHAL2000 B limit input as an AUX input I think could help many people get their existing systems a bit more flexible.

I'll talk it through on the PrintNC server and get back to you with a customised map. Thanks.

[p2mm]

and now the web builder only makes unbootable hexes

[Christo-Fourie]

Good day

Not sure if this is the correct place to post...

I'm generating the firmware for the MKS DLC32 V2.0 board and setting spindle 1 to on/off (see screenshot below):

The firmware compiles and downloads without a problem. I can also flash it using the ESP Flash Download Tool.

My problem comes in with the spindle. Having set it to on/off, I expect either 12V or 0V over the "Spindle + -" pins on the board depending on whether the spindle is on or off. The "Spindle + -" pins are connected to an SRD-12VDC-SL-C relay that switches the spindle's power supply. Board to Relay connections are as follows:

Spindle + <---> V+
Spindle - <---> IN
GND <---> V-

The "Spindle + -" output on the board is not doing anything, so no spindle for me at this time...

Regards
Christo

[stev-e]

Hello Christo. I am impressed that you managed to get it to generate firmware for an ESP32. I am trying to replace the MKS firmware in a Twotrees TTC-450 (would that the source code had not been deleted from the MKS DLC32 ~ ESP32 github pages, if anyone has these please shout). I want a version that can run files on an SD card, has a z probe and only needs spindle on/off but I can not get even the plainest web-builder configuration to output anything. I do note the warning that ESP32 takes a while but I keep getting "Failed : 0" .
Perhaps you can tell me, will GRBLhal run stand-alone (using the routers small screen) ?
If your selected options coincide with mine, or you can generate my preferred variant, could you fling me the binary?
Thank you

[terjeio]

I do note the warning that ESP32 takes a while but I keep getting "Failed : 0" .

My dev board fails to boot if GPIO12 has a pullup resistor mounted which my SD card breakout has, the DLC32 has a pulldown so it may still boot... If you flash an image without the SD card enabled does that work?

Perhaps you can tell me, will GRBLhal run stand-alone (using the routers small screen) ?

No, a LCD is not supported.

[terjeio]

My problem comes in with the spindle. Having set it to on/off, I expect either 12V or 0V over the "Spindle + -" pins on the board

The pin (GPIO27) supplies either 0V or ~3.3V) so you will have to add a driver (transistor) to drive the relay. And it works as it should for me when I test with a development board. What could be done is to switch the spindle on/off signal to GPIO32 instead as this has a driver transistor in place (it is used for PWM when a PWM spindle is enabled).

[Christo-Fourie]

So if I understand correctly, when the spindle is set to On/Off, GPIO27 is used for the On/Off signal? (In the firmware)

According to this literature (http://wiki.fluidnc.com/en/hardware/3rd-party/MKS_DLC32) GPIO27 is an LCD Reset (gpio.27 LCD_RST (On EXP1 via AHCT125 buffer. Output only)). This is from FluidNC so I'm not sure if it applies.

What could be done is to switch the spindle on/off signal to GPIO32 instead as this has a driver transistor in place (it is used for PWM when a PWM spindle is enabled).

What you suggest here is to modify the firmware to use GPIO32 instead of GPIO27? If this is the case, which file/s will have to be modified?

Thanks

[terjeio]

I have now changed the board map to use GPIO32 for spindle enable when On/off spindle is selected.

[Christo-Fourie]

And it works like a charm...

3 weeks of getting nowhere and now it finally works. Best feeling in the world.

Thank you very much.

[dangerousderek]

Hi!
I used the Web builder to compile for the BTT SKR2. I'm not good enough with code to compile grblhal myself, and no combination of settings on the web builder seems to work. The SKR 2 board works fine with marlin, but as soon as I flash the .bin from the web builder it fails to show up as a usb com port. After a bunch of googling, I found a blog that pointed me to this conversation:
https://community.platformio.org/t/difficulty-with-getting-usb-serial-usb-cdc-working/7501/9
They are discussing the usb not working with this chipset, but once again I don't know enough about coding to compile grblhal and fix the usb issue myself. Is there any way to fix this issue with the web builder?
Thanks!

[Witalij51]

Good afternoon. Add a button that will redirect to a file sharing service to share Gerber files. I think there will be a lot of people who want to share their Gerber files.

[phil-barrett]

I'm not sure that many of the boards listed have associated gerbers that have been shared. I would expect that if they are sharing gerbers, the homepage button directs you to a page that will tell you where to go.

[Witalij51]

Good afternoon. I installed your GrblHal firmware on the MKS DLS-32 board. The board contains a Wi-Fi module. The wireless network shows the presence of the GrblHal Wi-Fi module. Please tell me what is the password to connect to the GrblHal module.

[phil-barrett]

use iosender via USB to get to the grbl settings. you can set the wifi password there.

[Witalij51]

Thank you

[p2mm]

every time i try the web ui:

Command failed: pio run -etest -cjob.ini
src/driver.c: In function 'driver_setup':
src/driver.c:2666:23: error: 'LITTLEFS_MOUNT_DIR' undeclared (first use in this function)
fs_littlefs_mount(LITTLEFS_MOUNT_DIR, t4_littlefs_hal());
^
src/driver.c:2666:23: note: each undeclared identifier is reported only once for each function it appears in
*** [.pio/build/test/src/driver.c.o] Error 1
========================== [FAILED] Took 5.90 seconds ==========================

[terjeio]

Fixed now, thanks for reporting.

[p2mm]

Thanks, thde web builder does its thing now. Unfortunately, the teensy uploader now declares the donwloaded hex as unreadable.

[terjeio]

Odd, zip the hex ad add to a comment.

[p2mm]

these don't boot at all anymore
firmware-10.hex.zip

[terjeio]

Try with MPG disabled or with Real time command switchover:

The E5XMCS_T41 code crashes when MPG mode is switched by pin. I do not have a board to test with so will have to dig a bit to see if I can find out why.

[p2mm]

that does compile. But the reason i needed the aux input switchover is that realtime command switchover doesn't work. I can do ?, but that's about it.

[terjeio]

I found a bug - fixed now for the Web Builder. Will commit to github later.

[p2mm]

Ah, nice. Still no reaction after 0x8B on serial except for ? by the way.

[terjeio]

0x8B character code switchover is not available when aux pin switchover mode is enabled.
If it doesn't work in Real time command switchover mode for you then how do you send the character?

[p2mm]

firmwarex.hex.zip

I.ve been flashing almost all weekend and all of today hoping my mpg is gonna work, maybe iv overstressed the server a little:)