schildkroet / grbl-advanced Goto Github PK

Hello,
is there some simple way to use L6470 motor driver with GRBL advanced? I´m trying to investigate what place in code could be best for driving, but no any succes.

THX!

I noticed that your $140,$141 and $142 clash with the configuration numbers in https://github.com/gnea/grbl-LPC , is this going to be a problem further down the track for you?
Cheers
Mike

Hi Schilddkroet, I'am just running a cnc-mill (modified shapeoke) with an arduino-uno and grbl. I build also a cnc-foamcutter. I'am planning to go 32bit for both and found your thread. For that I have some questions. For mill: is it possible to have an output which signalize lasermode? reason: I would like to switch between 2 differnt powersources and pwm-voltage for laser and spindle by switching the mode. By the way, have I still to boot when switching the mode laser/ spindle? Another thing: i have 2 x-motors. For homing-purpose I would like to have a home-switch for each motor(automatic squaring the gantry) after homing thy should get the same informations.
For my foamcutter i need a 4th linear-axis. I found in one of the config-files to define the number of linear-axes. But I am not shure if that will work. If it would be possible, i would need a separat homing-switch input for that. I hope you can clarify my questions. Thanks

i will change the pwm for spindel(laser) from 1khz to 8khz. Where is the best place in the Software to do this?

BR
Hexxter

hi... can i use others STM32 board such as
STM32F103 [ARM Cortex M3] or The STM32F407 [ARM Cortex M4] ?

thanks ^^

Hi, I was wondering if this platform is capable of running a 3D Printer in the sense of pins and hardware. I would like to help add support if we can support up to 2 extruders at least, that being said it would need to support not only 2 extruders but uart and/or spi motor drivers too.
Some icing on the cake would to be able to also have SPI support for 2 MAX6675 chips.
Of course I would be able to write a huge amount of these drivers needed for this, but I was hoping you could provide insight to if this is the right platform for such things.

Hi, many compliments for your port of GRBL.
I've a question about Laser mode:
in an old issue I've read that at that time you had not yet ported the laser mode to STM32.
It is still true nowaday ?
Tank you

Gabriele

I have problem with the laser firing when its not supposed to, any idea what i need to change in the firmware?
Those thin diagonal lines, it seems to be random on each time i run the gcode.

Can i use the GRBL CNC shield V3.0 ? Is it pin compatible?

Hi,

I had try to your code using Atollic with stm32f407zgt and stm32f407igt,
(both china both), I got 2 different result. I Just testing the code without
connecting to any hardware - stepper, switch and etc.

a) 407ZGT it work -- luckily it.

b) 407IGT it work but it stopped with click the Unlock from the Candle2.0
- when the board start, on candle it got ERROR:Unknown or ERROR:9
- when candle send $X and received OK
- following candle send the spindle speed S10000 but not received OK
and stop here.

what could be the problem? 

BTW, I suggest when USART_NUM = 1, code under the stm32f4xx_it.c,
added this checking, it can work without any changes ... just like when
run it, wonder why the candle not showing the real time status and position.

a)
void USART1_IRQHandler(void)
{
if(USART_GetITStatus(USART1, USART_IT_RXNE) != RESET) {
/* Read one byte from the receive data register */
unsigned char c = (USART_ReceiveData(USART1) & 0xFF);

	//  Write character to buffer
if ( 1 == USART_NUM )         //  Added this checking
  ProcessReceive( c );
else
  FifoUsart_Insert(USART1_NUM, USART_DIR_RX, c);
}

b)
void ProcessReceive(char c)
{
...
...
}
else {
// Write character to buffer
if ( 1 == USART_NUM ) // Added this
FifoUsart_Insert(USART1_NUM, USART_DIR_RX, c); // Added this checking
else
FifoUsart_Insert(USART2_NUM, USART_DIR_RX, c);
}
}
}

TQ.

More of a question.. Would it me possible to run GRBL-Advanced on one of these?

https://emcraft.com/products/868

It's powered by an NXP i.mx8 which has an M4 Arm unit. if so would you be interested in receiving a unit to try?

Hi ich weis das gehört hier nicht hin... aber ich wollte dich mal fragen ob es eine Möglichkeit gibt eine Version zu bekommen die für einen STM32F104 wäre?

Wenn ich die Hex flash funktioniert das natürlich nicht mit der USB anbindung...

Danke! und Sorry das ich hier nerve ;)

Das grbl plugin in cambam erkennt nicht die grbl version bzw will sich nicht verbinden!
hast du eine lösung?

Hi
in the nucleo f446 the speed of each axis can be up to 500 khz?
thanks before

Hi,
I try to get GRBL-Advanced up and running on a NUCLEO STM32F446RE.
I followed the instructions in the Readme.md to build the software on a Linux system.
After making following adaptions to the Makefile:
-replaced the compile options -DSTM32F411xE -DSTM32F411RE by -DSTM32F446xx
-replaced -T../stm32f446re_flash.ld by -T../stm32f411re_flash.ld
I was able to "make all" and "make flash"
After flashing and rebooting the following appears on the serial port when pushing Enter( with minicom /dev/ttyACM0 115200 8N1):
-ok
-error:9
then the communication hangs...
Any suggestions?

This looks like an awesome project. It's fantastic that you've implemented backlash compensation. Such a needed feature for DIY / Hobby CNCs.

My question is: For standard Arduinos, there are several shields available that allow you to hook up stepper drivers to it. Like this one:

What's your setup with STM32? What do you recommend? I see that you have a Hardware section in your README, but I couldn't understand from there how do you hook up your stepper drivers to the main board.

Hello and happy new year,
The probe doesn't stop on contact (it was working with the arduino uno and a CNC shield)
I checked your code and it looks like the wiring pin should be the same.
Any advice ?
Thanks

Hello and thanks for this STM32F4 port.

I have a few Nucleo boards but none of them with F4. I tested the code on F4 Discovery and it works fine with a few wires :) https://twitter.com/hubmartin/status/1082408826874327041

My plan is to port your code to Nucleo with STM32F303. I will use STM HAL library. I'd like to ask if you would be interested in pulling my fork later to your branch or if I should not bother with compatibility and create a separate repo for it?

If you think this is a good idea, It would would be great if you give me some hints how to switch in the build between the platforms - if it is good idea to put #ifdefs to the code or choose the right "grbl HAL" in the Makefile.

Thanks
Martin

Can I use this code on my mega2560 with ramps 1.4 . I am not to good on coding side so please help me I am mechanical engineer & I made my own small milling machine and now I am facing backlash problem on my machine so I want to use backlash comp on my machine .

I got it to run on my nucleo board and configured it for laser engraving, thanks for sharing this!

Is there a way to turn on laser mode for this version? The settings menu doesn't provide an option for this, is it commented out or just not included?

When rastering (adjusting spindle speed/laser power after each block) programs run very slow, about 1 block per second. If the spindle speed is not being changed, such as for profiling, the controller has no problem cranking through the moves, but when it comes to turn the laser off and then back on again to get from one profile to the next, it has the same 1 second or so pause. Any ideas how to fix this?

It seems it will only connect at 115200 baud. This probably won't be a limitation, but for grayscale rastering it can push the limits of the bandwidth, I am running my STM32F103 at 230400 baud and the buffer occasionally does dip below half.

Thanks in advance for your consideration and help.

I am having regular problems with connecting to GRBL. After I deploy the binary, everything is fine and I can use GRBL. But after turning it off and then on - I cannot connect. There is no response on Serial port from the device.

I really doubt that this is a bug but no clue how to track this. Did anyone encountered such problem?

Hello, I'm a beginner when it comes to compilers etc and need your help. I managed to get Embitz running on my mac via virtual box and did a test build of your port without any problems (an unbelievable first for me) how do I change the build settings to target a F401RE board as I have a couple of these lying around. I assume by "flash hex file" I'm to drag and drop this to the mbed device.
Thanks
Mike

More a question than an issue: I'm looking for some cpu_map.h-type file, means a file where functions can be assigned to the different pins of the MCU. Where can I find that here?

Thanks!

Is it possible to compile for Nucleo 401RE ? What changes should I make?
Yours faithfully
Sergey.

Hello again!,
I've been testing your port again (I've been very busy these days),
After solving the problem of the low speed of processing, I was performing tests with the laser cnc and when exceeding speeds of F: 10000 appear blockages and lost steps when accelerating and decelerating, I use an X raster as a technique of engraving, also the noise has increased quite a lot compared to arduino.
My settings are:

$0=15 (Step pulse time)
$1=20 (Step idle delay)
$2=0 (Step pulse invert)
$3=0 (Step direction invert)
$4=0 (Invert step enable pin)
$5=0 (Invert limit pins)
$6=0 (Invert probe pin)
$10=1 (Status report options)
$11=0.010 (Junction deviation)
$12=0.002 (Arc tolerance)
$13=0 (Report in inches)
$20=0 (Soft limits enable)
$21=0 (Hard limits enable)
$22=0 (Homing cycle enable)
$23=0 (Homing direction invert)
$24=10000.000 (Homing locate feed rate)
$25=10000.000 (Homing search seek rate)
$26=250 (Homing switch debounce delay)
$27=1.000 (Homing switch pull-off distance)
$30=1000 (Maximum spindle speed)
$31=0 (Minimum spindle speed)
$32=1 (Laser-mode enable)
$100=20.000 (X-axis travel resolution)
$101=20.000 (Y-axis travel resolution)
$102=1203.000 (Z-axis travel resolution)
$110=8000.000 (X-axis maximum rate)
$111=10000.000 (Y-axis maximum rate)
$112=1200.000 (Z-axis maximum rate)
$120=700.000 (X-axis acceleration)
$121=6000.000 (Y-axis acceleration)
$122=80.000 (Z-axis acceleration)
$130=70.000 (X-axis maximum travel)
$131=300.000 (Y-axis maximum travel)
$132=200.000 (Z-axis maximum travel)

Use belts gt2 and pulleys of 16 teeth, the drivers are set to 1/4

Arduino worked perfectly with the same configuration (but was not able to process quickly with big data load).
I have been alternating between Arduino and STM32 NUCLEO to compare and the difference is remarkable, the arduino raster movements are smooth, silent and precise, while with STM32 Nucleo the noise increases a lot and only when accelerating and decelerating always in the same point (just before starting / finishing) he suffers a sudden braking and then continues.
I have modified config.h doing several tests with ACCELERATION_TICKS_PER_SECOND and other parameters but without success, I checked and matched the configuration with that of my arduino (by default of grbl) but I have not had any luck either.
These problems only happen when a speed of F: 10000 is exceeded, neither decreasing the acceleration the problem disappears. I think it's a problem of frequency (because of the noise) or with the acceleration algorithm.
I hope you can help me, Thanks!

It would be really nice if software was able to use TMCs driver SPI-configuration.
Especially now, when Trinamic released new TMC5161 drivers capable of running at 4A.
https://www.trinamic.com/products/integrated-circuits/details/tmc5161/

The 2130s had 'headless' mode and were able to run without SPI. 5161s are lacking this feature afaik.

hi! can you please say me what i must cange to control my spindle with relay?

i will only switch on and off.

thanks

Hello, I've noticed that if the machine has an alarm (exceeding soft limits as an example), once the error is cleared the remainder of the travel is completed. My uno board doesn't do this, in fact once I noticed the problem I swapped them straight away. I'm using a screw shield so there was no swapping of wires between swaps. I've used the same configuration file for both boards and I'm using USG.
Cheers
Mike

Hi,

I connected W5500 to the setup. It works fine unless I attach GRBL board to NUCLEO. After that I have a timeout on connection. After removing the shield I have no problems at all.

Did you discover similar issues? Are you able to point me where to troubleshoot?

Thanks!
Filip

Hi,
thanks for your work - it could give me finally the backlash compensation I really need. I even switched to from Arduino2560 to STM32 for this..

I came across an effect that my backlash setting for X was set, but not for Y.
With the debugger I found out that this is caused in MC_Init in MotionControl.c

void MC_Init(void)
{
for(uint8_t i = 0; i < N_AXIS; i++)
{
dir_negative[i] = DIR_NEGATIV ^ (settings.homing_dir_mask & (1<<i));
}

When settings.homing_dir_mask is not zero, which is the case for my mill ("3" for X and Y inverted, not Z), dir_negative[] is {0;3;1;1;1} after init is done. But it should be {1;1;0;0;0}
This leads to always skipping Y backlash calculation further down, as this expression never becomes true when comparing to 3:

if(dir_negative[i] == DIR_NEGATIV)

I replaced the line in MC_Init with

dir_negative[i] = 1 & (settings.home_dir_mask >> i);

and now I get {1;1;0;0;0} in dir_negative[] and thus backlash compensation also on Y!

The line is not yet complete, as I dropped the "DIR_NEGATIV ^" part

Best Regards

Peter

Hi Schildkroet,

thanks for sharing this project!

I'm planning to create a shild for the Nucleo board with RC filters for the limit switches, EEPROM, etc... For this reason, I have looked through the pin assignment, and I found that the pin description in the included picture (nucleof411re_pinout.png) is not up to date. Also, the pins for lim_Y and lim_Z are both the same (Pin_6). Are you planning to change the pin assignment in the near future? Then it wouldn't make sense to design a PCB for the actual pin assignment.

best Andi

I'm having an issue when running the spindle via pwm. I essentially copied the settings I had on grbl-mega into this (grbl-advanced is soo much smoother, thanks the work on it) the grbl mega iteration spun up fine at 24000rpm but without looking into the code on grbl-advanced, when issuing a M3 S24000 the spindle only spins at 18 18388rpm. I'm not sure if this is something I did or not. let me know and I will provide any extra information you need. Thanks

Hi, did you really test it with the STM32F446RE? Because the communication just seems to stall after a few commands.
I tried to build the firmware with MacOS & Windows, both gave me the same result.

$H let's it crash immediately (nothing is connected to the Nucleo board).

Hi,

So I was using GRBL on an Arduino but decided to move to your STM32 build. I have compiled and flashed it to a STM32F411 Nucleo board but I'm having a problem where there is a large delay (Around one second) whenever I send it setting commands ($x=val) and receiving the 'ok' response. There seems to be no problems with the other g-code commands.

Thanks.

Good afternoon.
I apologize for the bad English.
I have a free Nucleo 401RE card. Will your firmware work?
Sincerely, Sergey.

Hello Schildkroet!

With your advice, I uploaded GRBL on my Nucleo STMF429ZI.(USART2 -> USART3 was a solution)
So I can get response from Candle2(also your version) like below.

And now I have one problem in board status.
With STMF411RE, after press 'reset' button, status change to 'Idle'.
But with STMF429ZI, it does not change to 'Idle', stay in 'Hold' status.

Is there more things to change for use STMF429ZI?

Best Regards,
Seunghwan.

Hello,

First, thanks again for all your work!

I don't want to bother you to mutch, but I recognized that when I set DEFAULT_SYSTEM_INVERT_MASK = 0 in default.h, $0 does not become 0. If I set $0 = 0 after compiling, it stays at zero.
Also, I'm not completely sure what for the ENABLE_LIMITS and FORCE_HARD_LIMIT_CHECK flags are or what the difference to $21 (Hard limits) is.
Maybe you could put a small note in the readme about $0. Generally, I would support you with that, but since I don't understand $0 completely, I better don't write any documentation about it :D

Many thanks in advance!

Sorry... ich hab nun mein f411 ;)

aber irgendwie kann ich die A Achse nicht benutzen.. was mach ich falsch?

g91
g1 a200 f100

und dann kommt error 20...

gibts eine alternative kontaktmöglichkeit_?

Small-angle approximation may be used to reduce computation overhead further. A third-order approximation (second-order sin() has too much error) holds for most, if not, all CNC applications.
https://github.com/gnea/grbl/blame/eefe2bb95bb7b21ec2bb87e6ab6e20747e1626c4/grbl/motion_control.c#L135
I know the STM32F401RE has some cool specialized clock tools for the (DMA) direct memory access could hold larger planner buffers and could open the best implementation for g5.1 g5.2
http://linuxcnc.org/docs/html/gcode.html

is there a possibility to implement a 4 axis?

hello,thanks for your great work,i begin to upgrade my cnc form arduino uno to stm32f411re,using your project.but i can't get work the invert coolant flood,i uncomment the define in config.h but not work.
and the spindle enable pin on D14 always remain in 0v.
thanks for your help!!

I'm sure I'm being dumb here, but I threw a new stepper driver on and set the microstepping to 4000, did some calibration and attempted to issue a $100=2554.206 and received the error in the screen shot..

I'm sure I could hunt it down in the code but is this meant to defaut at that low? or didn't I read something and there's a setting I have to change to remove the cap?

what's the maximum resolution? I'll probably just lower the microstepping until I hear of a workaround

Hi Schildkroet!
Thank you for sharing your wonderful project.

Before, I tested GRBL with Arduino-Mega and successfully upload it.
And now I want to use GRBL on STM32 Board and found this project.

I tried to upload GRBL on Nucleo Board144(STM32F429ZI)
I followed your suggestion guide(#8). But finally failed ... :(
I think there are some missing point. Could you give advice about it?

Here is what I changed for STM32F429ZI
(1) Add linker file

• stm32f429zi_flash.ld
• stm32f429zi_sram.ld

(2) Change Compiler Setting

(3) Change Clock Tree

And I make bin file and put in to nucleo board by drag and drop on File Explorer.
But I can`t get response from board. Is there are more things to change?

Best Regards
Seunghwan

Hi,
My name is Emre. I am working on cnc machine project. I have a question. I am using the grbl library. In addition, I have added it in the encoder and I can compare the position information from the encoder with the position information from the stepper motor. However, if the motor has not reached the desired position, I want to move it in code myself. I know how to move from serial terminal. How do I write the move code? Since I am using Arduino, I could not fully understand the operation of the code because I could not debug the code. Can you please help?

Sincerely,
Mehmet Emre Öztürk

Hi, thanks for the great backlash feature!

Just curious though, when I enable backlash and run a simple gcode such as:

G17
G21
G90
G0 X0 Y20
G2 X0 Y20 I20 K20 F2000

The 40mm diameter circle that is meant to be drawn in a continuous smooth motion stops every 90deg. This only happens with backlash enabled. Without backlash enabled it works smoothly as expected.

The 90deg stops make sense as this is where each axis changes direction and injects the backlash steps, but isn't the planner meant to optimize the trajectory so it doesn't stop?

If I only enable backlash on the X axis, then there is only 1 stop at the 180deg point when the X axis changes direction.

Have I done something wrong, or are you aware of this and it is a limitation of this implementation?

Thanks heaps!
Dale.

Hello Schildkroet, I'm just about finished the mechanical side of my router and came across your GRBL port to the STM board. I have a couple of these lying around that I've not really used or followed for awhile. I downloaded your code and tried to compile it in the MBed online compiler and have lots of errors about unfound files. I know they are in your code but I would suspect some of them are also MBed library files (Error: Cannot open source input file "stm32f4xx_conf.h": No such file or directory in "HAL/SPI/SPI.h", Line: 6, Col: 29). My question is, shouldn't I be using the library files as this would mean that updates would be included when I compile again later? Any help for a novice would be greatly appreciated.
JappieMike

how hard is it to port your code to the black pill. its basicly the same core STM32F411....?

STM32F411CEU6

Hi,
Thanks for the good work, I have been using this for a mini mill for a while and it's great.
I am using a Nucleo F401RE and when wiring new buttons recently, I noticed the abort/hold pins were inverted compared with standard Arduino GRBL.
I have Abort on A1 & Hold on A0. I cannot find where this is defined in the code.

I also noticed they were not pulled up, but that's easy to change (line 187 of gpio.c).
Not a huge deal, but I thought I'd report it.

Hello, thanks for the project you are doing, I just bought the board STM32F411RE NUCLEO and tested with my cnc Co2 laser, previously it worked with Arduino Uno but it was too slow because of the speed of Arduino processing. I decided to jump to the platform of STM32 with your port of grbl 1.1, I expected to get more speed by 100Mhz compared to 16Mhz of Arduino but it has not been like that, the speed at which it starts to block is F: 6000 almost the same as Arduino (in the engraving of pictures with dithering, for images with no load as logos or text no problem, can reach speeds of F: 20000, that's why I think it's processing speed problem), I checked the code of the STM32 project and the core clock is the Right. I do not know what I can be doing wrong or it depends on more factors (I have also reduced the number of microsteps to 1/8). My settings are:

$0=15 (Step pulse time)
$1=20 (Step idle delay)
$2=0 (Step pulse invert)
$3=0 (Step direction invert)
$4=0 (Invert step enable pin)
$5=0 (Invert limit pins)
$6=0 (Invert probe pin)
$10=1 (Status report options)
$11=0.010 (Junction deviation)
$12=0.002 (Arc tolerance)
$13=0 (Report in inches)
$20=0 (Soft limits enable)
$21=0 (Hard limits enable)
$22=0 (Homing cycle enable)
$23=0 (Homing direction invert)
$24=10000.000 (Homing locate feed rate)
$25=10000.000 (Homing search seek rate)
$26=250 (Homing switch debounce delay)
$27=1.000 (Homing switch pull-off distance)
$30=1000 (Maximum spindle speed)
$31=0 (Minimum spindle speed)
$32=1 (Laser-mode enable)
$100=20.000 (X-axis travel resolution)
$101=20.000 (Y-axis travel resolution)
$102=1203.000 (Z-axis travel resolution)
$110=8000.000 (X-axis maximum rate)
$111=10000.000 (Y-axis maximum rate)
$112=1200.000 (Z-axis maximum rate)
$120=700.000 (X-axis acceleration)
$121=6000.000 (Y-axis acceleration)
$122=80.000 (Z-axis acceleration)
$130=70.000 (X-axis maximum travel)
$131=300.000 (Y-axis maximum travel)
$132=200.000 (Z-axis maximum travel)

Thanks in advance!

Hello,

I don't understand why, but the baud rate must be set to 57600 for this to work.
Yet we have 115200 in the config.h ...

Do you have any idea where this can come from?

Thank you and have a nice day, Stéphan

 -  My card : "NUCLEO F-411RE"
 -  Compiled with : "EMBITZ 1.11"
 -  Load with : "STM32 ST-LINK Utility"