Using CAN examples?

[jlmxyz]

I wonder if you have any documentation how to use CAN interface ? I'm building a CNC with dual closed loop controled motors for Y and Z axis, right now using STEP/DIR interface, but I think that CAN could be better....

[CAP1Sup]

I wonder if you have any documentation how to use CAN interface ? I'm building a CNC with dual closed loop controled motors for Y and Z axis, right now using STEP/DIR interface, but I think that CAN could be better....

As of right now, you cannot use CAN at all due to a pin allocation issue. I will be fixing CAN functionality soon. There is currently no support for position commands, but I suppose that the commands could be added with only a little work. I'll add it to the long term feature plan. As of right now, I'm still working on making sure that the motor steps correctly and hopefully getting it a little quieter

[jlmxyz]

didn´t knew about klipper.... and seems that klipper support CAN bus also!! https://www.klipper3d.org/CANBUS_protocol.html so if you´re about to rewrite the CAN layer, better make it compatible with klipper low layer protocol

using an octoprint + klipper plugin + usb2CAN + redesigned s42b closed loops boards begins to look like a really nice idea!!!

[jlmxyz]

I cloned the repository, opened it in vscode+platformio plugin, and begun to look at it for the CAN, right now just wanting to give a try without modifications, build went fine, but wonder why there is no main()?

[jlmxyz]

btw is the current state "likely usable"? I don´t speak about accuracy or noise, but just moving and reporting? I think I´ll first use serial line since I´ve an usb to serial to give a try to klipper and add the klipper protocol to it...

[CAP1Sup]

Unfortunately not yet. I've been having some issues with reverse direction on the motor, so it's been taking me a longer time. If you wanted to get started on Klipper, go right ahead and start with your own fork. I'll be happy to help and merge it once the feature is working. If you did start on it, would you mind making a separate branch for it? That way, when your changes are merged into Intellistep, we can continue to work on it without affecting the working build. Also, I'd like to see it set up Klipper as an added feature, so building the support into the existing parser and adding compiler directives (like #ifdef) would be ideal as it would keep the backwards compatibility. I'm hoping to get the code finished up soon, but I've been busy with work lately and haven't had time at home to work on the actual stepping procedure. It should be finished within the month. I'd love to have support on the boards for both Marlin and Klipper, as well as features such as position correction.

[CAP1Sup]

The only concern that I have is that the CAN voltage might not be 5v. You'd have to check on that, otherwise you could accidentally cause the magic smoke to escape.

[jlmxyz]

hummmm according to my investigations, 3.3V is provided by U8 which is a 1117 3.3 regulated power (either 1A or 800mA depending of manufacturer) the input pin of U8 is connected to a chip marked "BNOR 11" pin3or4 through a 15kohm restistor, there is a huge self right beside this chip and pin2or5 is connected to VMOT through a huge diode.... I made some mesurements, with +5V of DIR/STEP interface unpluged (just let the VMOT, GND, EN, DIR, STP), the board still works and VCC pin of PCA82C251 is 5V so it seems that if BTT software engineer are bad, electrical ones did the work by far better than the MKS servo engineers.... 1) they use VMOT to create internal power, which is by far better 2) VMOT is droped to 5V using the unknown chip BNOR11, 3) 5V is then regulated to 3.3V using the 1117 3.3 regulated power... too bad they didn´t used pins that could do analog sampling for the switches, else I could have replaced the switches by a connector and used them for end stops and thermal... without having to do huge redesing... so at least the CAN chip is correctly powered. I´ll start work on klipper, I think I´ll add a menu in the display to activate or not the support of the protocol, once I understood how they do it... don´t like too much #ifdef... it makes the code unreadble... better use compiler optimizations to whipe the dead code using #define KLIPPER_SUPPORT 0 if(KLIPPER_SUPPORT) { .... } this way you can still indent the code and allow folding Le sam. 15 mai 2021 à 20:11, Christian Piper ***@***.***> a écrit :

…

The only concern that I have is that the CAN voltage might not be 5v. You'd have to check on that, otherwise you could accidentally cause the magic smoke to escape. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#19 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AJUDLRDFHXKYB4EDCRBCQMLTN22MZANCNFSM442TREQA> .

[jlmxyz]

I've some question because I started schematics for the new board. the motor are driven by analog write on PB4 and PB5, but I don't find where theses pins are setup and if they are connected to PWM's timer1 of stm32f103R8.... where are the pins configured? thanks and regards Le dim. 16 mai 2021 à 00:22, jl forums ***@***.***> a écrit :

…

hummmm according to my investigations, 3.3V is provided by U8 which is a 1117 3.3 regulated power (either 1A or 800mA depending of manufacturer) the input pin of U8 is connected to a chip marked "BNOR 11" pin3or4 through a 15kohm restistor, there is a huge self right beside this chip and pin2or5 is connected to VMOT through a huge diode.... I made some mesurements, with +5V of DIR/STEP interface unpluged (just let the VMOT, GND, EN, DIR, STP), the board still works and VCC pin of PCA82C251 is 5V so it seems that if BTT software engineer are bad, electrical ones did the work by far better than the MKS servo engineers.... 1) they use VMOT to create internal power, which is by far better 2) VMOT is droped to 5V using the unknown chip BNOR11, 3) 5V is then regulated to 3.3V using the 1117 3.3 regulated power... too bad they didn´t used pins that could do analog sampling for the switches, else I could have replaced the switches by a connector and used them for end stops and thermal... without having to do huge redesing... so at least the CAN chip is correctly powered. I´ll start work on klipper, I think I´ll add a menu in the display to activate or not the support of the protocol, once I understood how they do it... don´t like too much #ifdef... it makes the code unreadble... better use compiler optimizations to whipe the dead code using #define KLIPPER_SUPPORT 0 if(KLIPPER_SUPPORT) { .... } this way you can still indent the code and allow folding Le sam. 15 mai 2021 à 20:11, Christian Piper ***@***.***> a écrit : > The only concern that I have is that the CAN voltage might not be 5v. > You'd have to check on that, otherwise you could accidentally cause the > magic smoke to escape. > > — > You are receiving this because you authored the thread. > Reply to this email directly, view it on GitHub > <#19 (reply in thread)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AJUDLRDFHXKYB4EDCRBCQMLTN22MZANCNFSM442TREQA> > . >

[jlmxyz]

what is really interesting, is that we can un-solder the dip switches and make this board a complete Klipper controler board...
PA15 used for calibsw have analog => can be used to attach a board for termal sensor
PB11 & PB10 are attached to TIM2_CH4 and TIM2_CH3.... maybe they can be used for PWM driving fans??? (EDIT : yes they can be used to drive FAN!!!! it will require to add control board but this can be done as a module attached near to the FAN)
PB3 can be used to attach a board and start/end switches.....

this is really interesting with this board we can have a klipper subsystem that can act as

1. motor driver
2. thermal sensor
3. one fan control and 2 end/start switches detector
   or
   3bis) two fan control and start switch
   or
   3ter) two fan, 2 switches in parallel, start or end is found according to motor direction... if one switch triggers, if we were moving toward start it's a "homing reached event" if we were moving toward end, it's "exceed limit event"

and more, s42b has no can, but this board as UART so can act as a proxy.... ok, I begin to love this idea!!!

[jlmxyz]

Ok, I started the design of the add-on board, and I'm a bit sad... in fact PA15 hasn't ADC input possibility...
so, I'm a bit annoyed....
it is possible but not using the dipswitches but the press button, must look if I find some connector that can go in place...
what can be done if using the v2.0 board as is

1. unsolder dipswitches, press buttons and C19 and r20
2. solder on fingerprints some connector
3. have the add-on board handle the press button with following remappings :
   1. PB0 from Mode-K3 to Thermal sensor
   2. PB1 from Valid-k4 to TIM3-Ch4 or TIM1-CH3 (FAN)
   3. PA3 from Next-k2 to TIM2_CH4 (FAN)
   4. PA15 from CALIB_SWITCH to END-START switch (with moving direction determining the end or start condition)

the issue with that is that I wanted to keep the possibility to just unplug the addon board and go back to previous behavior with for example a menu on the oled (as well as activating the klipper mode into the menu....) but with this hardware modifications the add on board will be required to operate the oled, thus there is no "software switch".....

another thing, I begun to look at klipper code, they don't use arduino libs but regular stm32 libs, arduino libs don't support CAN for stm32 if I'm correct, this means that we will have to include both libraries and I'm not sure it won't over consume flash...
is arduino libs mandatory?
didn't looked at your code, to know if it's an easy switch from one solution to another....

[CAP1Sup]

If you wanted to include the STM code, then we could covert/adapt it. The CAN library is already included in the betterReverse branch, awaiting debugging. Most of the stm32f1 libraries are included with the Arduino, so I'll have to take a look at the code you're referring to. Good luck on the addition board, I can add an optional setting to enable or disable the related code

[jlmxyz]

klipper CAN code : https://github.com/KevinOConnor/klipper/blob/master/src/stm32/can.c
anyway, if want to make a can bus, the R19 resistor must be removed, appart on one device, since the CAN require a single 120Ohm resistor per bus
I just begins to look at klipper code, in particular for messaging, I guess the easiest way will be to have the klipper code stored and compile only the required parts,
I don´t know if it might helps, but klipper can perform some kind of buffering, maybe this could help for acceleration computing and then noise reduction

[bilalibrir]

TTB_S57B_REV1.pdf
hello everyone I'm happy to share my version of the BTT S57B stepper controller.
there is still a lot of work to do especially because I don't own any module, and everything is done from some pictures and from the code.
I'm trying to make a design that uses the blue pill board.

[jlmxyz]

hummm ok you use the arduino lib.... I'm not used to it, I done freertos or the like dev, never arduino... so I don't know how the pins are configured... one other thing, I've also s42Bv1 board, the MCU is a STM32F030C8.... didn't looked into datasheet yet, but do you think your Firmware will be compatible? because they seems broken as well (sensor seems same and overall design seems the same only can isn't present on this board) Le dim. 16 mai 2021 à 20:28, Christian Piper ***@***.***> a écrit :

…

As for the timer, they are connected on channel 1 and 2 of timer 3 of the STM32. That's a hardware linkage that is handled automatically by the Arduino library. Keep in mind that you will have to figure out the correct CAN bus pins, as the ones that I had listed are wrong (I think, check with BTT's code) — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#19 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AJUDLRBYB2ZCIGK7JZXFQ2TTOAFFDANCNFSM442TREQA> .

[CAP1Sup]

I'm hoping it will be compatible. The pin configs will have to be moved to their own board files. Other than that, there isn't a lot of changes that need to be done

[IhorNehrutsa]

Here CAN is remapped to PB8, PB9, but the CAN driver is connected to PA11 and PA12 on the schematic.

Intellistep/src/lib/CAN/CAN.cpp

Lines 17 to 27 in 49a0e18

	void CANInit(enum BITRATE bitrate)
	{
	RCC->APB1ENR |= 0x2000000UL; // Enable CAN clock
	RCC->APB2ENR |= 0x1UL; // Enable AFIO clock
	AFIO->MAPR &= 0xFFFF9FFF; // reset CAN remap
	AFIO->MAPR |= 0x00004000; // et CAN remap, use PB8, PB9
	RCC->APB2ENR |= 0x8UL; // Enable GPIOB clock
	GPIOB->CRH &= ~(0xFFUL);
	GPIOB->CRH |= 0xB8UL; // Configure PB8 and PB9
	GPIOB->ODR |= 0x1UL << 8;

And I find commented CANInit() and can't find another CAN code

Intellistep/src/hardware/canMessaging.cpp

Lines 16 to 24 in 49a0e18

	// Function for initializing the CAN interface
	void initCAN() {
	// Initialize the CAN interface
	//CANInit(CAN_1000KBPS);
	// Clear the command string
	CANCommandString = String();
	}

Please clarify for me about the CAN.

[CAP1Sup]

I disabled CAN in the master branch because I didn't know the correct pins and couldn't find a library at the time. The CAN implementation is fixed in the betterReverse branch, which will eventually become the dev branch. I still need to do testing, but all of the code should work theoretically. I'll do testing and debugging on it later, but right now my main focus is getting the motor to step correctly.