search

bdring / FluidNC

The next generation of motion control firmware
Other
1.62k stars 386 forks source link

Feature: link 2 identical cnc controllers so we can drive more motors #1362

Open aguaviva opened 3 weeks ago

aguaviva commented 3 weeks ago

Machine Context

I am using a ESP32Duino (and it works like a charm) to control 4 motors of a 6 axis robot (ProArm RS2200). I would like to use a second ESP32Duino linked as a slave to the first one to drive more motors.

Feature Description

The 2 ESP32Duino could be linked somehow so they share the PWM clock.

Other Approaches

I believe alternate solutions wouldnt be feasible.

How I Can Help

I would like to hear and try your implementation ideas myself

MitchBradley commented 3 weeks ago

We already have a 6-motor solution with the various 6-pack boards and the Jackpot board. It doesn't make sense to expend a lot of effort to try and synchronize two ESP32 when we already have good, well-debugged, fully supported ways to drive 6 motors.

aguaviva commented 3 weeks ago

Sure I could buy some stuff, but I am here to learn and for the challenge ;)

MitchBradley commented 3 weeks ago

Okay, but we are not offering to help you. We prefer to spend our time doing stuff that helps everybody. Supporting all of the existing users with the myriad of existing hardware is challenge enough.

Maybe you could have a third MCU that sends commands to two 3-motor controllers.

aguaviva commented 3 weeks ago

Feel free to ignore my messages and lets see what others say.

AndreasEpp commented 2 weeks ago

I have had good results by using multiple senders. I open up a serial connection to all ESP32 controllers and start sending commands to all at the same time. The starting deviation between controllers is about 2ms in my case. To coordinate the timing between controllers I use G93 inverse time mode on all moves. I am working right now on a new approach where I pre generate gcode files for all controllers, upload them via xmodem and start them at the same time. The starting deviation is still at 2ms, but I don't have to keep senders busy and worry about buffers.