tuenhidiy / cnc-plotter-from-dc-motors-and-optical-encoders Goto Github PK

I'm using the dual channel L298N h-bridge instead of the shield and 2 identical geared motors w/encoders. It was a pretty straightforward port since the library( https://github.com/AndreaLombardo/L298N ) initializes similarly and has similar API.
example:
directionX = FORWARD;
becomes:
directionX = L298N::FORWARD;

motorX.run(directionX);
motorY.run(directionY);

becomes:
motors.runA(directionX);
motors.runB(directionY);

The odd effect I'm getting is when I test the DEADBW_X and DEADBW_Y, which I'd defaulted both to 4.5, by using my finger to move the encoder wheel on the back of the motor and when I move the Y motor the deadband stays at ~1/2 rotation of the encoder wheel and it triggers changes in the X motor at the ends of the deadband. Moving the X motor encoder does not cause feedback on the Y motor and the deadband feels more like I'd expect. ie motor hums and hums louder as I try to move further out of the deadband.

I swapped motors and lowered the DEADBW to 2.5 and indeed on the X axis the motor resists a much smaller movement but again the Y motor has a wide deadband , half a rotation of the encoder, and when it hits the ends the X motor jumps.

Since I have identical motors I also changed the PID to identical values(those of the default X axis).
Ideas?

Hi and great work and write-up on Instructables. I was wondering if you'd looked and/or done anything like a port of your project to something like the Arduino Leonardo for single axis control instead of doing dual axis in the Arduino Mega?

I have the idea of a larger drawing bot using L298N H-bridges to drive automotive wiper motors and it just seemed like it could be streamlined and packaged differently with a single axis control handled by a small Arduino Leonado for each motor/encoder/driver unit. I also have some large geared motor units with encoders mounted on the unit and would probably be used for the prototype.