drf5n

I ran into a problem trying to compile Marlin from the command line with a: HARDWARE_MOTHEBOARD=13 ARDUINO_VERSION=022 make that didn't like the https://github.com/ErikZalm/Marlin/blob/Marlin_v1/Marlin/Marlin.h#L34 line.

My german is awful, but it makes sense. If you interpret the D value with the units,(if I'm decoding the extra voodoo properly) it appears to be using the observed...

Looking into it more, PID controllers are used on lots of industrial systems systems with dead-time. The solution they use is to set the D (or Td) term to extrapolate...

I got interested in the PID tuning from the reliability and ease of configuration side of things. My thermistor circuit was pretty noisy, (ultimately because of poor power supply filtering)...

Branch 'issue74units' is simpler--it changes the comments, and lays the foundation for other changes by defining constants for: PID_SCALE_P: kP in counts/C into the internal unit mibicounts/qC PID_SCALE_I: kI in...

Branch "Issue74_ilim" changes the anti-windup logic from a hardcoded DEFAULT_I_LIMIT=384 to adaptive anti-windup, eliminating one of the voodoo PID variables. I think DEFAULT_I_LIMIT=384 would be appropriate for processes that operate...

The changes in "Issue74_ilim" would be similar to what would be needed to make a "bumpless" change to the P,I, or D values with the M130, M131, M132 -- you...

The issue74 branch allows per-heater definition of PID parameters in the config file. It lets you define the parameters in count-degreeC-second units and also lets you define the integral and...

issue74 now doesn't use or allocate space for the unused variables "watts" and "dead_time" stored in the DEFINE_HEATER macro calls.

Well, I'm learning things. Teacup is interesting in the way it manages precision. It was fun to follow RepRap's PID back through Marlin, Sprinter, Klimentclip, tonokip, FiveD, old fiveD http://sourceforge.net/p/reprap/code/3306/...