Comments on U1 controller

[jhinkle 1]

I'd like any positive/negative comments people might have.

I started on this 1.5 years ago but it got shelved because of illness.

I'm back into the development and wanted to see if anyone else is possibly working on something similar or has information/comments on my direction.

I purchased a U1 almost two years ago. I have a habit of redesigning things to make them better for me -- for it it was redesigning the U1 control board.

I designed a hi-end stepper motor controller that can be programmed to micro-step up to 128 or 256. That was phase one.

Phase two was the controller that drives the stepper.

The current board and software (Marlin) builds a trapezoid velocity profile for each segment using the axis that has the most steps in that segment. Marlin then uses Bresenham Line algorithm to step all 2 or 3 or 4 motors.

The issue I have using the Bresenham Line algorithm is that it was designed for raster graphics and "pulls" the secondary steppers as it controls the main axis for that segment. If you analysis the secondary axis tracks, you will find that they don't run at a smooth profile but can have their speed/velocity jumping back and forth by half speed.

I'm just attempting to take my good U1 and make it better for myself.

My change in control design is to build a trapezoid profile for EACH axis and have them run independent but in parallel and synchronized with each other. My goal is to have a much smoother segment deposited (you might have to look under a magnifying glass to see the difference).

Has anyone looked into something similar?

Thanks in advance for any comments.