Help locating the source of inaccuracy

You have “P” for g1 moves and “T” for g0 moves.

I am pretty sure you can adjust the Juction deviation on the /lcd. I would try setting it to you 500 line accel and lower the JD to see if that improves it.

Accel is hard to tune.
We use junction deviation to tell it when to even consider acceleration - eg if we go from 3mm/s to 5mm/s don’t bother even calculation an accel.
Then if an accel is triggered it gets smoothed out by the S curve acceleration so it is not a hard trapezoid.
MAX and Starting accel should be the same, I believe the terms are just mixed up in teh docs vs the firmware.
We also have another accel I can’t think of right now that was giving me horrible issues at high speeds during the LR3 testing.

It helps a little but we still have 4 steppers and 2 steel rails swinging around.

It is more about getting the machine to START moving smoothly to get rid of your artifacts. I only tuned the junction deviation for slow routers. You should tune it for fast movements. JD is when it will consider accelerating, so if this was set to high it would not even try to accel and just slam around at every corner. With tiny segments at high angle changes they might not even get going fast enough to trigger an accel. If you go too low though it will try to accel on tiny angle transitions and circles will get messed up.

Not yet but I have been working on the WIFI headless solution the last few days and just made some progress between our messages. Grbl tests 2022 - #80 by vicious1

1 Like