biesse rover 316 retrofit

welcome colleague i believe 322 is similar if i can help something. I have reached the stage of physical connection of inputs and outputs that I need ... I would like drills to work, but not nessesery for my job. I was looking at Bevins' file, but nothing is clear to me from python script, but I thought something simpler to try with o word. Thank you

Here's the first cut, I'm not very happy with two things, the first looks bad tunning y axis where like real waves on wood, dro shows very little deviation of the encoder back and forth, maybe 0.3mm ... the second thing is to reduce velocity to change direction , mostly on arcs, the movement is not very smooth but there is a pause at the entrance and exit to the curve, I tried to add G64 P0.1 to gcode but it's the same..does anyone have an idea how to solve constant velocity``? I'm also trying to somehow solve the program by waiting for the spindle to speed up to the given command and only then start the next line, but also without success
Thanks

 

A friendly advice, take 3 to 5 days off, do whatever but no machines.
Thank me later.
Back on topic: Nice to see it moving.
Upload the hal and ini files so we can have a look if something jumps as strange.
Use the Hal Scope to plot some traces of joint commanded position and joint feed error, that always helps when troubleshooting such issues.

You are right about that for rest, I would very much like to, but I have a mini series of chairs, so even with two machines, I am afraid that I will not be able to meet the deadline ... and I need this one even if it is as it is now. Take a look when you can...i almost upload all the files in one of the previous posts. Thanks a lot I'll try a little to focus on the Y axis, I think I put P very little and maybe that's why it doesn't hold the axis still.

Waiting for the spindle to come up to speed should be relatively simple. Linuxcnc has the Hal pins spindle.M.at-speed, and almost all VFDs have or can have using a programmable output an at speed singnal. Simply connect them.

As to your motion issues, be sure to check bearings for play. I have often seen such behavior caused by worn parts causing spongy or springy backlash in an axis. You may be able to reduce it's impact by lowering performance (accel and vel) and softening pid tuning, but it is much better to find the real source of the problem. On your style machine that cantilever axis, both it's ways and drivetrain are highly suspect and have to be in perfect condition to get good performance.

Thank you very much
I solved the moving of the next line by putting G4 Pxx (xx number of seconds pause) after the m3 command, since I changed the VFD, will try to find instructions (hitachi j300) ... Y axis works better I doubt the encoder but it can be as you said axial gap in the bearing of the ballscrew ... now it is not terrible, as you said I reduced the P (gain) on Y and it seems correct now. What bothers me more is the break at the entrance of the cutter into the curve and the exit..then linuxcnc drastically reduces the feed .. but not always, on larger arcs it happens about 0.5 sec stops and then continues the trajectory, not so smooth movement as he would say