Brother TC-225 / TC-229 adventure!

PCW wrote: OK some things here really make no sense

Shortening the cables should make no real difference, especially in dynamic performance
I can see that there might possible be more hum with longer cables, but that does not seem to be
the issue.
This may indicate some kind of ground loop. This can be an problem with drives that do not have
differential inputs. There are many things that can be done to eliminate ground loops if that is the issue.

Can you plot this overshoot/oscillation at .01 G with P very low (say 1) FF1 properly set and I/D/FF2 all 0

Have you checked the 7I77s 5V supply?

Also can you post your hal/ini files here?

ill check this tonight.
the 5v for the 7i77 is coming from the PC. I do have a 5v psu inside the machine, maybe I should use that...

stupid question, but linuxcnc is tuning the position loop and expecting a velocity mode drive with these pid settings correct?

if it was a torque drive, are there different settings? or is it just different tuning using the same position loop?

Velocity mode drives are more standard, and are generally easier to tune.

Torque mode drive can be tuned with the same position PID loop, but it is a bit trickier to tune. And it may be difficult to get good stiff response with the relatively slow servo thread rates typically achievable with Linuxcnc (slow compared to the velocity loop rates in most servo drives}
A torque mode drive can also be controled by two PID loops in Linuxcnc, one feeding the next. This works similar to the conventional position loop in the control commanding a velocity loop in the drive. This two loop setup tunes just like normal. (I have a machine setup running this way.)

i put the old control back on, we got a tool holder and tried some cutting. its pretty good!

1/2" 3 flute zrn coated cutter in aluminium. 1.5-1.75hp cut. the small amount of backlash doesn't seem to cause any chattering or issues.

were going to do some paying work on it for a bit, then we'll get back to the linux servo tuning.

we may also try to do a little cutting test with the linux tuning as it is now, to see what real world impact the tuning resonance has.

ihavenofish,

I recently purchased a TC-225 and have been wrestling with using LinuxCNC or an industrial controller. I am curious what pushed you towards LinuxCNC?

Thanks,
Graham

ive used linuxcnc for a while on my routers. so i was familliar, and knew it would function adequately especially with the atc and rigid tap.

unfortunately i cant seem to tune the tc225 servos with it. the X is close. maybe even acceptable. the Y is meh, the Z is not really usable.

i should not that that video i posted with the lumpy motion, that is all in the trajectory planner, not tuning as it turns out. so theres some big bug there - i did get the same thing just programming a circle. i never saw this in the older versions of linuxcnc (still with the new trajectory planner) so i think its just the newest version i installed being borked.

my options at this point are:
- get new servos and drives
- leave the stock control and move to the tc 229 with new drives, and step and direction option if analogue misbehaves

i suppose its entirely possible that reverting to an older linuxcnc could help the tuning as well, but that seem a long shot.

so how is your machine is it tuned and running or not there yet?