Servo Tuning, target values and base parameters for CNC-applications generally?

All things related to the Centroid Acorn CNC Controller

Moderator: cnckeith

Post Reply
Spartan117
Posts: 155
Joined: Mon Apr 11, 2022 3:37 pm
Acorn CNC Controller: Yes
Plasma CNC Controller: No
AcornSix CNC Controller: Yes
Allin1DC CNC Controller: No
Hickory CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No
Location: Germany, Stuttgart

Servo Tuning, target values and base parameters for CNC-applications generally?

Post by Spartan117 »

Hey guys,

i am about to tune my leadshine-servos now. I found some resources about the process itself, this shouldn't be a problem. But i can't find any statements about the base- and target-values for THIS certain application to be achieved?
I am talking about the target and basic values, that should be reached/used when tuning servos on such a cnc-lathe.

The tuning shall be done with a "worst case run". That means, it should be done with a move at maximum speed and acceleration of the application.
But can anyone tell me, what target acceleration is normal/should be used for a 11x27"-lathe in X and Y?
What acceleration of the axes is normal, what is fast, what is too slow?
Not trying to win anything here, just asking for "normal conditions"

In other words, i am searching for the time-period, in which the target-trapezoidal-curve of the move (looked at in the scope) rises from 0 to max. speed.

And what maximum position error should be reached for such an application? At least and ideally, on a lathe like this?
How much µm error is good, and when does "unneccessarily precise" begin? In µm-deviation on a lathe in X and Z?

Did you use any filter-functions of your drivers, too? Or should a machine-axis be tuned purely via the usual PID parameters?
I read, that most filters do affect the precision of the move, so they can worsen the positional-accuracy within the move severily

And what settling-time should be achieved after tuning? (the timespan from the end of the trapezoidal-curve to the "position-reached-signal" in the scope)
Most documents are talking about 2-8ms here, but again, these are mostly material-handling-applications, not cnc-machining.

And what deviation is a good "positioning-completed-range" for an application like this?
Yaskawa is calling this "positioning-done-signal" /COIN, Leadshine calls it "positioning-completed".
Both refer to certain range, a "positioning-completed-range", that is given in "degrees of turn", repectively encoder-counts, respectively millimeters of move. When the position reaches this range, the "Positioning-completed-signal" is activated.

In some tuning-manuals i read, that a good "positioning completed-range" is 0,01° of turn of the motor, but they all are not tuning machining-axis, but stuff like positioning-arms, material-handling, conveyor-belts and so on.
Is this value of "Within 0,01° around the targeted position" correct for this application, too? Pitch is 2mm/rev in X; resp. 5mm/rev in Z
What value (in µm-travel of axis would be also fine) is ok here?

It would also be very helpful, if you simply can post some screenshots of your "tuning-done"-curves in the scope (with diagrams scaling), the above asked values can be seen/measured there.

Short summary:

What is the target-acceleration for the axes of a lathe?

What max. positioning-error is good?

May i use filters on cnc-machine-applications?

What is an sufficient positioning-done-range for a lathe?

What settling time should be achieved on a lathe?


Thank you,

Marc
German-speaking, so please be patient with my english-skills :)


Post Reply