SourceForge.net wrote:
> Initial Comment:
> As I worked with positioning systems there ware always cascade loops 
> consisted of outer position and inner velocity loop.
> The most simple system is a proportional (P-regulator only) loop for position 
> error-> output=Kp*(Position_setpoint-Position_encoder);
> Output of position regulator is of course the needed velocity. The velocity 
> loop consists of proportional-integral (PI) regulator .
> 
> Utilizing only PID regulator for position loop as used in EMC2 is not the 
> right approach, It cannot work good. I had never seen PID regulator for 
> positioning tasks and professional servo systems use cascaded position and 
> velocity loops only.
If you truly believe a positioning-only loop can't work, please 
have a look at this document :
http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?PWM_Servo_Amplifiers
and especially the last graph.  I'm sorry the scale unit doesn't 
show there, but it is 200 uInch per division.  This was moving 
at 40 IPM, I think, during the continuous part of the jog.  I 
think the loop control is excellent, and far better than the 
physical machine can deliver at the cutting tool.  Note that the 
D term has a velocity component in it, so you get some of that 
velocity-loop response.    I agree that a straight 
positioning-only loop is a little less stable than nested 
position/velocity/torque loops.  You have to run with a lower 
gain to remain stable, and so feedforward help to keep error 
under control.

Of course, if you have a velocity servo amplifier with a DC 
tachometer, that will also work well.  We use the same PID logic 
for both velocity and pseudo-torque servo systems.

Jon

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
Emc-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-developers

Reply via email to