Tom Easterday wrote: > On Oct 17, 2011, at 3:58 PM, Peter C. Wallace wrote: > >> That simply looks untuned, have you varied FF1, added some P term etc etc >> > > I am having a really hard time finding any EMC PID values that make it better > then the graph I sent before. I can make it much worse by adding even just a > little P ie, p=1, p=1.1, p=1.01 - not to mention P=10 or higher. It looks a > little smoother, rolling down to zero but starts off at tenths or hundredths > of inches before doing so. If I change FF1 to anything other than 1, even > 1.01, I get a much worse result (tenths of inches of ferror) and sometimes > immediate faults if it is much higher than 1. I've tried some I term and > even some D term (and combinations of both just for the fun of it). I can't > get anything that has lower following error than having everything set to 0 > and FF1=1. > If you haven't seen my tuning page at http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?PWM_Servo_Amplifiers You might find it helpful. It is a completely ad hoc method, but has worked for me quite well. It is aimed at users of my (Pico Systems) boards, so some of the pin names will be a bit different on Mesa hardware, but the PID will work the same way.
Jon ------------------------------------------------------------------------------ All the data continuously generated in your IT infrastructure contains a definitive record of customers, application performance, security threats, fraudulent activity and more. Splunk takes this data and makes sense of it. Business sense. IT sense. Common sense. http://p.sf.net/sfu/splunk-d2d-oct _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
