Dustin Monroe wrote: > Ok, I messed around with a 9v battery and adjusted the reference gain > and loop gain and got all three axis working without any following > errors.
You need to arrive at a properly tuned drive before closing the loop with EMC and any further tuning, I'm not sure that you have done that. EMC can't correct poorly tuned drives. You can use Halscope to assist with this by plotting the position of the joint while applying a suitable command signal. Studying the curve produced will assist you in arriving at satisfactory tuning of the drive (you can see overshoot, oscillations etc.). To do this, you need to open up min-ferror and soft limits sufficiently, disconnect the command input from EMC and use your own command signal. Setting Halscope trigger can get fiddly in this situation. Remember that working like this is open loop. Do not be tempted to reconnect/switch EMC command back to the drive without first shutting everything down. Remember to reset min-ferror and soft limits when finished. Halscope and the Calibration tool are your friends. You will benefit enormously by getting to grips with Halscope. In common with Dave, I also find that the use of any I or D (in EMC) adversely affect the tuning and consequently leave these set to zero (as he says, you may find different). Once I commence tuning in EMC, I set P to 1, I,D,Bias,FF0,FF1, and FF2 to zero. I open up the min-ferror to a few inches, plot f-error, joint-position-cmd and joint-pos-fb in Halscope. Then increase P in steps and note the response and f-error. On my drives P tops out around 30 (too much gain). In practice, I find my best tuning set-up is achieved when P is set to 10 - 15 (you will understand why when you start studying the fine details of the curves plotted in Halscope). After that, FF1 and then FF2 in small increments. Expect to work to 4 dp, on one axis mine are set at FF1=0.153 and FF2=0.0004. For the feed rates I am interested in, my f-error is down to +-0.0001 inches. As already stated, your parameters will be different. Richard > Now I have a new problem to solve, I am getting oscillation that > I cant get rid of and when I adjust the gain enough to solve it, i get > following errors. The oscillation doesn't seem bad except on the y axis. > the z and x feel almost like they are going back and forth between faces > on the encoder but the y has a pronounced vibration. Im thinking i will > need to rent an oscilloscope to fine tune the motors. > ------------------------------------------------------------------------------ This SF.Net email is sponsored by the Verizon Developer Community Take advantage of Verizon's best-in-class app development support A streamlined, 14 day to market process makes app distribution fast and easy Join now and get one step closer to millions of Verizon customers http://p.sf.net/sfu/verizon-dev2dev _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
