Kirk Wallace wrote: > I have been having a problem lately. Once every few homing procedures > ends with a runaway. I am homing with index. It starts with the axis > movement towards the home switch, then I see it back off about .2", then > reverse towards the home switch again. It then moves very slowly for > about a second, then bolts off away from home at a rate I have never > seen the axis go. I have the max velocities set at just below the rated > encoder pulse rate. The runaway rate is much higher than this. I also > noticed that the position value at e-stop is in the 300 inch range.
What hardware are you using to count encoder pulses? The "moves very slowly for about a second" part is the latch velocity stage of the homing sequence. You do know that velocity is something that YOU are supposed to set for your machine, right? Likewise for the initial velocity and direction of homing movement. See the drawing at http://www.linuxcnc.org/docs/devel/html/config/ini_homing/index.html As for the run-away, if it is moving faster than the axis limits, the most likely cause is something funky going on with the index pulse. When the index pulse arrives, the position feedback to EMC is supposed to get reset to zero. The command from EMC also gets reset to zero, so the PID loop is happy. If the index is resetting but the command is not, or vice versa, the PID loop is going to see a huge error, and will zoom off into the distance in an attempt to correct it. > I plan on setting hal scope up to record what is happening, but in the > mean time, if anyone has ideas for solutions, I would appreciate hearing > about it. You need to scope it. Capture the position feedback, position command, PID output, and index enable. Trigger on the rising edge of index-enable. After you capture a trace, adjust the vertical scales so everything is readable, make a screen-shot (I believe alt-print-screen works), and post it somewhere (imagebin.org perhaps). Then post a link to the pic here, along with a description in words of what you saw. Regards, John Kasunich ------------------------------------------------------------------------- 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-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
