Kirk Wallace wrote: > On Mon, 2007-10-08 at 21:52 -0500, Jon Elson wrote: > ... snip > >>>problem? Maybe missing pulses or noise adding pulses? Thanks for any >>>replies. >> >>It could be anything, so you have to investigate. First, put >>halmeter on the HAL pin ppmc.0.encoder.2.count, rotate the >>spindle to a known position, and read thhe halmeter value. >>(The above assumes you have the spindle encoder on axis 2, if it >>is on axis 3 use that number between encoder and count.) Now, >>turn the spindle exactly one turn and read again. The >>difference should be equal to the expected 5000 counts. > > > Thanks Jon and Chris. I verified the encoder had 5000 pulses/revolution. > I then did the above and got varying counts in the mid 4ks. Upon > inspecting the encoder I found the disk had a slight wobble towards one > side of the sensor. I shimmed it away from the sensor a little and I now > get 5000 counts per revolution plus or minus one count on occasion. My > threads now come out fine. I like these US Digital disks, but > apparently, I still haven't learned the proper way to fit them to a hub. I got a big eBay motor that had a fancy 6-channel (ABC, plus UVW for brushless motor commutation) encoder made by Renco in it. It has a tab you pull out from the side that causes fingers to reach in from the encoder body and center it on the hub. Then you tighten the hold-down screws before pushing the tab in. This aligns the read head with the tracks on the disc. I had to realign my encoder when I got the motor. I think US Digital may have some little moded collar that serves the same purpose. But, I guess if the disc hub is running eccentric on the shaft, that won't help. > I wonder if there is a way to have EMC check encoders by comparing pulse > count against the index. If index doesn't appear when the count predicts > it, a warning could be issued. >
You could pretty easily make up such a program that would command the motor to run at a slow rate but open-loop, and put the encoder counter into index-enable mode, so it counts up and resets on every index pulse. The program would check the highest number it sees every cycle, maybe reporting a histogram of the highest count registered every turn. With no belt or coupling, you could run this test for a half hour or so. I've never actually written a stand-alone HAL program, but it should be doable. 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-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
