Kirk Wallace wrote: > Here is the latest on by Shizuoka mill problem. I haven't spent a lot of > time on this, but I did replace the hard drive with a new install of > EMC2 2.3.1 and Ubuntu 8.04. > > At the beginning of the mill conversion when I installed the VFD, I had > a problem with noise on the analog input, which I made usable at the > time. I looked into this a little more. I shut down different parts of > the system to see if the noise went away. There was no effect, so the > noise must be from the VFD, which has an input mains filter and shielded > motor output lines, or the DAC may have a problem. I need to figure out > how to better study the VFD signal. > > Hoping that the PC work might have helped, I ran the encoder wheel > program four more times with an empty tool holder, which ran the part > normally, but the backplot had an anomaly. The program calls for the > center hole first, then the slots and finally the outer detents. On the > last two runs, at some point (which I didn't see) the red center hole > paths disappeared, leaving the white lines. > > Since the actual tool paths seemed to complete properly, I put the tool > holder with an end mill in and ran the program. The machine stopped > again at the same spot at the nineteenth hole. I got a screen capture of > the EMC2 status at the point the motion stopped: > > http://wallacecompany.com/machine_shop/Shizuoka/Screenshot.png > http://wallacecompany.com/machine_shop/Shizuoka/Screenshot-1.png > http://wallacecompany.com/machine_shop/Shizuoka/Screenshot-2.png > > I am not familiar enough with the status screen to know what is > significant, so if something jumps out, please let me know. > > The symptoms don't seem to indicate what the problem might be. The only > difference between a successful run and not, is whether a tool touches > the workpiece. > > The plan now is to move the VFD to a steel enclosure.
That might make a big difference. Not knowing other details, make sure your grounds are all connected to one point. I wonder if there is a voltage potential between the electronics ground and the spinning tool which gets large enough to "break the program" when you reach a certain location on the part? This reminds me of all kinds of disk drive problems when the spindle was not grounded properly. Digital and HP drives had a brass piece touching graphite in the center of the spindle. Remember that spinning metal can generate a lot of voltage and it might be isolated as it spins on lubricated bearings. After all the changes you've made, I don't believe there is a computer problem. You seem to have issue with hole 19. My guess is that number 19 has nothing to do with it, it's the angle under which the part is positioned that matters. Encoder problem? Can you change the G-code in such a way to come to the same spot from the "other side" or skip one of the segments you are working on? Another option is to skip hole 19 and see if that's the only spot where milling stops or it would happen on the next hole also. How about running simplified task on the part with a simple "touch and go" where the holes are supposed to be without actual milling? You need some kind of a debugging script, i.e. G-code to simulate most basic CNC moves without actual cutting. I would even try to run the whole program using styrofoam or light wood instead of using metal part. -- Rafael ------------------------------------------------------------------------------ Are you an open source citizen? Join us for the Open Source Bridge conference! Portland, OR, June 17-19. Two days of sessions, one day of unconference: $250. Need another reason to go? 24-hour hacker lounge. Register today! http://ad.doubleclick.net/clk;215844324;13503038;v?http://opensourcebridge.org _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
