@ Hansjakob Thank you for the summary of your experiences. The whole homing thing seems to be easier with linear encoders, especially absolute ones. There one could just adjust the offset and then slave the second axis to the position of the master axis plus the offset. The ones you describe seem to be relative ones with index pulse. This probably makes it easier to adjust the gap between the two axis. But the base problem seems to be the same. Mines are relative resolvers without index pulse.
Somehow I imagine the homing sequence like that: The master axis does it's homing sequence, slaved by the second axis of course. When the home position is found the homing sequence of the second axis starts, again both axis moving together. When the second home position is found the offset can be calculated and corrected. From this point both axis are syncronized. Thats how I imagine it, but I have no clue how to do it in EMC. I read an IRC conversation from 2004 about this topic, but no idea what the developers finally decided to do. Thanks for your answer. @ Aram Thanks for you too. I agree, this would be the simplest and maybe most robust solution. But somehow I think it should be possible in software. This would be a more elegant solution. @ Jarl Thanks for the links. Seems to be interesting. I am pretty attracted by the gantrykins from jepler: http://sourceforge.net/tracker/index.php?func=detail&aid=1075625&group_id=6744&atid=356744 But only the fact, that the 2 axis can be independently jogged, if nessesary. But this also is a very dangerous setup, one wrong mouse click - bang! :-) Also I would rather prefer a master axis with a "normal" position loop setted by the motion controller, and another position gap loop for the slave axis. The homing sequence described by jepler seems to be close to what I imagine. Except that there is the possibility that the two axis do not move in unison after the first axis is homed. And I really would prefer to not have the slave axis shown as an additional (say A) axis which can be moved independently. An independent move should only pe nessesary after a crash (this could be made with a special 4 axis configuration for after-crash situations), and of course for the offset correction in the homing sequence - or everytime the home switches are passed over. Jarl, you say you slave both axis. Can you explain more about your hal setup? Or how you plan it. What changes in EMC are you thinking about? I would be willing to help implementing, however I do not really know the EMC sourcebase yet. Only played around with .comp's and HAL. (by the way, is the project interested in any .comp files, even for simple components like a differential analog output?). regards Marc ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
