hello again benjameanos, just a few more things :

- pushing is all about timings, collet opens, and the feeder start to push until it hits something; the feeder does not know that the collet is open, it simply pushes continuous, all the time ... but things get a bit. just a bit, more complex during the bar change sequence : feeder must retreat the remnant, and he does a few things only after the collet opened : actually, the feeder does not know that the collet was opened, but only that there was a code that sends a signal to open the collet, thus the feeder does not move based on a confirmation code there are some parameters for "dwell" inside the feeder, only to achieve a propper timing

- if you take those 2 things, from above, written with blue, then is easy to see that the feeder is (almost) always in tension; at least on iemca, on some models, etc, this leads to origin lost, and is ok to 'zero it' from time to time : i ' zero it ' every day 1st time when i 'zero it', it was shifted with 9mm, but this was at the begining ... ' zero ' may be lost after a long term operation, or after a sudden failure; there are ( at least ) 2 methods to verify the 'zero' :
... inside the cnc cabinet. look how long is the new bar when it is bringed for the 1st cut off operation : if 'zero' moves, then the bar will also shift
... operate the feeder in manual mode, and move it to retreat position, and check the coordinate displayed on the feeder remote


ok ... now, about sudden failures when changing : sometimes, the feeder keeps on pushing, aparently without a reason, but, i noticed that actually it repeats the ' bar change macro ', and something happens, something gets lost, i don't know what, and he pushes like crazy : to handle this, i have used variables ( couting parts ) inside the feeder macro, and if it accidentaly tries to run that macro, it will be stopped if this action was performed recently ( no more than 'x' parts were cut ); in my case, the feeder was ocasionaly repeating only after the 1st part, so i needed to 'block' this potential behaviour that may have occured on 2nd part ... i was lucky, because there was a patern, thus ocasionaly only after 1st part was machined / kindly