I am having a little bit of trouble with a series of blended one-axis moves. I am moving a slide table back and forth. Each "move profile" consists of a series of blended moves to carry the slide across the table and then back. The velocities change at specific positions across the table, creating a 'trapeziod' effect across the table and back. If I create the exact same velocity/accelleration/position profile across the table and back (except reversed), I would expect the actual velocity profile, as graphed by LabVIEW in realtime, to be identical. (See example below). However, it seems that the blended moves are delayed somewhat, so that the velocity profile across the table appears offset from the profile of the return trip. I have to use t
rial and error to get an exact overlap of the proviles, by changing the across trip positions to be a little less than the return trip positions.
My question is, since the blended move funtion surely operates as it should, then exactly how is the blend accell/decel superimposed on the current velocity move? I would like to know these equations so that I can properly calculate my profiles instead of relying on trial and error.
Thanks,
Michael Hampson
XL Automation, Inc.
EXAMPLE:
If the slide table is 10" travel, then a blended profile may consist of the following:
Start position is 0"
MOVE Position Velocity Accel/Decel
1 2" 100RPM xxx rpm/s
2 5" 200RPM xxx rpm/s
1 10" 100RPM xxx rpm/s
1 5" 100RPM xxx rpm/s
1 2" 200RPM xxx rpm/s
1 0" 100RPM xxx rpm/s
This profile should create an offset trapezoid both on the across trip and the return trip. This it does, but the trapezoid created on the return trip is offset from that created on
the across trip. I would like them to everlap exactly.
Michael Hampson
President
XL Automation, Inc.