Brian,
I suppose that all that needs to be considered is two motion stages, the third is an idependant stage that has the same function as one of the others. So there is one rotary stage to spin the object and one linear stage to move the laser beam up & down. I have analysed the actual positions of the linear stage during motion and it does indicate an incosistant speed that seems to match the wobbly lines, although this was using software time position polling. I think with contoured moves the normal accel and decel have no effect, I will need to deal with that myself in the position arrays that I send to the NI card. At the moment there is no accel or decel, I command constant move lengths with a constant time interval throughout the move. The cards I have are PCI-7358's that do support on the fly PID parameter changes (I think), I just hope that they can be changed very quickly if I need to go that way! I guess I'll find out soon, but until I can tune the stage for the slow / accurate move I do not know if there is a requirement for on the fly PID parameter changes. Maybe this is an issue with the mation stage its self, although It was sold to us as a nano-positioning stage.
Thanks / Martin.
Certified LabVIEW Architect