Panasonic MINAS A6 Multi Series Technical Reference page 129

5.2Manual adjusting function
Usage example of velocity feed forward
The velocity feed forward will become effective as the velocity feed forward gain is gradually increased
with the velocity feed forward filter set at approx. 50 (0.5 ms). The positional deviation during operation at
a constant velocity is reduced as shown in the equation below in proportion to the value of velocity feed
forward gain.
Positional deviation [unit of
command] =
With the gain set at 100%, calculatory positional deviation is 0, but significant overshoot occurs during
acceleration/deceleration.
If the updating cycle of the positional command input is longer than the driver module control cycle, or
the pulse frequency varies, the operating noise may increase while the velocity feed forward is active. If
this is the case, use positional command filter (1st delay or FIR smoothing), or increase the velocity
forward filter setup value.
Usage example of torque feed forward
To use the torque feed forward, correctly set the inertia ratio. Use the value that was determined at the
start of the Real-time auto tuning, or set the inertia ratio that can be calculated from the machine
specification to Pr 0.04 Inertia ratio.
The torque feed forward will become effective as the torque feed forward gain is gradually increased
with the torque feed forward filter is set at approx. 50 (0.5 ms).
Positional deviation at a constant acceleration/deceleration can be minimized close to 0 by increasing
the torque forward gain. This means that positional deviation can be maintained at near 0 over entire
operation range while driving in trapezoidal speed pattern under ideal condition where disturbance torque
is not active.
Zero positional deviation is impossible in actual situation because of disturbance torque.
As with the velocity feed forward, large torque feed forward filter time constant decreases the operating
noise but increases positional deviation at acceleration change point.
5-45
command velocity [unit of command/s] / positional loop
gain [1/s] x (100–velocity feed forward gain [%]) / 100