Data Sheet

Closed Loop Relative and Tracking Position Modes
136 Advanced Digital Motor Controller User Manual V1.8 August 28, 2017
Differential
Gain
Propor tional
Gain
Output
Measured Position
-
x
Σ
x
dE
dt
Integral
Gain
x
dE
dt
E= Error
Sensor
Desired Position
Integrator
Limit
FIGURE 11-4 . PID algorithm used in Position Mode
The Differential component of the algorithm computes the changes to the error from one
ms time period to the next. This change will be a relatively large number every time an
abrupt change occurs on the desired position value or the measured position value. The
value of that change is then multiplied by a user-selectable Differential Gain and added to
the output. The effect of this part of the algorithm is to give a boost of extra power when
starting the motor due to changes to the desired position value. The differential compo-
nent will also help dampen any overshoot and oscillation.
The Integral component of the algorithm performs a sum of the error over time. In the
position mode, this component helps the controller reach and maintain the exact desired
position when the error would otherwise be too small to energize the motor using the
Proportional component alone. Only a very small amount of Integral Gain is typically re-
quired in this mode.
In systems where the motor may take a long time to physically move to the desired
position, the integrator value may increase significantly causing then difficulties to stop
without overshoot. The Integrator Limit parameter will prevent that value from becoming
unnecessarily large.
PID tuning in Position Mode
As discussed above, three parameters - Proportional Gain, Integral Gain and Differential
Gain - can be adjusted to tune the position control algorithm. The ultimate goal in a well
tuned PID is a motor that reaches the desired position quickly without overshoot or oscil-
lation.
Because many mechanical parameters such as motor power, gear ratio, load and in-
ertia are difficult to model, tuning the PID is essentially a manual process that takes
experimentation.