PID tuning in Position Mode

PID tuning in Position Mode

The Roborun PC utility makes this experimentation easy by providing one screen for chang- ing the Proportional, Integral and Differential gains and another screen for running and monitoring the motors.

When tuning the motor, first start with the Integral Gain at zero, increasing the Proportional Gain until the motor overshoots and oscillates. Then add Differential gain until there is no more overshoot. If the overshoot persists, reduce the Proportional Gain. Add a minimal amount of Integral Gain. Further fine tune the PID by varying the gains from these posi- tions.

To set the Proportional Gain, which is the most important parameter, use the Roborun util- ity to observe the three following values:

Command Value

Actual Position

Applied Power

With the Integral Gain set to 0, the Applied Power should be:

Applied Power = (Command Value - Actual Position) * Proportional Gain

Experiment first with the motor electrically or mechanically disconnected and verify that the controller is measuring the correct position and is applying the expected amount of power to the motor depending on the command given.

Verify that when the Command Value equals the Actual Position, the Applied Power equals to zero. Note that the Applied Power value is shown without the sign in the PC utility.

In the case where the load moved by the motor is not fixed, the PID must be tuned with the minimum expected load and tuned again with the maximum expected load. Then try to find values that will work in both conditions. If the disparity between minimal and maximal possible loads is large, it may not be possible to find satisfactory tuning values.

Note that the AX3500 uses one set of Proportional, Integral and Differential Gains for both motors, and therefore assumes that similar motors, mechanical assemblies and loads are present at each channel.

AX3500 Motor Controller User’s Manual

99

Page 99
Image 99
RoboteQ AX3500 user manual PID tuning in Position Mode, Applied Power = Command Value - Actual Position * Proportional Gain