Theoretical basis

The PID speed controller needs to be set and optimised depending on the application.

The precise mathematical description of all parameters of the control circuit has been shown often to be rather extensive and difficult in practical applications. Therefore, a simple procedure shall be presented here by which the controller can be set practically.

For that, a speed step-change (without ramp) needs to be given as a reference variable at the input of the controller. The step response (speed actual value) should be recorded for evaluating the controller setting. When specifying the speed step, make sure that the drive remains operating below the torque limit. 

 

Set the controller as follows:

  1. Setting ID100 'Speed control proportional gain KP' Kp, with ID101 = 0 (Tn), ID102 = 0 (Td)
  2. Setting ID101 'Integral-action time speed control TN' Tn, with ID100 = const. (Kp), ID102 = 0 (Td)
  3. Setting ID102 'Differentiating time speed control TD' Td, with ID100 = const. (Kp), ID101 = const. (Tn)

 

Step response of the optimised speed control circuit 

 

For an optimally set PID controller, the actual speed value may overshoot a setpoint step-change by no more than 20%.

 

Two PT1 filters can be configured at the output of the speed controller.

See ID32928 'Time filter 1' and ID32929 'Time filter 2'

 

 

Relevant parameters:
Parameter Name

ID100

'Speed control proportional gain KP'

ID101

'Integral-action time speed control TN'

ID102

'Differentiating time speed control TD'

ID32928 

'Time filter 1'

ID32929 

'Time filter 2'

Setting the proportional gain Kp

Set ID102 ('Differentiating time speed control TD', Td) and ID101 ('Integral-action time speed control TN', Tn) to 0, the controller then works as proportional controller.

By increasing ID100 'Speed control proportional gain KP' KP, the controller should be made to overshoot (50 % overshoots). The actual speed has a course then similar to the curve with the solid line:

 

 

Halve the determined value for 'Speed control proportional gain KP' KP and enter the halved value in ID100.

Setting the reset time Tn

Using the integral proportion (I-proportion) in the controller, the controller deviation resulting from the P controller is adjusted.

The integration time is reduced (starting at an initial value e.g. 100ms) until the settling time is minimal. If the reset time is set optimally, the actual speed value curve (jump answer) roughly follows the curve with the solid line:

 

For an optimally set PI controller, the actual speed value may overshoot a setpoint jump by no more than 20% as an answer.

Setting the differential time Td

The differentiating time Td is extended until the desired dampening of the jump answer is reached. The curve with the solid line serves as a reference point for setting the D-share.