Tuning with a Plant Measurement

Take a plant measurement. All traces have been turned ON in the plots below. The plots shown below are from a Trust TA9000. The following plot was aquired with the following settings:

Disable the amplifier and enter a trial tuning setup. The first example will be Kp = 100, Kd = 1000. Refresh the data by either clicking View->Refresh or pressing F5 (shortcut key for the same).


Kp = 100 Kd = 1000

You can see Kp and Kd reflected in the dark green trace (current filter). The dark red trace shows generally good closed loop behavior (no major underdamped peaking). There is an obvious resonance at ~680 Hz. A resonator filter will be used to cancel this resonance as a further example.

Now enter a postfilter of 680 Hz center frequency, 80 Hz bandwidth, and -25 dB gain. Click refresh (F5).

The above plat shows that the resonance was largely reduced. This appears to be acceptable behavior. The stage can now be enabled and moved at this point.

NOTE:
This example is not intended to be used as a tuning tutorial. The tutorial simply shows how to operate plant simulations. Keep in mind that you can tweak the filter parameters and refresh (F5) the data as many times as you necessary before enabling the amplifier.

Tuning with plant testing with non-zero tuning parameters

Tuning with the plant model with non-zero tuning parameters is virtually the same as with zero tuning parameters.

The same procedure as above was followed, but the starting tuning parameters were Kp = 100, Kd = 1000 instead of Kp = 0, Kd = 0. You can see that the "As Measured Filter" is a non-unity response (not 0dB, 0 degrees). At this point, you may want to disable the amplifier while experimenting with tuning.

To follow the next step from the previous example, a resonator with 680 Hz center frequency, 80 Hz bandwidth, and -25 dB gain is added. When you are happy with the tuning in simulation, enable the amplifier and verify your performance.

Previous | Next