[MadTooler]

How are these test variables for the Bode plots to be determined?

Noise Injection

• Amplitude
• Cutoff Freq Hz

Measure

• # Samples

Thanks.

[TomKerekes]

Hi MadTooler,

Regarding Stimulus have you read this?
Bode Screen

Regarding # Samples: - more samples will take longer and provide a more averaged result. Usually a value of 20 works well.

HTH
Regards

[MadTooler]

Tom,

Yes, read it already, and was left unclear as to how I know what amplitude and cutoff frequency apply to my system.

It states "Noise Amplitude has the same units as Position Measurement." I understand then that the amplitude is in encoder "steps," but how to determine correct for my machine? Is there a use in results of varying amplitudes?

What parameters of my/any machine direct the cutoff frequency value?

Thanks.

[TomKerekes]

Hi MadTooler,

Please see if this article helps:
Dynomotion

Regards

[MadTooler]

Tom,

I had also read that one. It still seems like such a best guess as long as it is not shaking itself to death approach. I am reaching for some values of calculations that start things in the right ballpark.

I have found, from pure experimentation, that 100-300 amplitude seems to treat my machine nice, but have no idea if this range is appropriate. I can achieve acceptable plots, but again do not know if I have set the bar at the wrong level.

I am mostly using the 20 hz for the cutoff. Is this a typically suitable value for BDC motors?

I notice I do not typically get the crossover bar and results at the bottom of the plot when I set the frequency higher than 100. I think I understand why if my crossover is under that value.

It seems clear to me that I can fake my understanding of the process well enough to get my machine usable at this point. In the future, I plan to study up on the deeper subject to refine the tune if I choose/need to improve optimization.

Thanks again.