how does the "pid control input filter" know the frequency of the input data? What's the difference between "Butterworth Filter" and "PID Control Input Filter"

[iyearning]

Hi all,

I am using low pass filters in my PID control. I found I have 2 choice, one is the "Butterworth Filter " and the other is "PID Control Input Filter". For "Butterworth Filter", I need to clearly define the frequency of the input data. However, the "PID Control Input Filter" does not need information of frequency. How does it know the frequency? They can give similar results and "PID Control Input Filter" seems give results closer to the real signal. What's the difference between them? How to use them correctly?

Thanks

best

[PiDi]

As it is said in PID Control Input Filter documentation:

Applies a fifth-order lowpass finite impulse response (FIR) filter to the input value. Filter cut-off frequency is designed to be 1/10 of the sample frequency of the input value.

You do not need to know actual sampling frequence when using digital filters (FIR, IIR) - they're designed for normalized frequency, i.e. 1 = sampling frequency, 0.5 = half of the sampling frequency, etc. This is just how filters math works.

[iyearning]

Dear PiDi,

What should I do if I want to cutt off specified frequency when using PID control input filter?

Thanks

best

[johnsold]

From what PiDi posted it appears that you cannot change the cutoff frequency of the PID Control Input FIlter. It is fixed at 1/10 of the sampling rate.

If you need to filter at a different frequency, you will need a different filter.

Lynn

[iyearning]

Hi Lynn,

So I have to use the butterworth filter instead of pin control input filter? Or there are other options? I am measuring analog voltage signal. I have to use analog filter, not the digital filter, right?

Thanks

best

[johnsold]

The Butterworth filter will allow you to set the cutoff frequency.  There are many filter options. It is not clear what you expect to gain by filtering so that makes it difficult to make specific recommendations.

All software filters are digital filters.  Many of them perform very similarly to hardware linear circuit filters of the same type. Some things can be done in software filters which are difficult or very expensive to do in hardware.

Lynn