Input signal -->abs block-->ideal low pass filter block-->output signal - Ideal low pass filter in Simulink

Hello everyone. I have the following time-continuous system: input signal -->abs block-->ideal low pass filter block-->output signal. In simulink I make the abs block with the Fcn block. My problem is to get ideal low pass filter with a 3000Hz band and 1 amplitude (linear scale). How could I get it? Maybe I should use Lowpass filter block? In this case what are the right parameters?

Thank you for your time.

Answer 1

[Julian Rosker]

Simulink does not support an ideal filter block. As a substitute, a non-ideal filter with a very small transition band can be created. A filter structure of "Direct-form FIR" with a passband just below 3000Hz and a stopband of just over 3000Hz should achieve this. Below shows one example of this filter.

Frequency response for the above filter:

Alternatively, you can design your own filter using the Analog Filter Design block. This won't be an ideal filter either, but may be more desirable for your implementation. To examine the frequency response of, for instance, a custom Butterworth filter, see https://www.mathworks.com/help/signal/ref/butter.html. Other options include cheby1, cheby2, ellip, and besself.

Further information on the pre-R2015b Lowpass Filter Simulink block can be found here: https://www.mathworks.com/help/dsp/ref/lowpassfilterobsolete.html .

[Gennaro Arguzzi]

Hi @Julian Rosker, I tried to run the simulation, but Matlab shows the following error (below there is also the block diagram):

[Julian Rosker]

Depending on your data, you may be able to add a zero-order hold after the signal to discretize the input. Information on that is here: https://www.mathworks.com/help/control/ug/continuous-discrete-conversion-methods.html

If your data is not well suited for that, try using the Analog Filter Design block. Most likely you would want a Bessel or Chebyshev filter.