Measuring Transfer Functions of Discrete-Time Systems

 Learn how to quickly measure transfer functions of discrete-time systems in Simulink®. The device under test used in this demonstration is a completely custom 1 MHz Harris elliptic bandpass filter built out of Simulink primitives. You’ll start with an already designed filter, but you still need to verify the implementation. In other words, does the filter implementation meet the design objectives in the transition band, stopband, and passband?

In this example, drive two similar bandpass filter implementations in parallel using random noise excitation and measure their responses simultaneously. Overlay their magnitude responses in one spectrum analyzer and their phase responses in a second spectrum analyzer. You’ll also learn how these transfer functions can be tuned while the simulation is running.  

You can download the R2023b examples used in this demonstration here: https://github.com/kschutz68/harris_b...

Related Resources:

- Cross Spectrum–Based Transfer Function Measurement:    • Cross Spectrum–Based Transfer Functio...  

- Deconstructing the Cross Spectrum–Based Transfer Function Measurement Technique:    • Deconstructing the Cross Spectrum–Bas...  

- See more videos:    • Kschutz  

- Discrete Transfer Function Estimator: https://bit.ly/4aobuWh

Chapters: 

00:00 Introduce the Device-Under-Test, a harris bandpass filter, in block diagram form

01:00 Why Measure Transfer Functions in Simulink?

01:35 Another implementation of the device-under-test in state-space form

02:45 Running the model

04:35 Under the hood of the Transfer Function Measurement block

07:20 Viewing the source code

09:42 Using the spectrum analyzer as a display device

10:38 Modifying the DUT while the model is running

11:45 Review the mask of the transfer function measurement block

12:30 Navigating to the Discrete Transfer Function Estimator block

14:30 Downloading this example from github