Partly Serial Systolic FIR Filter Implementation
This example shows how to implement a 32-tap lowpass FIR filter using the Discrete FIR Filter block.
The two filter blocks in the example model implement an identical partly serial 32-tap filter. The top block configures the serial filter by specifying the number of cycles N between input samples. This spacing allows each multiplier to be shared by N coefficients. The second block is configured to use a certain number of multipliers, M. These two configurations result in the same filter implementation. For 32 symmetric coefficients, there are 16 unique coefficients. Therefore the filter shares each of 2 multipliers between 8 coefficients.
The model shows two ways of applying input samples, depending on the rate of the rest of your design.
Open the model. Inspect the top block parameters. Filter structure is set to
Partly serial systolic and Specify serialization factor as is set to
Minimum number of cycles between valid input samples. Number of cycles is set (to 8) using a variable,
numCycles. In the lower block, Specify serialization factor as is set to
Maximum number of multipliers. Number of multipliers is set (to 2) using a variable. The variables are defined in the
PostLoadFcn callback function.
From the color coding, you can see the rate of both filter blocks is the same, while the rate of the generated input samples is different.
Run Model and Inspect Results
Run the model. Observe the input and output signals in the generated plots. The code to generate the plots is in the
PostSimFcn callback function.
Use the model toolbar to open the Logic Analyzer. If the button is not displayed, expand the Review Results app gallery.
Inspect the rising edges of
Generate HDL Code
To generate HDL code from either Discrete FIR Filter block, right-click the block and select Create Subsystem from Selection. Then right-click the subsystem and select HDL Code > Generate HDL Code for Subsystem. The two blocks generate the same HDL code.