Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
16 years ago

FIR Compiler: some questions

Hello,

i have made some interesting observations using the Quartus II 9.0 FIR Compiler.

I have designed an FIR Hilbert Transformer of order 200 as a FIR filter using MATLAB.

Then i have imported the calculated coefficients to FIR Compiler 9.0.

There under 'simulations' it is possible to force the system clock in the testbench. Here i have chosen 40 ns, but at the testbench the DUT gets an external clock of tclk = 10 ns. It seems to me that this adjustment does not work, or am i wrong?

Another Question:

Specifying the parameters, there is a button 'use nonsymmetrical structure' but it isn't possible to set the check mark. Does this behaviour appears due to using the web edition? Detects the FIR compiler automatically if there is a symmetrical coefficient structure and reduces the 'multiplications'?

Using the coefficients for the Hilbert Transformer calculated by MATLAB, every second coefficient is zero. I don't think that this behaviour will be observed and these 'multiplications' will be reduced by the FIR Compiler, but i have not found any information about reducing 'multiplications' having zeros as coefficients in the user guide.

Thanks for your answers,

T. Wollmann

2 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor

    I'm relativity new to Digital filter design and I was wondering if you could explain how you designed the FIR Hilbert Transformer in Matlab and then exported the coefficients to the FIR compiler? I would like to be able to do something very similar to what you have done.

    Regards

    David
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor

    Hi,

    i saved the coefficients calculated by matlab in an *.txt - File. It is important that every coefficient is separated by its own row - do not use "," or any other character after each coefficient. You can also use scientific notation.

    It is necessary, that places and decimal places are separated by "." and not by ",", like it is in german, or anything else.