1
$\begingroup$

I have calculated the transfer function of an FIR filter

$$ y[n] = x[n] + α · x[n − R] $$ This is what I have $$ H(z) = 1 + αz^{-R} $$

Now I should plot the square of the amplitude response. So I calculated the frequency response by inserting e^jωTs for z. I should choose α = 1/2 and R = 3. So I get

$$ H(e^{jωTs}) = 1 + αe^{-RjωTs} = 1 + \frac{1}{2} e^{-3jωTs} $$

$$ |H(e^{jωTs})|^{2} = (1 + αe^{-RjωTs})^{²} = 1 + \frac{1}{2} e^{-6jωTs} $$

I am also not sure here about the algebra of the square of the magnitude.

Now I should plot from 0 to ωs and I think I am missing Ts. I should probably just choose a value, right? How is the usual approach?

$\endgroup$
  • 2
    $\begingroup$ You have (at least) one issue with your "squared amplitude" the first equality in the last line does not hold, one should multiply $H()$ by its conjugate $\endgroup$ – Laurent Duval Jun 27 '18 at 14:59
  • $\begingroup$ I don’t know what Ts is but I don’t like it. Also, that isn’t how you square numbers. Once you change it to multiply by the complex conjugate, don’t forget to FOIL! $\endgroup$ – Dan Szabo Oct 23 '18 at 12:41
1
$\begingroup$

If the impulse response $$h[n]=\delta[n]+\alpha \delta[n-R]\hspace{1cm}=\{1,0,0,...\alpha\}$$ with the ellipsis being R presumably, the Fourier transform $$H(\omega)=e^{-j 0 \omega}+\alpha e^{-jR \omega}=1+\alpha e^{-jR \omega}\hspace{1 cm}\exists\hspace{1 cm}\mathbb{C}$$ where magnitude is defined as $\sqrt{x^{2}+y^{2}}=|x+iy|$

Since we want the square of the amplitude response $=|H(\omega)|^2$, we can square the real and imaginary parts of the transform$$\Re{\{H(\omega)\}}^{2}+\Im{\{H(\omega)\}}^{2}\hspace{.5cm}=\hspace{.5cm}(1+\alpha \cos(-R\omega))^{2}+\alpha \sin(-R\omega)^2\hspace{.5cm}\exists\mathbb{R}$$ Then plot from $\omega=[0,\pi)$

Might look something like this with your parameters
graph

Please tell me if this gives the expected response because I'm relatively new to continuous/algebraic Fourier theory

$\endgroup$
  • $\begingroup$ Looks like you are using a and alpha interchangeably. Also you lost your factor of a/alpha on the sine term. $\endgroup$ – Dan Szabo Oct 23 '18 at 12:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.