- Tue Jul 28, 2009 9:11 am
#77828
Maybe google + the following hints will help:
Use a FIR architecture for the filter with as few coefficients as possible.
Choose ADC sample rate high enough (>> 25 kS/s).
Probably, deriving its coefficients using 'frequency sampling' method is the easiest way: calculate values of Gc(f) for several f, do inverse FFT, this gives you the impulse response h(n) of the filter. All this you can do in Excel (fourier function).
Now simulate the frequency response of the real h[n] with the finite number of coefficients and with the limited resolution (can again be done in excel). Tune the coefficients to optimize the response.
Vrms calculation is done by summing the squares of a number of samples. Take care of scaling..
(I see that most digital filter design software provide no options for random frequency transfers so these are of limited use here. However matlab (or maybe scilab) could be useful)
hope this helps.. good luck...