Low order narrow FIR band pass filter

I need to design a 0.15-1.5Hz FIR band pass filter for the TMS320F2808PZA TI processor with sampling frequency of 300Hz and a maximum order of around 500 if symmetric (or 250 if not) because of RAM. I tried using the Parks-McClellan algorithm using two low pass filters and subtracting them as given in the code below. However, I could only get a 0.4-2.9 Hz filter since we need to use a maximum order of around 500. The passband ripple of the filter should be as low as possible. What type of filter should I use? Thanks for your help.

The 0.4 low pass filter code is given below:

%% LP1Hz filter Designrp = 0.5; % Passband ripple

rs = 65; % Stopband ripple

fs = 300; % Sampling frequency

f = [0.00 1.309]; % Cutoff frequencies

a = [1 0]; % Desired amplitudes

% Compute deviations

dev = [(10^(rp/20)-1)/(10^(rp/20)+1) 10^(-rs/20)];

[n,fo,ao,w] = firpmord(f,a,dev,fs);

b = firpm(n,fo,ao,w);

bb=min(floor(log2((2^31-1)/max(abs(b)))),ceil(64-32-log2(sum(abs(b)))));

B=round(b*2^bb);

LP_Scaling = round(2^bb*sum(b));

The 2.9 Hz low pass filter code is given below:

%% LP2.9Hz filter Design

rp = 0.01; % Passband ripple

rs = 120; % Stopband ripple

fs = 300; % Sampling frequency

f = [2 5.001]; % Cutoff frequencies

a = [1 0]; % Desired amplitudes

% Compute deviations

dev = [(10^(rp/20)-1)/(10^(rp/20)+1) 10^(-rs/20)];

[n,fo,ao,w] = firpmord(f,a,dev,fs);

b = firpm(n,fo,ao,w);

bb=min(floor(log2((2^31-1)/max(abs(b)))),ceil(64-32-log2(sum(abs(b)))));

B=round(b*2^bb);

LP_Scaling = round(2^bb*sum(b));

Answer 1

[Wayne King]

Hi do you have the Signal Processing Toolbox? First I would question whether you really need a bandpass here when you want your lowest passband frequency at only 0.1 Hz, why not just remove the mean (DC) from the signal and then use a lowpass filter.

For example:

d = fdesign.lowpass('Fp,Fst,Ap,Ast',1.5,3,0.5,40,300);
Hd = design(d);

Gives you an equiripple FIR design of order 326.