filter

in butter(1,wn) i know wn is normal frequency =(cut off/nyq.freq) in amplitude modulation what will be wn=? if carrier freq=fc and i calulat 0:.001:1 sample and message freq fm

Answer 1

[Rick Rosson]

Code:

   dt = 0.001;         % sampling time (seconds per sample)
   Fs = 1/dt;          % sampling rate (samples per second)
   Fnyq = Fs/2;        % Nyquist frequency (hertz)
   Fco = fm + bw;      % Cut-off frequency (hertz)
   wn = Fco/Fnyq;      % Normalized cut-off (dimensionless)

Explanation:

• dt is the sampling time (in seconds per sample)
• Fs is the sampling rate (in samples per second)
• Fnyq is the Nyquist frequency (in hertz)
• Fco is the cutoff frequency (in hertz)
• bw = bandwidth (in hertz)
• Fc is not needed to compute wn

You should make sure that the following conditions are true:

   bw << fm << fc

and

   bw << fm << Fs

HTH.

Rick

[mohamed al-asklany]

excellent ,now if i have a message like cos(2*pi*fm*t), and i want to build function which make frequency modulation

for this message it does not work and if it work(some times) the receiver can not be bone my problem is to how to make discriminator and how to put the message in frequency and how to see spectrum of signal and somthing else i add to am signal a noise (awgn) is this code correct

snr=10^(snr/10);

vn=var(tx)/snr; %AWGN channl

n=sqrt(vn)*randn(1,1001);

tx=tx+n;

assuming t=0:.001:1

and tx is the transmitted signal

Answer 2

[Rick Rosson]

Please try:

   dt = 0.001;
   Fs = 1/dt;
   Fnyq = Fs/2;
   Fco = fm + bw;
   wn = Fco/Fnyq;

HTH.

[mohamed al-asklany]

i can not understand what you write please, more details.

thanks

[Rick Rosson]

Try it.

[Walter Roberson]

Fs - sampling frequency

Fnyq - nyquist frequency

fm - message frequency

Fco - cutoff frequency

wn - wn to use for butter()

I have not figured out what bw is, or where fc fits in.

[Rick Rosson]

bw = bandwidth (in hertz)

Fc is not needed to compute wn

[Rick Rosson]

The idea of bandwidth is that the message has a spectrum that extends across a range of frequencies centered at f = fm. You can assume that the range is [ fm-bw ... fm+bw ].

[Rick Rosson]

You should also make sure that the following conditions are true:

bw << fm

and

fm << fc