This is not normalised (since the normalisation is different for different FFT implementations). It also gives an absolute amplitude plot. I would prefere a dB plot re arbitary myself.Code:
fs=2000; %sampling frequency
dt=1/fs; %sampling interval
y=(cos(300*pi*t)+cos(500*pi*t)).^3; %1 seconds worth ofsignal
spp=abs(sp); %absolute value of frequency component
%(you could square this as well to get something like a power)
df=1; %frequency resolution (in Hz)
f=0:df:999; %frequency scale for spectrum
spp=spp(1:1000); %cut out the part of spectrum below Nyquist frequenct
ok..someone told me I could also use this:
Y = fft(y,512);xlabel('frequency (Hz)')
Pyy = Y.* conj(Y)/512;
f = 20000*(0:256)/512;
title('Frequency content of y')
can you expalin to me what each one does?
y is a function
well in fact this code was use to find the frequency components of a signal buried inside a noisy time domain signal using the fft function in MATLAB.
the assignement says :
Using the same technique, find in MATLAB the frequency components in the followingy(t) = (cos(300πt) + sin(500πt))^3
so i think i need to use the same code
ur code does not work anyhow
Undefined function or variable 'dt'.
btw is there a command that can clear the command window?becauase each time i want to write something new i have to exit then reenter it