spectral analysis, frequency spectrum, power spectral density
2 vues (au cours des 30 derniers jours)
Afficher commentaires plus anciens
I have power=2000 float data points and time= 2000 float time intervals. in 200ns.
is there any way i can put this in matlab to see some spectral analysis of frequency and other things. basically i have x and y values of power vs time. I want to manuplite this data to show frequency values which i expect to be in 500-1000MHz range.
or may be draw a fft plot. but the problem is i just know these two variables (power and time) is it possible? if yes then how?
2 commentaires
Réponses (3)
Rick Rosson
le 12 Mai 2012
N = 2000;
StartTime = 0;
StopTime = 200e-9;
dt = (StopTime - StartTime)/N;
Fs = 1/dt;
dF = Fs/N;
f = -Fs/2:dF:Fs/2-dF;
0 commentaires
Wayne King
le 12 Mai 2012
"thanks for ur answer....it helps a lot but some clarity i need...when i do freqz(w) for the power data. i see x values of freq as normalised freq as api rad/sample. can i change this to actual frequency. say like in MHZ?"
Have you read the documentation for freqz()? It shows that you can use the sampling frequency as an input argument
[H,F] = freqz(B,A,N,Fs);
For example:
Fs = 1e4;
% Butterworth filter with 3dB frequency of 1 kHz
[B,A] = butter(10,(2*1e3)/Fs);
[H,F] = freqz(B,A,[],Fs);
plot(F,20*log10(abs(H)));
grid on; xlabel('Hz');
0 commentaires
Voir également
Catégories
En savoir plus sur Parametric Spectral Estimation dans Help Center et File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!