The input function is a 60Hz square wave.
From Fourier Transform I know that the frequency on the spectrum is 60,180,300....(2k-1). But there's a series of 120,240,360...(2k) signals appear. 
I've googled but all I found is Fourier Transform issues, not Fast Fourier Transform. I'm wondering why there are some 2k terms appearing on the spectrum.
Thanks!
$\begingroup$
$\endgroup$
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
2
Looks like aliasing.
Your 360 Hz is probably not a real component at 360 Hz but it's the 9th harmonic (540 Hz) that aliases back to 360 Hz at a 900Hz sample rate.
That's easy to test: change the sample rate to, 940 Hz and see of the 360Hz moves to 400Hz or stays put. It moves to 400Hz, it's clearly aliasing.
-
$\begingroup$ Why does aliasing happen? I know if it violent the Nyquist Theorem it will appear. But the sampling rate is 10 times the input frequency. $\endgroup$ Oct 6, 2019 at 16:08
-
$\begingroup$ A square wave contains an infinite amount of harmonics. The Nyquist criteria applies not just to the fundamental frequencies but to any frequency in the signal which is infinite for a square wave. All harmonics that are below the Nyquist frequency will be sampled correctly. All harmonics above the Nyquist frequency will alias. $\endgroup$– HilmarOct 6, 2019 at 18:50