I have some biomedical signals. If I calculate their magnitude spectrums, I see power-line interference at 50 Hz. I was excepting a harmonic at 100 Hz, but it turns out there is none. However, there is a harmonic at 150 Hz. Is this common or should I worry that I have done something wrong?
3
$\begingroup$
$\endgroup$
-
$\begingroup$ Are you sure that your frequency scaling in Fourier domain was right? $\endgroup$ – Eddy_Em Aug 7 '13 at 22:21
-
$\begingroup$ Yes I think so. There's also harmonic at 250 Hz, but not 200 Hz. $\endgroup$ – argh Aug 7 '13 at 22:25
-
$\begingroup$ If you are confident that you have scaled your frequency axis correctly, then what you see is what you get. $\endgroup$ – Tarin Ziyaee Aug 7 '13 at 23:10
-
$\begingroup$ Maybe this might help: ibiblio.org/kuphaldt/electricCircuits/AC/AC_10.html I see in the page that you can miss a few harmonics. Just went through it. $\endgroup$ – Sudarsan Aug 7 '13 at 23:12
2
$\begingroup$
$\endgroup$
Yes, this is common. Common non-linearities (such as op-amp clipping) preserve the symmetry of the upper and lower section of the waveform; and as such, they do not introduce even harmonics - just odd harmonics.
-
$\begingroup$ non-linearities with odd symmetry produce odd harmonics, and even symmetry produce even harmonics (for sine wave input) $\endgroup$ – endolith Aug 28 '13 at 2:24
