two small peaks besides the fundamental frequency

[shanmei]

I have a signal, and I do the SFDR simulation. I found two peaks as the figures show near the fundamental frequency. What's that? Thanks.



The original time domain signal is shown as bellowing figure, it changes its amplitude at around 4.3e-4 second. Can the changing amplitude cause the two peaks on the freuency domain? Thanks.

 

[dick_freebird]

Probably mix-tones, their offset from the fundamental
is the modulating frequency. Looks like ~ 5kHz. Also
looks like you want to play some with your resolution
bandwidth, too much rounding of what ought to be
narrow.

 

[FvM]

The original time domain signal is shown as bellowing figure, it changes its amplitude at around 4.3e-4 second. Can the changing amplitude cause the two peaks on the frequency domain? Thanks.

A single amplitude step doesn't transform to discrete spectral lines. But we need to know the FFT window used when calculating the spectrum for an exact analysis.

Performing the SFDR measurement with an amplitude step seems questionable in any case.

 

[KlausST]

Hi

The FFT output is always like the signal (in time domain) is repeating.
Imagine an identical signal is appended to the current signal (window).
Now it seems that the step is repeating.

The signal then shows like an amplitude modulated carrier.
The modulation signal (here it looks like a square wave) with it´s frequency components will appear right and left to the carrier frequency.

Maybe this is what you see in your picture.

If the bottom window is the complete FFT window, then we see a total time of 800us. This gives a repeat frequency of 1250Hz.
But the difference frequency seems to be around 8kHz...so it doesn´t fit.


Klaus