National Instruments NI Spectral Measurements Toolkit - Manual (Page 31)

© National Instruments Corporation 31 NI Spectral Measurements Toolkit User Guide
Peak Search and Amplitude/Frequency Estimation
The SMT Spectrum Peak Search VI uses interpolation to precisely locate
frequency peaks in the amplitude or power spectrum and to estimate the
amplitude of each peak. You can enter a real spectrum in any units or
scaling. You can also specify whether to locate a single maximum peak or
multiple peaks that exceed a specified threshold amplitude.
A single frequency tone appears in the frequency domain as a sampled
version of the window the SMT Spectrum Peak Search VI applies to the
input signal. If the VI does not apply a window, a finite sample size is
equivalent to rectangular windowing. If you specify which window the VI
applies to the input signal, the VI uses a curve-fitting algorithm on the three
points around each detected peak to estimate true frequency and amplitude.
The amplitude/frequency estimation method works best on an averaged
power spectrum because the averaging reduces the noise and provides a more
consistent measurement. Figure 11 illustrates the curve-fitting algorithm.
None of the three FFT bins falls exactly on the frequency peak, but the VI
uses the known frequency response of the applied window to estimate the
true peak location, which is offset from the maximum FFT bin by
Δ.
Figure 11. Amplitude/Frequency Estimation Algorithm
1 Amplitude
2 Window Shape
3 Threshold
4 Two FFT Bins
1
3
4
2