34 MS6000-EU-EN V1.4 8/12
FFT Window Selection: The Math FFT function has Five FFT Window options. There is a trade-off
between frequency resolution, Spectral Leakage, and amplitude accuracy for each type of the window
choices. Determine which one to choose according to the desired object to be measured and the source
signal characteristics.
Window Measurement Characteristics
Hanning Random Waveform
Good frequency resolution, Fair amplitude accuracy, and
Good Spectral leakage.
Flat Top
Sinusoid
Waveform
Poor frequency resolution, Best amplitude accuracy, and
Good Spectral Leakage.
Rectangular
(None or Boxcar)
Pulse or Transient
Waveform
Special-purpose window applicable to discontinuous
waveforms. Best frequency resolution, Poor amplitude
accuracy, and Poor Spectral Leakage.
Bartlett Random Waveform
Good frequency resolution, Fair amplitude accuracy, and
Fair Spectral Leakage.
Blackman
Random or Mixed
Waveform
Poor frequency resolution, Best amplitude accuracy, and
Best Spectral Leakage.
4.3.1.4 FFT Aliasing
Problems occur when the time-domain waveform acquired by the oscilloscope contains frequency
components higher than the Nyquist frequency. The frequency components above the Nyquist
frequency will be under sampled and displayed as lower frequency components that ‘fold back’ from the
Nyquist frequency. These erroneous components are called aliases.
4.3.1.5 Eliminating Aliases
To eliminate aliases, use the following methods.
Click the TIME/DIV key to set a faster sample rate. Since the Nyquist frequency increases as the
sample rate is increased, the aliased frequency components will be displayed correct. If too many
frequency components appear on the screen, use the FFT Zoom option to magnify the FFT
spectrum.
If there is no need to observe the frequency components above 20MHz, set the CH Bandwidth
Limit option to Limited.
Filter the signal input and limit the bandwidth of the source waveform to lower than the Nyquist
frequency.
Identify and ignore the aliased frequencies.
Use zoom controls and cursors to magnify and measure the FFT spectrum.
