AnsweredAssumed Answered

FFT Window Issue and differences

Question asked by vj on Sep 10, 2014
Latest reply on Sep 17, 2014 by algoss

Like • Show 0 Likes0
Comment • 4

Hello,

I am working on FFT measurements using DSA -X 91604A series oscilloscope and was analyzing the different FFT windows -Flat,Han,Ham,Rect & Blackman Harris which the scope provides.

While comparing i see that with the same setup the measured Magnitude of a spectral peak for a *blackman harris window is
8-10dBm less compared to the other windows available.*

Is it a bug? Can you please provide some insight?

Regards,
VJ

Edited by: VJ on Sep 10, 2014 11:19 AM

No one else has this question

Outcomes

4 Replies

ksmith Sep 12, 2014 2:13 PM
Mark Correct
Correct Answer
Homework time!

Spectral Analysis Using a Deep-Memory Oscilloscope Fast Fourier Transform (FFT) AN 1383-1
http://literature.cdn.keysight.com/litweb/pdf/5988-4368EN.pdf

Most measurements will require the use of a window such as the Hanning or Flattop windows. These are the appropriate windows for typical frequency analysis measurements. Choosing between these two windows involves a trade-off between frequency resolution and amplitude accuracy.
Hanning
•     Advantage: Good Frequency Resolution
•     Disadvantage: Less Amplitude Accuracy
Hamming
•     Advantage: Good Frequency Resolution (slightly better than Hanning)
•     Disadvantage: Less Amplitude Accuracy
Flattop
•     Advantage: Good Amplitude Accuracy
•     Disadvantage: Less Frequency Resolution
Rectangular
•     Advantage: Good Frequency Resolution, Good Amplitude Accuracy (For self-windowing waveforms)
•     Disadvantage: Possible Leakage Effects
Blackman-Harris
•     Advantage: Good for signals with a strong interference component that is fairly distance from the frequency you want to see, good general purpose window

Disclaimer: For more reliable response, you should consider calling your local Agilent Technical Call Center. The Agilent Forums are monitored on an "as available" basis, and aren't necessarily the fastest way to get technical questions answered.
Like • Show 0 Likes0
Actions
- algoss @ ksmith on Sep 15, 2014 7:24 AM
  Mark Correct
  Correct Answer
  Remember that every window has the effect of modulating the waveform, and spreading the energy from the peaks into the sidebands. The B-H window does that the most, and while the peak moves down, the sidebands move up. It also has a large effect on the Resolution BW.
  
  Al
  Like • Show 0 Likes0
  Actions
  - vj @ algoss on Sep 15, 2014 10:46 PM
    Mark Correct
    Correct Answer
    Thank you for the response.I agree that each window has its own advantage & disadvantages.
    And the Blackman Harris would reduce the amplitude but my concern is that it has reduced the peak by 10dBm which could make a measurement test fail with respect to a spec limit.
    
    So i m concerned whether the result i get through Blackman Harris window is acceptable or is it some issue with scope?Anybody have tried using this window in Agilent scope?
    
    I want to understand if the amplitude accuracy is good when using B-H window for FFT measurements in this condition.
    
    VJ
    
    Edited by: VJ on Sep 15, 2014 11:44 PM
    Like • Show 0 Likes0
    Actions
    - algoss @ vj on Sep 17, 2014 5:22 AM
      Mark Correct
      Correct Answer
      "I want to understand if the amplitude accuracy is good when using B-H window for FFT measurements in this condition."
      
      Then you need to learn some FFT theory and run some experiments with other tools, such as MATLAB or any of the free alternatives, or even Excel. Save the captured trace file to a csv, import it into your favorite program, and play with the various windowing options.
      
      Spreading of the peaks means that some of the displayed energy moves to other buckets, thus reducing the energy shown in the higher peaks and increasing the energy on the lower peaks. It's like running a smoothing function over the 'real' FFT (which doesn't exist, by the way). The total energy included in all of the buckets should be the same for all windowing functions (except for the loss caused by the actual window itself).
      
      It appears to me that the B-H window is not the right one to use if you are worried about amplitude accuracy.
      
      Al
      Like • Show 0 Likes0
      Actions

FFT Window Issue and differences

Outcomes

Related Content

Recommended Content