LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

FFT Power Spectrum weighting mode exponential

Solved!
Go to solution

In the FFT Power Spectrum VI using weighting mode exponential from the link:

https://zone.ni.com/reference/en-XX/help/371361H-01/lvanlsconcepts/average_improve_measure_freq/#Wei...

Weighting is applied according to the following equation.

where Xi is the result of the analysis performed on the ith block, Yi is the result of the averaging process from X1 to Xi, N = for linear weighting, and N is a constant for exponential weighting (N = 1 for i = 1).

 

For exponential weighting, N is a constant. What is its value?

 

0 Kudos
Message 1 of 7
(3,331 Views)

Hi KBerger,

 

That's an excellent question, I agree that the language used there is a bit vague, I'll look into it and see if I can find anything more conclusive. Thanks!

 

0 Kudos
Message 2 of 7
(3,297 Views)

You define the N value. The FFT functions have a count input for the averaging. This is your N.

The higher the N, the lower the weight of each new sample compared to the average (= the less change each new sample will cause on the average).

For instance, when your N is 10 you have Yi = 0.9 Yi-1 + 0.1 Xi, with 100 you have Yi = 0.99 Yi-1 + 0.01 Xi.

 

0 Kudos
Message 3 of 7
(3,279 Views)

dan_u

Thats what it states/uses for linear weighting "N = for linear weighting". Looking for when using exponential.

 

0 Kudos
Message 4 of 7
(3,271 Views)
Solution
Accepted by topic author KBerger

Yes, but N=i doesn't mean it is constant. For linear weighting N = i means N=1 when i=1, N=2 when i=2, ... And averaging stops when i reaches your defined count.

For exponential weighting N is constant (the count you define) which leads to an exponential weighting. The averaging never stops in this case.

 

 

Message 5 of 7
(3,269 Views)

OK Thank you.

From the NI FFT Power Spectrum VIs help, it states for N and weighting:

number of averages specifies the number of averages used for RMS and vector averaging. If weighting mode is exponential, the averaging process is continuous.

 

This could be stated more clearly (to me indicates for exponential N is not used).

 

Message 6 of 7
(3,262 Views)

@KBerger wrote:

From the NI FFT Power Spectrum VIs help, it states for N and weighting:

number of averages specifies the number of averages used for RMS and vector averaging. If weighting mode is exponential, the averaging process is continuous.

 

This could be stated more clearly (to me indicates for exponential N is not used).

 


I agree this can be misunderstood. 

You can post an idea in the LabVIEW Idea Exchange.

 


@KBerger wrote:

OK Thank you. 


There's a Kudos button and you can mark posts as solution Smiley Very Happy

Message 7 of 7
(3,255 Views)