From 04:00 PM CDT – 08:00 PM CDT (09:00 PM UTC – 01:00 AM UTC) Tuesday, April 16, ni.com will undergo system upgrades that may result in temporary service interruption.

We appreciate your patience as we improve our online experience.

LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

filter out oscillation

I'm trying to learn how to get rid of some noise and high peaks that I'm getting when I'm recording a signal from a flowmeter. I made an example simulating a sine with a noise and using a filter that I guess works fine to eliminate the peaks, however the noise is still there,I cannot use  the  Sample Compression Express VI  because I'm working in another computer that has Labview V. 5 and I read in another thread that shift registers can be use to store the data and then calculate the averages. I'll appreciate if somebody can explain to me how this work. Attached is my little example.
0 Kudos
Message 1 of 5
(2,881 Views)
First thing you could do is to suppress the subtraction in your test vi !
Your filter works perfectly, but of course, there is some phase shift. The subtraction re-introduce the noise. Test it, see what happens !
 
Chilly Charly    (aka CC)

         E-List Master - Kudos glutton - Press the yellow button on the left...
        
Message 2 of 5
(2,868 Views)
Hi,
I would simply try using a butterworth filter or some other filter.  Please let me know if I am missing the question of do not understand the issue. I have attached a VI that shows the use of a filter compared to your current results.
Eric A.
National Instruments
Distributed I/O Product Support Engineer
Message 3 of 5
(2,838 Views)
I made this program which reduce the noise in two ways: by using a filter and by using a moving average smoother. In order to determine the frequencies of the noise and change in an educated fashion the parameters of the filter (i.e low cut-off frequency, stop passband, etc) , I'm trying to analyze the signal in the frequency domain, maybe I'm confuse about the meaning of the frequencies showed in the charts because I was expecting a peak at 1 Hz, that is the frequency at which I'm generating the sine wave, however the scale of the frequencies goes up to 0.5 and shows the peak somwhere between 0 and 0.1, but I also notice that the low cut-off frequency for this low-pass filter should be greater than the value of the peak showed in the charts in order to eliminate the noise. For example, if the low cut-off frequency is 0.02 it shows less noise than when the low cut-off is say 0.125, so I would like to understand what these frequencies mean.
 
Another question is there is a vi to show the estimated frequency peak in the amplitude spectrum chart?
 
I couldn't open the file that Eric sent, my labview is v. 8.0 your is 8.5
 
Thank's for your help
 
Diana
0 Kudos
Message 4 of 5
(2,810 Views)
Hi,
Sorry about posting the VI in 8.5.  I think your current issue is coming from building the waveform.  I have attached a pic of how I edited your code to show how your array is probably closer to what you are expecting.
Eric A.
National Instruments
Distributed I/O Product Support Engineer
0 Kudos
Message 5 of 5
(2,785 Views)