LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

autocorrelation with labview

Sorry, I forgot to add, I cross checked the diadem output to another more sophisticated software and they produce the same curve.  

 

The bottom line is that I want labview to produce the same autocorrelation curve as the other to softwares. 

 

Thanks

Ay7973 

0 Kudos
Message 21 of 25
(1,117 Views)
I noticed that there are three normalization methods that can be set in the autocorrelaion subVI. Have you tried all three settings?
0 Kudos
Message 22 of 25
(1,108 Views)
yes i have tried all three methods of normalization, The unnormalized and biased modes both produce essentially the same curves, and the unbiased produces something different but still not what I am looking for.  
0 Kudos
Message 23 of 25
(1,104 Views)
Ay7973-

Please go ahead and post your code.  This may have to do with the way the data is being parsed from your text file.  Looking over your text file, it has other numbers associated with each point that may be affecting it. 
Regards,

Mike S
NI AE
0 Kudos
Message 24 of 25
(1,079 Views)

Are you SURE your data produces an exponential corrolation decay?  The light intensity scattered from a solutions of mono-disperse spheres should produce a single exponential decay under autocorrolation however if you have noise in your signal or a homodyne component, such as light reflected off the contanter walls, you may not get the results you expect.  Keep in mind that the autocorrolation function of a sine curve is, in fact, a sine curve.  If you have periodic noise in your signal that is of significant magnitude you will not get a decaying autocorrolation curve...it will be periodic as well.  Make sure your DATA consist of the light intensity of light scattered only from your fluid sample and not contaminated with surface reflections.  You might want to image the light through a pinhole or something to get rid of any background light before sending it into the pmt or whatever you are using to detect.  Make sure there is no 60htz noise in your signal.  You might run an FFT on your raw data and see if there is a strong peak at 60htz or any other frequency for that matter.  Very often DLS issues arise, not from the software or equipment, but simply due to lack of sample purity or light/electronic noise contamination of the signal.

 

By the way, you might also try looking up the wiki for autocorrolation function.  It would be instructive to write a program (you can do it in labview) to calculate an autocorrolation.  If you do it in labview, however, it will be crazy slow.  Then you will have a better idea of exactly what the function labview calls is doing and also it might allow you to check and see if, in fact, your data DOES produce an exponential decay.

 

Kevin

0 Kudos
Message 25 of 25
(632 Views)