LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Identifying 2nd and 3rd Loop in a Graph

Solved!
Go to solution

Hello again Dave,

I understood why the Y axis is from 0 to 36.7097 but for the Y axis I am expecting to get values from 1460 to 1620. If you check my other graphs in the file I've attached for you,  you will see the exact range I am looking for.

Add 273 on whatever Celcius said.
-Kelvin
0 Kudos
Message 21 of 33
(1,704 Views)

Then it's time to use "Array Max & Min Function" again Smiley Wink

Look at my suggestion:

 

Expand Range.png

 

This sets the X- axis range to a range which is [expand range] higher than the distance of the 2nd and 3rd peak.

 

Greets, Dave
0 Kudos
Message 22 of 33
(1,697 Views)

Okay, that is a very intelligent solution Dave but nothing has changed in the graph. I mean scale is still in 10000 something. Because the current position of the second and the third peak is already in that range. So, eventhough when we would like to isolate the only needed parts range will still be in the scale of 10xxx.

Add 273 on whatever Celcius said.
-Kelvin
0 Kudos
Message 23 of 33
(1,693 Views)

OK, got it. Use "Array Subset" from the array palette to extract the interesting part of the array. Then you have to adapt the cursors Smiley Wink

Greets, Dave
0 Kudos
Message 24 of 33
(1,688 Views)

I will check it tomorrow and let you know about it Dave.

Thanks again for your kind effort! Appreciate it.

Add 273 on whatever Celcius said.
-Kelvin
0 Kudos
Message 25 of 33
(1,686 Views)

Hello daveTW, it is me again! I have one issue that I think you might help of. daveTW, I have to determine the distance which I have marked as "X" in the image. For your reference, I have attached the program that is in 2014 version for your reference. I would like to know if you have any suggestions about this case.

Add 273 on whatever Celcius said.
-Kelvin
Download All
0 Kudos
Message 26 of 33
(1,629 Views)

Here's a few more thoughts which might help:

1) The reason the X scales aren't correct are that the graphs are not showing the correct offset and multiplier.  You can do something like this to make them show correctly (this is equivalent to creating an X array as in your original VI, but much simpler!).

GraphScale.png

 

2) To compute the distance marked X, then I think the easiest is to subtract a Median Filtered waveform from the data like this:

SubMedianFilter.png

 

3) The computation of the peak locations using the peak detector looks like a good approach, but I'm not sure if your calculation of W is correct - the code is finding a FWHM of the maximum value, but that conflicts with your original figure which shows W as the width of the first peak - which should lie inside the second and third peaks.  You might need to be more specific about what the value is that you want to calculate, but one option might be to look for the points of maximum derivative either side of the first peak, and take the width between those.  This gives a W=0.322 shown by the green lines on the graph below.

Calculations.png

 

0 Kudos
Message 27 of 33
(1,606 Views)

That was a great answer GregS. But unfortunately I am not able to open the .PNG file. Can you upload it again please? File seems broken.

Add 273 on whatever Celcius said.
-Kelvin
0 Kudos
Message 28 of 33
(1,598 Views)

Hello again GregS,

I am now able to open the graph and control it. In the program you send I just didn't get, why did you choose 250 as left rank input parameter in the Median filter? Could you please explain me briefly if there is any reasion lies behind that?

Add 273 on whatever Celcius said.
-Kelvin
0 Kudos
Message 29 of 33
(1,581 Views)

With the Median Filter, the idea is that you use a rank that is approximately twice as large as any small deviations (such as the spike you want to measure).  Then the filter will give you the underlying trend, which you can then subtract from the signal.  So I got 250 by seeing that the spike is about 100 pixels wide - 200 or 300 will probably work just as well.  The right rank is the same as the left if you don't separately wire it.

 

Were you able to open the snippets?  You will probably need to open the full-size image and save that to disk, then drag it into a LabVIEW block diagram.

 

Any thoughts about measuring the width of the main peak?

0 Kudos
Message 30 of 33
(1,565 Views)