Calculate rms value of an an array

[Kaly]

Somehow, I can't find an 'rms' function in Daqfactory. Is there truly no way to calculate the actual rms value of an array? The array may not be a sine wave so I cannot simply divide by the square root of 2. How would one go about obtaining the actual rms value of an array?

[AzeoTech]

There is no built in function for calculating RMS, but if you can point me to a page with the formula you'd like to use, I can probably show you how to write it in DAQFactory.

[SteveMyres]

I thought 'Root Mean Square' WAS the formula?

[AzeoTech]

Yup, but I'm not sure how well it will work with quantized data.  The accuracy is going to depend on the sampling interval, and some sort of curve fit might be helpful. 

But ignoring that and assuming you have a channel / array "X" with your data, the RMS calculation in DAQFactory would simply be:

mean(x^2) ^ 0.5

 Remember, doing x^2 returns an array the same size as x, where each value of x is squared.  Mean() takes the mean of whatever array is provided to it.  And ^ 0.5 raises whatever is to the left to the 1/2 power, which is the same as square root.

 

[Kaly]

OK So I think I have it all figured out. I attached a .ctl file showing how I did it. The file uses an original daqfactory example as a starting point.

Does all seem well?

Thanks.

[AzeoTech]

I don't think so.  You were on the right track at the bottom, but the field labelled RMS is:

max(captureddata) / 2^0.5

when it should be:

mean(captureData^2) ^ 0.5

 

[Kaly]

The field labeled RMS is from the original Daqfactory example. That field calculates RMS value by first determining the peak. That would only work for a true sine wave. I added the 'TrueRMS' field (and some other things). The 'TrueRMS' field should show the true RMS value for any wave shape. Accuracy depends on the field labeled 'capture length'.