Hi
From the spectrum measured with AC coupling and imported, I apply a transformation to get the RIN spectrum of a signal of a DC average given as V_DC. First I convert the frequency to Hz and the ChA to V. I also extract the frequency spacing Df between each point of my spectrum. Then my RIN spectrum is (ChA/V_DC)^2 / Df.
To check if I am doing the right transformation, I sum up all the RIN spectrum values and I multiply by Df, like an integral. Then I squareroot it and multiply by 100 to get the result in %.
By doing this and comparing with the RMS value calculated from 'AC RMS' in the signal window, divided by V_DC and multiplied by 100, I get something slightly different. Summing up the spectrum always gives a result about 1.3 times bigger than using the ratio 'AC RMS / V_DC'. I have tried on several spectra to be sure.
I could understand if I got a ratio of sqrt(2) or 2, but a ratio of about 1.3 is surprising. Am I doing anything wrong in the calculation of the RIN spectrum?
Thanks or any help or suggestion!
Best regards
JC
RIN spectrum vs RMS
RIN spectrum vs RMS
 Attachments

 dataset#1884.psdata
 (60.69 KiB) Downloaded 175 times
Re: RIN spectrum vs RMS
In the meantime, I tried the 'rectangular' windowing (meaning no window applied it seems), instead of the default 'Blackman', and it solved the problem. I have no discrepancy anymore.
So to display a noise spectrum, rectangular is preferrable.
JC
So to display a noise spectrum, rectangular is preferrable.
JC