Making RMS Measurements Correctly
This Knowledgebase article applies to:
- SigQC version(s) 3.07.11 on the following platforms: All Supported
There are several methods RMS can be caculated using SigQC and SigAnalyzer and some of these methods will have mathmatical error associated with it.
There are issues with the PTA/Pulse Overall Measurement – These issues will not be discussed a this time.
Using SigAnalyzer and the M44 the RMS measurement is a function under the channel setup. This RMS value is computed from the autospectrum setup. When you apply a “HANNING” window to the rms the spectrum is corrected for amplitude. This in turn has a effects the scaling of the RMS function. The end result is that the RMS measurement is higher than it should be. The correction factor should be .833333. This can be added as a Linear Units Conversion with the RMS function.
When adding an RMS in SigAnalyzer under the post-processing options then there is no need for correction. If the user desires to add a digital filter to the data you should use the butterworth filter with the desired rolloff. The FIR filter has ripple that will distort the time waveform and contaminate your results.
Calculating RMS as a post-processed function in SigQC.
>> When creating a post-processed function in SigQC you should calculate from a time history function. There are issues in computing RMS from an Autospectrum within SigQC.
>> Use the Butterworth Filter when you wish obtain the RMS of a filtered range of data. The FIR filter is not sharp enough and the ripple assoiciated with a FIR filter distorts the results.
An application technical support document should be generated about RMS calculations in SigQC.