Spectrum Analyzer Python script Analog Discovery 2.

[Ostin Flores]

Hi all,

I am writting an script to determine the Spectrum Analysis of a signal with the Analog Dicovery Board 2. When using the AnalogIn_spectrum.py I obtain the FFT data and then plot it, it seems like the data comes in as dBV/HZ but I want to convert this value to V/Root(HZ). Is there an easy way to do this?

The following Lines in the AnalogIn_spectrum.py example get the FFT data in Volts (rgdBins1) and convert it into dBV/HZ 

 

for i in range(nBins):

    rgdBins1[i] = 20.0*math.log10(rgdBins1[i]/sqrt2) # to dBV

    rgdBins2[i] = 20.0*math.log10(rgdBins2[i]/sqrt2) # to dBV

 

So what should I do to rgdBins1[i] to convert the data into V/Root(Hz)?

My second concern has to do with a small DC value that is recorded at 0Hz (See picture). What does it represent? I do not have an input voltage to the signal displayed. Your help is highly appreciated.

Software Version 3.20.33 

 

[attila]

Hi @Ostin Flores

V/sqrt(2)/sqrt(WindowNENBW*Frequency/Samples)
FDwfSpectrumWindow(...., byref(WindowNENBW))
FDwfAnalogInFrequencyGet(..., byref(Frequency))

The first BIN is double of the DC value.

 

[Ostin Flores]

Thanks @attila your answer was very helpful, one more thing that is bothering me is if there are any examples on how to do an RMS Averaging. 

[attila]

Hi @Ostin Flores

For averaging calculate the average of multiple frequency domain results.
For triggered repetitive patterns/signals time domain averaging may provide better result.

[Ostin Flores]

This is for a Linear RMS Averaging like the one perfomed in the Waveforms application. I collected various samples and then performed a Root Mean Square of the incomming samples and I do not get the results that you get.

Root Mean Square formula Xrms= sqrt[(X1^2 + X2^2 + X3^2 ... + XN^2)/N]

N is the number of averages

XN is the FFT Data

[attila]

Hi @Ostin Flores

It is linear average of Vrms values.
The latest version adds "RMS average" option.