FFT method and preprocessing in QZFM UI v3.2

[Dominic Labanowski]

Hello,

We have some QZFM Gen-2.0 magnetometers we have been using with the v3.2 UI. When we look at the FFT generated in the UI, we see low noise in the 1-10 Hz band (as we'd expect given the datasheet for the magnetometers). When we process the data ourselves, however, we see a substantial increase in low-frequency noise - both for digital data saved using the UI as well as for analog data taken manually using an NI-DAQ. Doing some basic preprocessing (DC subtraction, applying a hanning window) removes some of this added low-frequency noise, but does not get us down to the noise level seen in the UI.

Would you be able to share what processing gets performed on the time-series data to generate the FFT shown in the v3.2 UI?

Thanks,

-Dom

[James Osborne]

Hello Dom,

We use the LabView Spectral Measurements Express VI to calculate the power spectral density (PSD), which has units of pT^2/Hz. We then take the square root of the output to display the PSD in units of pT/sqrt(Hz) in the PSD graph in the UI. We use the settings displyed below and update the PSD every 1000 data point be default. The updates are averaged unless the "Restart Averaging" button is pressed. 

Please feel free to email me directly if you have further questions about this.

Best regards,

James