Correcting DC offset in AC Coupled Accelerometers - Solution more so than question

[LabViewer35242]

I'm acquiring data on four PXI-4472B cards, and all are set to AC coupled in MAX.  I'm using the "AI Accelerometer" create channel VI and noticed that my channels VDC doesn't equal 0v like I would expect with AC coupled channels.  I'm decided to use the "Amplitude and Levels Express VI" to get the DC offset, and subtracting my waveform array by this.  Sound right?  Hopefully it is, and anyone else with this problem can use this solution.

[LabViewer35242]

LabView VI is atttached:

[LabViewer35242]

Well it works for simulated data, but not real data.  Any thoughts?

[majoris]

If you have a handy sma dc block, that might be the first thing to try.

In terms of software, wouldn't a high pass filter or simply taking the derivative of the signal get you close to what you want?

No experience with that card, sorry!

[LabViewer35242]

Reply, yes.  I have a highpass filter at 1Hz on my actual program, but not on the mock program that I attached.  I think I just need to bump it up to 2Hz.

[johnsold]

1. The spectrum of white noise is a constant.  If you have enough noise in your signals, that may elevate the baseline of the spectrum.

2. If you acquire fractional cycles of your signal, you will have a DC component, although that does not offset the spectrum.

3. Why do you ADD twice the DC component to the signals before calculating the spectrum? The usual approach is to subtract the mean.

Lynn