DC offset in analog inputs

[imperial-aero]

Hi,

I'm reading analog inputs from 8 microphones and I find that they all have different mean values (i.e. the waveforms are displaced vertically by different amounts as shown on the attached pictures). I suspect that this might be due to a DC offset of the signals.

In my data analysis, I carry out the calculations on the Y-axis readings at a particular time level. So since the waveforms are displaced by different amounts, at a particular time level they will all have different Y-axis readings (whereas the actual amplitude of the signals are the same). This leads to inaccurate readings.

Is there anyway in LabVIEW where you can adjust the readings so that their mean is at zero?  Or somehow make the changes to my results after I obtain them?I know that in LabVIEW there's an option where you can just display the analog part of the signal. But I'm not entirely sure whether that will be a solution to my problem, as it may deteriorate my readings.

Thank you.

[Pnt]

Have you tried "AC & DC Estimator.vi" ?

See Labview help for details

[imperial-aero]

I couldn't find the mentioned VI anywhere in the Example Finder. I'm using LabVIEW 8.5.

[Eagle Eye]

High Imperial,

It looks from your pictures that you DO have a different offset for each input.

Offset is relative, usually referring to 0 level.

What happens to your calculations when you use the "with signal" value and subtract the "no signal" value?
I suspect ti should be correct then.
Please let me know if this helped you.

Eagle Eye

[imperial-aero]

Thank you for the reply.

I think if I did subract them, then I should get the signal I want. But I am trying to avoid it, simply because then I'd have to take one set of results with no signals and then subract that from every other test I do with signals (which I reckon will be around 60 tests).

I was wondering whether there's a software solution for this. I guess I could always obtain the DC part of the signal within LabVIEW and then obtain the AC part from there. But I'm not sure what effect will this have on the signal, whether it's just going to shift it to zero line or whether it's going to attenuate the signal.

[Pnt]

---

imperial-aero wrote:

I couldn't find the mentioned VI anywhere in the Example Finder. I'm using LabVIEW 8.5.

---

Try to search the Labview folder....

[PAG]

Hello,

I already had such problem in the past.

It seems to me that the better way is effectively to substract the DC part. You can have the DC part making the average of your signal, or using Basic Averaged DC-RMS.vi for exemple:

Signal Processing => Waveform measurements => Basic Averaged DC-RMS.vi

1) Acquire your channels without signal

2) Calculate DC part of each

3) Remove DC part from channels without signal (you can just use the "substract" function with 1 waveform and 1 DBL as inputs).

4) Acquire your channels wtih signal

5) Remove DC part from channels with signal

6) Possibly check that DC part is the same with or without signal.

Have a nice day.

Pierre-Alain