How to overcome zero offset before integration to get velocity

[Desesperado]

Hi I retrieve acceleration signal from a PCB sensor with a NI DAQCard and try to integrate numerically the signal using the routine (called vi's in Labview enviroment) the program has to integrate but I get a lot of clearly spurious low frequency peaks. How can I get rid off this problem? I would like to kip the concept of integrating within the program instead of using an integrator circuit. Is there a vi that performs this job beter than the one in LabView
Thanks

 

[IRstuff]

How do you know the data is bad?

TTFN

FAQ731-376

 

[Skogsgurra]

There is a forum for NI-related questions. Have a look at

http://www.eng-tips.com/threadminder.cfm?pid=1106

But, I think IR has a very valid point. If the integrator outputs something that you do not expect, it is usually the expectations that are wrong. Accelerometers do not know what you want them to output. They output everything they sense. Including Earth's gravity, earthquake acceleration components and also small vibration amplitudes that you are not aware of existed. An external integrator would do exactly the same thing. If not AC-coupled.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Desesperado]

To IRstuff.I know that because I have compared my spectrum with the output of a Pruftechnik and a IRD spectral analyzers

 

[Desesperado]

To Skogsgurra. Thanks for the tip. I will go to the link you provide. About the possibility ofnthe data being valid I have questioned myself in the same direction. But the results of comparison I have made with spectral analyzers would need an explanation. I've checked noise floor of my card which is very tiny and calibrated the card. Since my device is DC coupled (and has no other option) I will AC couple by soft and/or put a hipass butterworth filter stopping anything below 2 Hz. Thank you and also thank to IRstuff. I'll try the NI link