We work with an ultraprecision lathe measuring vibration during machining using a NI Labview acquisition system and 2 PCB accelerometers. Signal from accelerometers goes through a conditioner box, where it is also amplified, then to the acquisition board. This signal is treated in Labview with a bandpass filter (which is supposed to take off all DC) and it�s also sent to a FFT box that shows it has a strong component around 600 Hz .
This signal resembles a sine wave and represents of course acceleration. In order to have displacement we tried to integrate it twice (2 Integral x(t) Labview boxes).
First problem is that the integrated signal becomes very oscillatory, much like a cosine wave superimposing a ramp that switc
hes from increasing to decreasing (and vice-versa) almost randomly (we didn�t expect to have any DC component).
Also, changing �dt� value of de Integral x(t) box in Labview changes our signal completely. What should be the best �dt� value?
We have no experience with that. Has anyone successfully obtained displacement with accelerometers? Your help is highly appreciated.
Thanks
Erick
