Filter the signal waveform

[camerond]

Well, you grabbed an Express Filter VI and set the parameters for a 50 Hz highpass filter... If you set your filter to 50 Hz, that means you will have 1/2 (or 1/sqrt(2), I don't remember whch at the moment) the amplitude of the 50 Hz component of the input coming through. So, you still see it. Also, for the conditions you chose, the rolloff is pretty shallow, considering that your signals come in at about 200 Hz. I'd suggest to try a 16-pole Butterworth highpass filter at about 100 Hz cutoff and see what the output looks like. There's nothing magic about 16 poles, it just should have a much sharper cutoff.

Now, if that works better, I'd suggest digging into at least the LabVIEW help files about filters, then finding a book (or even Web content, but you may find you have to go several places to find something in sufficient detail to be of practical use) to help you decide just what lengths you have to go to in order to minimize this background. You can't just throw a filter in and expect it to work unless you have an idea what the filter's general characteristics are in the first place.

FWIW, since you're talking about chips coming off a drill, I bet you've got about as good an idea as you are going to of the transients which occur when the chips pass the detector. They won't be uniform in either size or shape, so each one will probably have its own unique peak profile. Just because you can make the baseline flatter, don't expect the peaks to be larger. You still haven't addressed the fast noise signal.

Cameron

[camerond]

With your huge signal and low noise level, adaptive noise cancelling is like shooting a mosquito with a cannon. It'll be real expensive (in terms of effort required) and negligably better than standard filtering.

Cameron

[rahulmore]

Dear Cameron,

Please refer below thread. This was generated when I started working on this signal.
http://forums.ni.com/t5/LabVIEW/identify-the-disturbances-in-a-signal/m-p/2446038#M751878
I will try your suggested option and let you know how effective it is.

As shown in the below image, the real signal is the continuous reflection of the light ar the rake face of cutting tool giving a negative spike.
And when any kond of chip passes over the tool rake face, the complete signal is not reflected to the detector hence there  is some kind of disturbance in detected signal. I have shown one such disturbance in the below image.
I hope this will help to understand what I am looking for.

- Rahul