accelerometer filter for noise

[zain1612]

Hi

I am taking data from an accelormeter in real time and displaying it to a user. I am struggling to find a valid way of seperating out the acceleration due to gravity and noise (which I want to display as I want to show a persons hand shaking). I am thinking of using a butterworth filter to achieve this. So I should use a high pass to seperate remove the noise and a low pass to remove the acceleration due to gravity?

Thanks in advance 

[Bob_Schor]

This comes under the heading of "What works well in my situation?".  Often it is helpful to simply "Do the Experiment".  You apparently know something about filters (being a non-engineer, I'm always confusing High Pass with High Cut, for example, and think Butterworth is a particularly good brand of margarine).  Put together a VI with some front panel controls to set corner frequencies, hook up your instrument, hold it in your hand and give it a shake while watching the screen.  Are you pleased with the results?  Does it look like it will give you the kind and quality of data you need for your study?  Tweak the filters and see what seems to work best.

The Virtue of a Virtual Instrument, after all, is how quickly you can replace Capacitors and Resistors (to say nothing of op amps) with absolutely no soldering or drilling for a new switch or dial in an aluminum chassis!  Go Design, Test, and Conquer!

Bob Schor

[zain1612]

I am just struggling to use the butterworth filter in labview. As it an accelometer the data is going to be changing (from 700 to 300 with noise of +-6). I have come up with the following VI but it doesnt seem to be working.