which kind of digital filter/algorithm is suitable for load cell under the influence of external noise?

Hi Greetings!

Consider a load cell system measuring weight under the influence of external non-periodic noise present at many low frequencies (as low as upto 0.04Hz!). The desired data is at DC. Making a sharp low pass digital filter makes response very very slow. Is there any way/ algorithm/ filter processing, which can give me the desired response in lowest transient time?? Please suggest.

Thanks, SSP

Answer 1

[Star Strider]

If the desired signal is d-c, it is not going to change with time (the definition of d-c), so just take the mean of your signal.

That aside, you did not say what sort of filter you are using. A Chebyshev Type II filter, doing the filtering with the filtfilt function, would be my choice.

From your description, consider redesigning your data acquisition hardware to use a hardware lowpass Bessel filter with a 0.04 Hz cutoff. Even that might not be easy to build, but it would be faster than a long digital filter. A Chebyshev Type II hardware filter would be my second choice, depending on whether phase distortion was an important consideration.

[Star Strider]

ssp’s ‘Answer’ moved here:

Thank you! It helped me a lot! Zero phase filter makes the response significantly faster. I found that filtfilt function has filter as its subfunction.

However y=filter(b,a,x) seems easy to implement, but I am stuck at implementing y=filter(b,a,x,zi). What would be the new sequence of the following sequence after applying initial condition zi?

y(n) = b(1)*x(n) + b(2)*x(n-1) + ... + b(nb+1)*x(n-nb)
                 - a(2)*y(n-1) - ... - a(na+1)*y(n-na)
Would you please guide me?

Thanks! SSP

[Star Strider]

My pleasure.

I have very little (essentially no) experience with filter, since I always use filtfilt for its obvious advantages. It also internally implements the initial condition calculations.

I strongly recommend using filtfilt. I am not familiar enough with filter to help you with its intricacies.