Configuring Moving Average Filter Lengths for Decimated Signal Processing
Matching moving average filter length to the decimation factor eliminates alias leakage by placing transfer function zeros directly on downsampling fold points.
Sinc Response
Mathematical Filter Characteristic
Frequency response profiles define the magnitude of a digital signal across a spectrum based on a rectangular time domain window. A sinc response is the natural result of averaging samples over a fixed period, creating a profile with a main lobe and a series of diminishing side lobes. This shape is fundamental to the behavior of delta-sigma analog to digital converters used in high resolution weighing and temperature sensing.
Sampling Notch
Precise null points exist where the gain drops to zero at specific frequencies. By timing the sampling window to match the period of line noise, the sinc response can effectively eliminate 50 or 60 Hertz interference. These notches are located at integer multiples of the reciprocal of the integration time.
Passband Droop
While the notches are beneficial for noise rejection, the curved shape of the main lobe causes a gradual loss of signal strength at higher frequencies. Digital compensation filters are often applied after the sinc response to flatten the output within the desired bandwidth. This two stage process allows for both high noise rejection and an accurate representation of the signal.
Settling Time
The duration of the averaging window dictates the speed at which the filtered output can respond to a sudden step change in the input signal. Longer windows slow the response.