Smoothing Filter - HBM SomatXR CX23-R User Manual

4 Setting up and running a test
Application Note: Piecewise Linear Relationships
In the following example, the desired output of the Signal Calculator channel y is defined as follows, based on the value of
the input channel x .
1. Define the required set of logical channels as follows:
s1: x>100
s2: x>50 && x<=100
s3: x>0 && x<=50
s4: x<=0
2. Define the required set of arithmetic channels as follows:
y1: 2.1*x+100
y2: 2.2*x+90
y3: 2.3*x+80
y4: 2.4*x+70
3. Define the final channel y as follows:
y: y1*float(s1) + y2*float(s2) + y3*float(s3) + y4*float(s4)
NOTE
It is not necessary to define intermediate variables and it is less efficient from a processing point of view
when intermediate variables are not used more than once in the set of computed channels. However,
intermediate variables have been used above to clarify the general approach.
60
4.4.14 Smoothing filter
The Smoothing filter computed channel generates an output that is a smoothed
representation of the input without generating any phase lead or lag.
NOTE
The Smoothing filter can result in loss of data significance if not used properly. It
is provided primarily for digital pulse counter inputs. In general, it should not be
used for analog input channels with digital anti-aliasing filters.
Filter length: Specify the filter length as an odd number between 3 and 201
samples. The filter is a simple boxcar filter where each output sample is the linear
average across the filter length. For example, for a filter length of five, the filter
averages the current sample, the two samples before and the two samples after.
Note that the channel backfills the initial output samples with the first fully filtered
a4275-8.0 en HBM: public CX23-R and EX23-R