All manner of phenomena including temperature can be measured by converting them into electrical signals with a sensor. The finer or smaller the phenomenon, the smaller the electrical signals that can be received from the sensor, making them more easily susceptible to the effects of a variety of noise which is a factor inhibiting accurate measurement. Given that, we will herein describe the fundamental principles and operation of the functions of industrial recorders, taking as an example those of the μR series (filtering, moving average), which reduce or suppress the effects of noise that contaminates input signals.

The µR series industrial recorders (pen model) are equipped with the filter function designed to reduce or suppress the effects of noise that contaminates input signals. This filter function measures input signals and subjects them to first-order lag filtering (damping) with firmware processing upon final output of the measured value. It acts in the same way as a low-pass filter composed of a resistor and capacitor at the final output stage of a recorder. (see Figure 1)

First-order lag refers to the output of a response curve as illustrated in Figure 3, which occurs when step input as shown in Figure 2 takes place. The T on the time constant in Figure 3 refers to the time taken (in seconds) to reach an output value of 63.2%. This means that lengthening time constant T requires a prolonged period to reach 100% output causing the filter function to be applied even more strongly. In the μR series, it is possible to choose from a 2, 5, or 10 second time constant.

By applying the filter function in this way, it is possible to reduce or suppress unstable or contaminated input signals caused by noise, and achieve smooth measurement results.

The µR series industrial recorders (dot model) are equipped with the moving average function as standard, that suppress the effects of noise that is riding onthe signal. The average value of the m most recent values acquired at the scan interval is used as the measured value of the channel.

The number of moving-averaged data points (m) can be set in the range from 2 to 16. The figure below shows an example indicating the operation of the buffer for the moving average computation when the number of moving averaged data points is set to 5.


Using the moving average enables you to reduce or suppress unstable or contaminated input signals caused by noise, and achieve smooth measurement results.