Configuration and operation
akYtec GmbH · Vahrenwalder Str. 269 A · 30179 Hannover · Germany · Tel.: +49 (0) 511 16 59 672-0 · www.akytec.de
16
4.8.3 Signal filtering
To protect against electromagnetic interference, the module is equipped with a digital low-pass filter. Digital
filtering is carried out in two stages.
1. At the first stage, the pronounced "dips" and "overshoots" are filtered out from the useful signal. The
difference between the last two measurements is compared with the
Filter bandwidth
parameter. If the
difference exceeds the bandwidth, the measurement is repeated with the doubled bandwidth. If the new
measuring confirms the correctness of the previous one, its result will be taken as a new stable state to
which the bandwidth reduced back to the set value will be applied. If not, the result will be discarded.
This algorithm protects the input from the single-pulse interferences often generated by industrial plants.
The
Filter bandwidth
parameter is specified in measurement units for each input. Decreasing the
bandwidth improves the interference immunity of the input, but leads to a slower reaction to rapid chang-
es of the input signal. Therefore, when the interference level is low or when working with rapidly chang-
ing processes, it is recommended to increase the filter bandwidth or disable the filter stage by setting the
parameter to 0.
When working in conditions of strong interferences, it is recommended to decrease the bandwidth to
eliminate their influence on the operation of the module.
2. At the second stage of filtering, the signal is smoothed (damped) in order to eliminate electromagnetic
noise components. The main parameter of the damping filter is the
Filter time constant
.
Filtering takes place according to the formula:
𝑆𝑆
=
𝑆𝑆
𝑛𝑛
∗ 𝑇𝑇
+
𝑆𝑆
𝑛𝑛−1
∗
(1
− 𝑇𝑇
)
where
S – stored signal value
S
n
– signal value measured at the last sampling
S
n-1
– signal value measured at the previous sampling
T – damping factor
𝑇𝑇
= 1/(
𝐾𝐾
10
+ 1)
K –
Filter time constant
The
Filter time constant
is set in seconds for each input. The increasing of the filter constant improves the
noise immunity of the input, but at the same time increases its inertia i.e. slows down the reaction to rapid
changes in the input signal.
The second filter stage can be disabled by setting the parameter to 0.
4.8.4 Sensor curve correction
The characteristic curve of the sensor can be corrected with two parameters
Curve offset
and
Curve slope
.
The
Curve offset
can be applied in the following cases:
−
to compensate the lead wires resistance in a two-wire RTD connection
−
to compensate the R
0
drift of a thermocouple
The
Curve slope
can be used, to compensate the errors of the sensors themselves or inaccuracy of the
shunt resistor. The
Curve slope
is set in dimensionless units in the range 1…10.
4.8.5 Measured input values
Measurement results of the input signals are presented in two formats (Tab. 4.6):
AIn REAL
– 4-byte floating point value
AIn INT
– 2-byte integer value
where
