Login
logo
Compare

Crimson: Nonlinear Scaling

06 Jun 2025

Industrial Automation Tech Note 27 - TNIA27

Abstract:

This document describes using Crimson® 3 programming to scale a nonlinear input signal.

Products:

CR Series HMI / G3 Series HMI / G3 Kadet HMI / Graphite® HMI / Graphite® Controllers / Modular Controller / Data Station Plus / ProducTVity Station

Use Case: Scaling a Nonlinear Input Signal

Many applications involve a nonlinear relationship between the input signal and the engineering units that it represents, such as the level of a conical or horizontally mounted cylindrical tank. This document explains how to scale the signal using Crimson programming, as well as replacing the functionality of the obsolete CSINI8L0 and CSINV8L0 modules.

Required Software:

Crimson 3.0 / 3.1

Theory

When the equation that relates the input signal to the engineering units is unknown the alternative is to divide the curve into multiple lines. The more lines that are used the closer the engineering units will be. In order to generate these lines, a series of X (input signal) and Y (engineering unit) coordinates are entered into arrays. The input signal is compared to the elements of the X array in order to find which line segment should be used to calculate the engineering units. Figure 1 is a visualization of the data points used later in the document.

Fig_1.PNG

Figure 1.

Application

1. Create and populate X and Y arrays.

• The arrays can be exposed for entry from the user interface or populated by a program.

 // populate arrays with known data points 

 X[0] = 1;
 Y[0] = 1;

 X[1] = 10;
 Y[1] = 100;

 X[2] = 20;
 Y[2] = 400;

 X[3] = 30;
 Y[3] = 900;


• For a more accurate result, use more data points. Table 1 below shows the inaccuracy of using only 4 points compared to the true equation output:

Input Squared 4 Segment Calculation
1 1 1
2 4 12
3 9 23
4 16 34
5 25 45
6 36 56
7 49 67
8 64 78
9 81 89
10 100 100
11 121 130
12 144 160
13 169 190
14 196 220
15 225 250
16 256 280
17 289 310
18 324 340
19 361 370
20 400 400
21 441 450
22 484 500
23 529 550
24 576 600
25 625 650
26 676 700
27 729 750
28 784 800
29 841 850
30 900 900

                  Table 1.

2. Create the program (CalcVal).

a. Referring to Figure 2, edit the program’s prototype.

Fig_2.png

Figure 2.

1) Return Type – Data Type: Floating-Point or Integer
2) Parameters – Floating-Point or Integer (Input)

b.Write the program code:

 // declare locals
 int i;
 float PVSpan, InputSpan;


 // loop until we get to the point in the arrays where the actual input value is 
 while ((Input > X[i]) && (X[i] != 0)) i ++;


 // if it is right on a point, return the display value with the offset
 if (Input == X[i])
      return Y[i];


 // if it is between points, calculate the display value
 else {

      // calculate PV change in this segment
      PVSpan = Y[i] - Y[i-1];
      // calculate input change in this segment
      InputSpan = X[i] - X[i-1];
      // if there is an input change, calculate the PV (y=mx+b)
      if (InputSpan > 0)
           return (((Input - X[i-1]) * PVSpan) / InputSpan) + Y[i-1];
      else
           return Y[i-1];
 }
 // if the input falls outside of the scaled range, return a high number
 if ((i != 0)&&(X[i] == 0))
      return 99999.1;

3. Use the program as the Source of a tag.

Referring to Figure 3.

a. Create a new Numeric Tag.
b. Change its Source to General.
c. Type in the name of the program, with the name of the source value as its argument: CalcVal(Reading).

Fig_3.PNG

Figure 3.