Linear Scaling SM 331 Thermocouple Values to HMI S7-300 Reference

Overview

When a Siemens S7-300 SM 331 analog input module (order number 6ES7 331-7KF02-2AB0) reads a thermocouple, the value placed in the Process Image Word (PIW) is a raw integer count. To display a temperature in degrees Celsius on a WinCC HMI, the raw count must be converted to an engineering unit. The conversion is a two-point linear scaling operation that maps the module's raw range to the process variable range defined by the transmitter (lower engineering limit UGR and upper engineering limit OGR).

This reference documents three scaling paths used in production: (1) an in-line STL function with literal constants, (2) a parameter-driven function block backed by a data block (DB) that holds the upper and lower range limits, and (3) WinCC linear scaling applied directly to the HMI tag. The math, the DB layout, and the integration points for TIA Portal V20 and STEP 7 V5.x are covered, with a verification procedure, a troubleshooting matrix for common field errors (including the 1.4x error introduced by selecting the wrong sensor type), and notes on S7-1200, S7-1500, and ET 200SP equivalents.

Module Identification and Key Specifications

The 6ES7 331-7KF02-2AB0 is an 8-channel analog input module from the SIMATIC S7-300 family. The order number suffix "-2AB0" is significant: the -2AB0 revision supports the 0-10 V, plus/minus 10 V, plus/minus 5 V, 1-5 V, 0/4-20 mA, RTD (Pt100, Pt1000, Ni100, Ni1000), and thermocouple (Type J, K, T, E, R, S, B, N, U, L) measuring modes selectable per channel group in the hardware configuration. Earlier -0AB0 and -1AB0 revisions do not support the full thermocouple type list and have different default ranges.

Hardware Configuration and Raw Value Ranges

1. Insert the SM 331 in the S7-300 station at the correct slot (slot 4-11). Open the device configuration and double-click the module to open properties.
2. Open Properties > Analog Inputs and select the measuring type per channel group. Each group of 2 channels shares a measuring type. For one thermocouple on channel 0, select the group, then pick the thermocouple type (for example "TC-Type K" or the linearized variant "TC-K linear").
3. Set the reference junction compensation: None, Internal, or External via Pt100. For most field installations with thermocouple-grade wire run directly to the front connector, Internal is used.
4. Select the temperature unit (deg C or deg F) and the resolution (0.1 deg C or 0.01 deg C per count).
5. Select the integration time: 2.5 ms, 16.67 ms, 20 ms, or 100 ms. Longer integration times reduce noise at the cost of conversion speed. For thermocouple inputs with 50 Hz line noise, 20 ms is the default. 100 ms is preferred for low-temperature-difference measurements.
6. Note the PIW address from the address overview (for example PIW 288 for channel 0 of slot 4). This is the address the STL program must read. Compile and download the hardware configuration to the CPU.

The SM 331 reports thermocouple values in three modes; the choice in HW config determines the math the program must apply.

For Type K thermocouples in linear mode with 0.1 deg C resolution, the full-scale range of approximately -270 deg C to +1370 deg C maps to -27648 to +27648 counts at 0.1 deg C per count. In the source project, the user observed PIW values around 250 for a 25.0 deg C calibrator reference, confirming 0.1 deg C per count. The applied fix was to switch the measuring type to the linear compensation mode and divide the PIW by 10 to obtain degrees Celsius as a floating-point value for the HMI.

Linear Scaling Formula and Conversion Math

The two-point linear scaling formula converts a raw count X to an engineering value Y:

Where:

• X = current PIW value (raw count, integer)
• MIN = raw count at the lower range limit (typically 0 for unipolar, -27648 for bipolar)
• MAX = raw count at the upper range limit (typically +27648 for unipolar and bipolar)
• UGR = Untergrenze (lower engineering limit) - the value the lower raw count represents in engineering units (e.g. 0.0 deg C)
• OGR = Obergrenze (upper engineering limit) - the value the upper raw count represents in engineering units (e.g. 100.0 deg C)

For a 0-10 V input mapped to 0-100 percent display with a 0-27648 raw range:

For a 4-20 mA input mapped to a 0-200 deg C transmitter with live-zero:

The 4 mA live-zero is handled at the transmitter side: 4 mA corresponds to 0 deg C and 20 mA to 200 deg C. The PIW at 4 mA is approximately 0 (after module calibration) and at 20 mA is 27648.

For a bipolar plus/minus 10 V input mapped to -100 to +100 percent:

STL Implementation: In-line Scaling Function

The in-line approach uses a single STL function (FC) with input parameters for the raw value, UGR, and OGR. The function returns the scaled real value at output OUT. The original snippet from the source project follows the S7-300 STL programming model, where PIW is loaded, type-converted to REAL via ITD then DTR, and the linear math applied.

FUNCTION FC 100 : VOID
// Inputs
//   IN0   : INT    - raw PIW value
//   IN1   : REAL   - UGR (lower engineering limit)
//   IN2   : REAL   - OGR (upper engineering limit)
// Outputs
//   OUT3  : REAL   - scaled engineering value
VAR_TEMP
  TEMP4 : REAL;
END_VAR

BEGIN
  L   #IN0
  ITD                            // Integer to Double Integer
  DTR                            // Double Integer to Real
  T   #TEMP4                     // store raw value as REAL

  L   #IN2                       // OGR
  L   #IN1                       // UGR
  -R                             // OGR - UGR
  L   2.764800e+004              // MAX (27648)
  /R                             // (OGR - UGR) / 27648
  L   #TEMP4                     // X
  *R                             // X * (OGR - UGR) / 27648
  T   #OUT3                      // scaled result
END_FUNCTION

Notes on the code:

• ITD (Integer to Double Integer) and DTR (Double Integer to Real) are required because the CPU-315/317 family used with the SM 331 does not provide a direct INT-to-REAL conversion. The intermediate DINT step prevents sign-extension artifacts on negative raw counts from bipolar thermocouples (which would otherwise be misinterpreted as 16-bit signed values and could overflow on a -27648 input).
• The constant 27648 (unipolar full scale) is the canonical SM 331 raw range for the non-linear thermocouple type and 0-10 V / plus/minus 10 V / 4-20 mA ranges. For the bipolar plus/minus 80 mV range used in the source project (thermocouple in non-linear mode), MIN should be -27648 and MAX should be +27648, with the formula: Y = UGR + (X - (-27648)) * (OGR - UGR) / (27648 - (-27648)).
• The TEMP4 local variable is needed because the accumulator must hold the raw value across the multi-step scaling operation. Without the temp, the result of the -R would overwrite the raw value.
• Call this FC in OB1 (or in a cyclic OB at the desired update rate). The HMI tag is then wired directly to FC100.OUT3 or to a DB location that holds the last result.

DB-Driven Parameterized Scaling Block

The parameterized version is preferred for installations with many scaled points because the upper and lower engineering limits are stored in a DB. The programmer changes the DB values without recompiling the FC, and the same FC is reused for all analog inputs in the project. The DB layout below is the structure used in the source project, with field names translated from the comments (PEW is the German/Italian abbreviation for PIW; OGR = Obergrenze; UGR = Untergrenze).

The FC reads PIW into DBD 0, performs the scaling math against the DB parameters, and writes the result to DBD 32. The HMI tag points to DBD 32.

FUNCTION FC 200 : VOID
// Inputs
//   #DB_NUMERO       : BLOCK_DB - data block containing the scaling parameters
//   #PT100_Direccion : INT      - PIW address of the raw value
// Outputs
//   Result written to DBD 32 of #DB_NUMERO
VAR_TEMP
  DB_No : INT;
END_VAR

BEGIN
  AUF #DB_NUMERO                        // open DB

  L   #PT100_Direccion                  // load PIW
  ITD
  DTR
  T   DBD 0                             // PEW_REAL := PIW (as REAL)

  L   DBD 12                            // OGR
  L   DBD 16                            // UGR
  -R
  T   DBD 20                            // OGR_UGR := OGR - UGR

  L   DBD 0                             // PEW_REAL
  L   DBD 8                             // MIN
  -R
  T   DBD 24                            // PEW_MIN := PEW - MIN

  L   DBD 4                             // MAX
  L   DBD 8                             // MIN
  -R
  T   DBD 28                            // MAX_MIN := MAX - MIN

  L   DBD 20                            // OGR_UGR
  L   DBD 24                            // PEW_MIN
  *R
  L   DBD 28                            // MAX_MIN
  /R
  L   DBD 16                            // UGR
  +R
  T   DBD 32                            // XA := UGR + (PEW - MIN) * (OGR - UGR) / (MAX - MIN)
END_FUNCTION

Usage examples from the source project and common field cases:

• PT100 temperature transmitter, 0-100 deg C: OGR = 100.0, UGR = 0.0, MAX = 27648, MIN = 0.
• Danfoss VLT drive frequency feedback, 0-60 Hz from 4-20 mA: OGR = 60.0, UGR = 0.0, MAX = 27648, MIN = 0.
• Velocity in machine units where the input is a -10 V to +10 V bipolar signal mapped to -100 to +100: OGR = 100.0, UGR = -100.0, MAX = 27648, MIN = -27648.
• Pressure transmitter 0-10 bar with 4-20 mA output: OGR = 10.0, UGR = 0.0, MAX = 27648, MIN = 0.

WinCC / HMI Linear Scaling Configuration

For projects where the scaling should be performed on the HMI side rather than the PLC, WinCC Unified, WinCC Professional, and WinCC Flexible all support tag-level linear scaling. This eliminates the need for a dedicated FC and reduces the PLC scan-time impact when the value is only used for display.

1. In the TIA Portal project tree, open the HMI device and select HMI Tags.
2. Create a new tag pointing to the PLC address (e.g. DB200.DBD0 for the raw REAL, or directly to PEW 288 if the tag type is INT).
3. In the Inspector window, select Properties > Properties > Linear scaling.
4. Click Linear scaling to enable the function. The PLC area shows the raw value type (INT, DINT, REAL) and the raw range endpoints; the HMI area shows the engineering range endpoints.
5. Enter the engineering range endpoints that match the transmitter. For a 4-20 mA / 0-200 deg C transmitter, enter PLC range 0-27648 and HMI range 0.0-200.0. For a -10 V to +10 V bipolar signal mapped to -100 to +100 percent, enter PLC range -27648 to +27648 and HMI range -100.0 to +100.0.
6. Assign the scaled tag to the IO field on the screen. Set the display format to match the engineering unit (e.g. "999.9 deg C" for 0.1 deg C resolution).

For WinCC Unified in TIA Portal V20, the official Siemens documentation describes the procedure in detail: Applying linear scaling to a tag (RT Unified) - WinCC Unified.

For WinCC Professional on Comfort Panels and PC Runtime, the path is HMI Tags > Properties > Linear Scaling on the tag dialog. WinCC V7.x uses Tag Management > Properties > Linear Scaling in the channel/tag dialog. The configuration is identical across versions: PLC range and HMI range are entered as two endpoints each, and the runtime applies the linear mapping.

Compensation Methods and Configuration

Thermocouples generate a voltage proportional to the temperature difference between the measuring junction and the reference (cold) junction. The SM 331 must know the cold-junction temperature to report the absolute measuring-junction temperature. The module offers three compensation modes.

The source project initially used internal compensation with a constant 1.4x error. The recommended fix is to use the linearized thermocouple type (e.g. TC-K linear) which keeps the internal compensation but outputs the value already in 0.1 deg C increments, removing the need for in-program linearization against a non-linear table. The 1.4x error in the original configuration was a symptom of selecting a non-existent "TC-I" type, not a compensation failure; the right fix is to use the correct thermocouple type letter for the physical sensor (TC-K for Type K, TC-J for Type J, and so on). With the correct type selected, internal compensation produces an accurate reading at the calibrator to within the module's plus/minus 0.3 percent of full scale.

Platform Variants: S7-300, S7-1200, S7-1500, and ET 200SP

The SM 331 is specific to the S7-300 platform. The same scaling math applies to all Siemens platforms, but the module, the address scheme, and the HW config steps differ.

The scaling math, the ITD/DTR instructions, and the linear formula are identical across all four platforms. The only differences are the module order number, the slot/address scheme, and the HW config dialog. The DB-driven FC shown above can be reused on any of these platforms with no code changes.

Safety, Commissioning, and Field Checks

For heater control applications (the source use case), the scaled temperature value drives an output that controls a heating element. Three safety and commissioning practices are mandatory.

1. Independent high-temperature cutout. Never rely solely on the scaled value for over-temperature protection. Install a separate safety thermostat or safety PLC (e.g. SIMATIC F-CPU with F-AI) that trips the heater contactor independent of the S7-300 standard program. The S7-300 is a non-safety PLC and the SM 331 is a standard (non-Fail-Safe) module.
2. Wire break and out-of-range detection. Add code that checks for PIW = -32768 (underflow / wire break) and PIW = +32767 (overflow / over-range) and forces the heater output off when these conditions are detected. A broken thermocouple in the source project would otherwise report a very low temperature and the controller would command full heater output indefinitely.
3. First-article calibration. Before relying on the scaled value for closed-loop control, apply a calibrator at three points (0 percent, 50 percent, 100 percent of range) and verify the scaled output matches to within the module accuracy. A 1-LSB error at 0 percent is acceptable; a 1 percent error indicates a configuration mistake (wrong type, wrong compensation, wrong divisor).
4. Commissioning sequence. Use this sequence when bringing a thermocouple loop into service: (a) verify the HW config type matches the physical sensor; (b) verify the wiring at the front connector with a continuity check; (c) apply a calibrator at the lower range limit and confirm the PIW and the scaled value; (d) apply a calibrator at the upper range limit and confirm; (e) apply a calibrator at mid-scale and confirm; (f) enable the closed-loop control only after all three points are within tolerance; (g) enable the safety cutout and test it by intentionally opening the thermocouple wire.

Verification, Calibration, and Field Diagnostics

1. Open the watch table. In STEP 7 / TIA Portal, open a Monitor/Modify table and enter the PIW address. Verify the raw value changes as the thermocouple is heated or cooled.
2. Apply a known reference. Use a thermocouple calibrator (e.g. WIKA CTI9100 or Beamex MC6) at 0.000 mV (Type K reference at 0 deg C) and at 4.096 mV (100 deg C Type K, per NIST polynomial). Record the PIW and the scaled output. They should match the calibrator within the module accuracy spec (plus/minus 0.3 percent of full scale for the -7KF02).
3. Check the end points. With the calibrator at the lower and upper range limits, the FC output should match UGR and OGR to within 1 LSB. A 1-LSB error in the scaled output corresponds to 0.1 deg C (or 0.01 deg C at the higher resolution).
4. Verify on the HMI. Force the FC to write a known value (e.g. write 50.0 to DBD 32 directly) and check that the HMI tag shows 50.0. This isolates the HMI tag configuration from the PLC scaling. If the HMI shows 50.0, the tag is correctly configured; if it shows 0.0 or "########", the tag type is wrong (INT instead of REAL) or the format string is mismatched.
5. Check the scaling direction. A common error is to swap UGR and OGR when the input is inverted (e.g. 4-20 mA where 20 mA is the low end and 4 mA is the high end for some flow sensors). Verify against a single calibrator point that the FC output increases as the calibrator output increases.
6. Check overflow / underflow. Values outside the raw range produce PIW of -32768 (underflow, often wire break) or +32767 (overflow, often over-range). The FC should clamp these or generate a fault bit. Add a check: if PIW < MIN or PIW > MAX, set a fault flag and stop using the scaled value for control.
7. Read the diagnostic buffer. If the SF (group fault) LED on the SM 331 is on, open TIA Portal online, expand Online > Diagnostics > Diagnostic Buffer, and read the SM 331 entry. Common codes: "Wire break" (check thermocouple), "Measuring range violation" (input out of configured range), "Configuration error" (HW config not downloaded or wrong type for the channel group).

Troubleshooting Matrix

Frequently Asked Questions