Skip to content
Instrumentation HubCalibrate your knowledge
Measurement / Temperature calibration

Temperature Calibration: RTD & Thermocouple Loops

A temperature loop has three parts that can each be wrong: the sensor, the wiring and the transmitter. Check them separately, and you will know exactly where any error comes from.

Field work needs a valid permit, isolation and the site's procedures. Tell operations before you touch a loop that feeds control or a safety system, and put the loop in manual or bypass only with their agreement. Device menus, limits and tolerances vary by make and model: always check the manufacturer's manual.

Calibrate the parts separately

  • Transmitter: disconnect the sensor and inject a known resistance (RTD) or millivolt (thermocouple) from a calibrator. This checks the electronics and the range.
  • Sensor: put the sensor in a dry block or bath next to a reference probe and compare. This checks the element.
  • Loop: with sensor and transmitter connected, compare the DCS reading against the reference. This catches wiring and configuration errors.

RTD (Pt100) checks

A Pt100 is 100 Ω at 0 °C and about 138.5 Ω at 100 °C, roughly 0.385 Ω per °C. Useful reference values (IEC 60751):

TemperaturePt100 resistance
−50 °C80.31 Ω
0 °C100.00 Ω
50 °C119.40 Ω
100 °C138.51 Ω
200 °C175.86 Ω
300 °C212.05 Ω
IEC 60751 classTolerance
AA± (0.10 + 0.0017 |t|) °C
A± (0.15 + 0.002 |t|) °C
B± (0.30 + 0.005 |t|) °C
Field example: 2-wire error
A Pt100 was wired 2-wire with 1 Ω total lead resistance. The transmitter sees 1 Ω extra, about 1 / 0.385 = 2.6 °C high. Converting to 3-wire (or 4-wire) cancels most of the lead resistance.
  • Open element: reading goes to full scale or the transmitter flags a sensor fault.
  • Short or moisture in the sheath: reading low and unstable; check insulation resistance between element and sheath (with the transmitter disconnected).

Thermocouple checks

A thermocouple produces a small voltage proportional to the difference between the hot (measuring) junction and the cold (reference) junction at the transmitter terminals. The transmitter adds cold-junction compensation (CJC).

FaultWhat you seeCheck
Extension wire reversedReading falls when temperature rises, or reads low by about twice the terminal temperaturePolarity at every joint: in IEC colours the negative is white
Copper wire used instead of extension wireError that changes with ambient temperatureUse the matching extension or compensating cable all the way
Wrong type configured (K vs J)Large error that grows with temperatureType in the transmitter vs sensor tag
Open thermocoupleUpscale or downscale burnout per configurationContinuity of the loop of wire
Ageing / drift at high temperatureSlowly reads wrong after months at high temperatureCompare with a reference in a dry block; replace if out of tolerance

Typical type K tolerance (IEC 60584 class 2) is ±2.5 °C or ±0.0075·|t|, whichever is larger. Read more in RTD vs Thermocouple.

Transmitter calibration procedure

  1. Record the tag, range, sensor type and wiring (2/3/4-wire or TC type) from the datasheet.
  2. Disconnect the sensor. Connect the calibrator in resistance (RTD) or mV/TC simulation mode. For TC simulation, use the calibrator's own CJC or a matching wire to the terminals.
  3. Connect a mA meter in the loop (or read the transmitter's output with a calibrator).
  4. Apply 0, 25, 50, 75 and 100 % of the range going up, then back down. Record the mA at each point.
  5. Calculate the error in % of span at each point and compare with the tolerance.
  6. If out of tolerance, do a sensor trim / output trim as the manual describes, then repeat the test and record As-left.
  7. Reconnect the sensor, check the reading is sensible, and return the loop to operations.
Expected mA = 4 + 16 × (T − LRV) / (URV − LRV) Example: range 0–200 °C, inject 138.51 Ω (100 °C) → 12.00 mA expected Error % span = (measured − expected) / 16 × 100

Sensor check in a dry block

  • Use an insertion depth of at least 10–15 times the sheath diameter so the tip is not cooled by the air above the block.
  • Put the reference probe in the same insert, as close as possible to the sensor under test.
  • Wait for stability at each point (often several minutes). Rushing is the most common source of error.
  • Test at least two points across the working range, for example 0 °C (ice point or block) and near the normal operating temperature.
Field example: loop reads 3 °C high
A tank temperature read 3 °C higher than the local thermometer. Transmitter test with a resistance source: within 0.1 °C. Sensor in a dry block: within 0.2 °C. The remaining error was found in the wiring: a 2-wire connection on a 3-wire configured transmitter with a long cable. Correcting the third core removed the error.

Records

Record As-found and As-left at every point, the reference standard used (with its certificate number and due date), ambient temperature, date and technician. As-found data is what tells you whether the calibration interval is right.

Q&A quiz

Related: RTD vs Thermocouple · Temperature · RTD app · Thermocouple app