Thermocouples: types, cold junction, cables
How thermocouples generate a voltage, the common types and ranges, cold junction compensation, extension and compensating cables, polarity and typical errors.
Principle
Two different metals joined at the measuring point produce a small voltage (Seebeck effect) that depends on the temperature difference between the hot junction and the point where the wires connect to copper — the reference (cold) junction. Reference functions and tables are defined in IEC 60584-1 (based on ITS-90; NIST publishes the same data).
Common types
| Type | Materials | Typical range | Sensitivity (approx.) |
|---|---|---|---|
| K | NiCr – NiAl | −200 … 1200 °C | 41 µV/°C |
| J | Fe – CuNi | −40 … 750 °C | 52 µV/°C |
| T | Cu – CuNi | −200 … 350 °C | 43 µV/°C |
| N | NiCrSi – NiSi | −200 … 1200 °C | 39 µV/°C |
| S / R | Pt-Rh – Pt | 0 … 1600 °C | 10 µV/°C |
| B | Pt-Rh – Pt-Rh | 600 … 1700 °C | ≈ 8 µV/°C at high t |
Exact ranges and tolerance classes depend on the standard class and wire diameter.
Cold junction compensation
Because the voltage depends on the difference, the instrument must know the reference junction temperature: an internal sensor in the terminal block of the module/transmitter, an external Pt100 at a terminal box, or a fixed reference temperature. A wrong or disabled compensation shifts the reading by the terminal temperature — the classic "thermocouple shows room temperature when cold" check.
Cables
Between sensor and instrument use extension (same alloys) or compensating cable (cheaper alloys matching within a temperature range) of the right type. Copper cable in between moves the reference junction to the wrong place. Polarity matters: reversed wires give readings going the wrong way.
Thermocouple or RTD?
Thermocouples: higher temperatures, robust, fast, cheap — but less accurate and need compensation. RTDs: better accuracy and stability below ~600 °C — see RTD and Pt100.
Connections
Open in Knowledge Graph →Loading the map…
Sources
Updated 2026-10-02