How thermocouple conversion really works
A thermocouple generates a small voltage that depends on the temperature difference between its measuring (hot) junction and the reference (cold) junction where it meets your copper wiring. The standard ITS-90 tables and polynomials all assume the cold junction sits at exactly 0 °C — which it never does in practice. So every real conversion is a three-step dance:
- Measure the loop EMF (mV) and the cold-junction temperature
- Add the EMF that the cold-junction temperature would produce (from the same table)
- Convert the summed EMF to temperature via the reference function
This calculator performs all three steps, using the ITS-90 reference polynomials for each type and numerical inversion for the mV → temperature direction. Note that you always add EMFs, never temperatures — because the mV-temperature curve is nonlinear, adding 25 °C is not the same as adding the EMF of 25 °C.
Worked example
A Type K loop reads 20.500 mV at the meter, and the terminal block is at 25 °C:
- Type K EMF of 25 °C = 1.000 mV
- Table EMF = 20.500 + 1.000 = 21.500 mV
- Inverting the Type K function: 21.500 mV → ≈ 520 °C
Skip the compensation and you'd convert 20.500 mV directly to ≈ 497 °C — a 23 °C error, silently wrong.
Type ranges and typical use
| Type | Range (°C) | Sensitivity | Typical use |
|---|---|---|---|
| K | −270 to 1372 | ~41 µV/°C | general purpose workhorse |
| J | −210 to 1200 | ~55 µV/°C | older plants, reducing atmospheres |
| T | −270 to 400 | ~43 µV/°C | low temperature, food, cryo |
| E | −270 to 1000 | ~68 µV/°C | highest sensitivity |
| N | −270 to 1300 | ~39 µV/°C | improved K, better drift |
| R / S | −50 to 1768 | ~10 µV/°C | furnaces, precious-metal |
| B | 0 to 1820 | ~9 µV/°C | very high temp, no CJC needed |
Field notes
- Use the right extension cable. Ordinary copper between the thermocouple head and the panel moves the effective cold junction to the head — where the temperature is unknown. Use matching extension/compensating cable all the way to the CJC point.
- Polarity matters: a reversed pair reads roughly correct at ambient and drifts the wrong way as the process heats — a classic commissioning catch.
- Millivolts are tiny. At ~41 µV/°C for Type K, a 1 mV measurement error is ~24 °C. Use a meter with µV resolution for bench work.
- Type B quirk: its EMF is essentially zero below 50 °C, so readings below ~250 °C are meaningless — and this calculator's inversion for Type B starts there.
Frequently asked questions
How do I convert thermocouple mV to temperature?
Add the cold-junction EMF to the measured millivolts, then look up (or numerically invert) the ITS-90 polynomial for your thermocouple type. Skipping the cold-junction step is the most common conversion error — the tables assume a 0 °C reference.
What is cold junction compensation (CJC)?
A thermocouple measures the temperature DIFFERENCE between its hot junction and the point where it connects to copper wiring (the cold junction). CJC measures the cold-junction temperature and adds its equivalent EMF so the reading reflects the true hot-junction temperature.
How many mV does a Type K thermocouple produce at 100 °C?
With the reference junction at 0 °C, Type K produces 4.096 mV at 100 °C. With the cold junction at a typical 25 °C room, a meter would read about 3.096 mV (4.096 minus the 1.000 mV of 25 °C).
Why does Type B not need cold junction compensation?
Type B output is nearly zero below about 50 °C (0.002 mV at 25 °C), so the cold-junction contribution is negligible at room temperature. It is also why Type B readings below roughly 250 °C are unusable.
Provided for reference and education. Verify independently before use in safety-critical work. See our disclaimer.