A K-type thermocouple reading corrected for its cold junction, and the same raw Pt100 resistance interpreted three different ways depending on wiring configuration.
| Thermocouple type | K-type |
| Measured (raw) thermocouple voltage | 10.153 mV |
| Cold-junction (terminal block) temperature | 25°C |
| RTD | Pt100 (R0 = 100 Ω) |
| Measured resistance | 108.5 Ω |
| Lead resistance (if known) | 0.5 Ω per conductor |
A thermocouple only measures the voltage difference between its hot and cold junctions — the cold junction's own temperature has to be converted to an equivalent mV and added back in, using the same trusted NIST ITS-90 inverse polynomial (inverted numerically) rather than a separately-fitted forward curve.
Using the same raw 108.5 Ω measurement and 0.5 Ω/conductor lead resistance in each case.
| Wiring | Compensated resistance | Resulting temperature |
|---|---|---|
| 2-wire | 107.5 Ω (both leads subtracted) | 19.24°C |
| 3-wire | 108.0 Ω (lead resistance substantially cancelled) | 20.53°C |
| 4-wire (Kelvin) | 108.5 Ω (unchanged — no lead error) | 21.82°C |
| Check | Requirement | Actual | Status |
|---|---|---|---|
| Thermocouple compensated hot-junction temperature | n/a (this is the result) | 274.5°C | ✓ PASS |
| RTD reading spread across wiring methods (same raw resistance) | n/a (informational) | 19.24°C to 21.82°C — a 2.58°C spread | ✓ PASS |
Key insight: A raw sensor reading is meaningless without knowing exactly how it needs to be interpreted — a thermocouple's raw millivolts are useless without cold-junction compensation, and an RTD's raw resistance carries real, quantifiable error from lead-wire resistance unless the wiring configuration (and ideally 4-wire Kelvin sensing) removes it. Both errors are entirely predictable and correctable, which is exactly why instrumentation specifications call out cold-junction compensation method and RTD wiring configuration explicitly rather than leaving them to assumption.
Every input in this example is editable in the live calculator — free, no signup.
Open Thermocouple & RTD calculator →The correction only works as well as the assumed lead resistance value actually matches reality — a 2-wire configuration has no independent way to measure its own lead resistance, so any correction applied depends on a separately estimated or datasheet lead-resistance figure, which is exactly why the 2-wire note calls it 'the largest source of error over long runs': the correction is only as good as that external assumption, unlike 4-wire sensing which measures true resistance directly regardless of lead length.
Thermocouple output is genuinely non-linear across their working temperature range — a linear approximation (using a single fixed mV/°C slope) introduces increasing error the further the reading is from the calibration point it was based on, which is exactly why NIST publishes higher-order polynomial coefficients spanning specific voltage ranges, each fitted to accurately track the real non-linear thermoelectric behavior of that thermocouple type.