Reference data
Thermocouple reference tables
Millivolts against temperature for the five common letter types, computed from the NIST ITS-90 reference functions rather than transcribed — with the Seebeck coefficient, so you can price an instrument's microvolt of noise in degrees, and the IEC 60584-1 tolerance bands.
Read these against a 0 °C reference
Every value below is the emf a junction at that temperature produces with its reference junction held at 0 °C. A thermocouple measures a difference, not a temperature, so unless you are using an ice-point reference your meter is reading something smaller: the table value minus the emf of your own terminal block. The calculator does that subtraction properly — in millivolts, which is the only way that gives the right answer.
Type K: -270 to 1372 °C
Nickel-chromium / nickel-alumel. The general-purpose workhorse: wide range, cheap, but the least stable of the common types above about 800 °C.
| t (°C) | Emf (mV) | dE/dt (µV/°C) | Class 1 (±°C) | Class 2 (±°C) |
|---|---|---|---|---|
| -270 | -6.458 | 0.7 | — | — |
| -250 | -6.404 | 4.9 | — | — |
| -200 | -5.891 | 15.3 | — | — |
| -150 | -4.913 | 23.6 | — | — |
| -100 | -3.554 | 30.5 | — | — |
| -50 | -1.889 | 35.8 | — | — |
| 0 | 0.000 | 39.5 | 1.50 | 2.50 |
| 50 | 2.023 | 41.2 | 1.50 | 2.50 |
| 100 | 4.096 | 41.4 | 1.50 | 2.50 |
| 150 | 6.138 | 40.3 | 1.50 | 2.50 |
| 200 | 8.138 | 40.0 | 1.50 | 2.50 |
| 250 | 10.153 | 40.7 | 1.50 | 2.50 |
| 300 | 12.209 | 41.4 | 1.50 | 2.50 |
| 350 | 14.293 | 41.9 | 1.50 | 2.63 |
| 400 | 16.397 | 42.2 | 1.60 | 3.00 |
| 450 | 18.516 | 42.5 | 1.80 | 3.38 |
| 500 | 20.644 | 42.6 | 2.00 | 3.75 |
| 550 | 22.776 | 42.6 | 2.20 | 4.13 |
| 600 | 24.905 | 42.5 | 2.40 | 4.50 |
| 650 | 27.025 | 42.3 | 2.60 | 4.88 |
| 700 | 29.129 | 41.9 | 2.80 | 5.25 |
| 750 | 31.213 | 41.5 | 3.00 | 5.63 |
| 800 | 33.275 | 41.0 | 3.20 | 6.00 |
| 850 | 35.313 | 40.5 | 3.40 | 6.38 |
| 900 | 37.326 | 40.0 | 3.60 | 6.75 |
| 950 | 39.314 | 39.5 | 3.80 | 7.13 |
| 1000 | 41.276 | 39.0 | 4.00 | 7.50 |
| 1050 | 43.211 | 38.4 | — | 7.88 |
| 1100 | 45.119 | 37.9 | — | 8.25 |
| 1150 | 46.995 | 37.2 | — | 8.63 |
| 1200 | 48.838 | 36.5 | — | 9.00 |
| 1250 | 50.644 | 35.7 | — | — |
| 1300 | 52.410 | 34.9 | — | — |
| 1350 | 54.138 | 34.2 | — | — |
| 1372 | 54.886 | 33.9 | — | — |
Type T: -270 to 400 °C
Copper / constantan. The type for cold-chain and cryogenic work — well behaved below 0 °C and the most accurate common type there.
| t (°C) | Emf (mV) | dE/dt (µV/°C) | Class 1 (±°C) | Class 2 (±°C) |
|---|---|---|---|---|
| -270 | -6.258 | 1.0 | — | — |
| -250 | -6.180 | 6.3 | — | — |
| -200 | -5.603 | 15.7 | — | — |
| -150 | -4.648 | 22.3 | — | — |
| -100 | -3.379 | 28.4 | — | — |
| -50 | -1.819 | 33.9 | — | — |
| 0 | 0.000 | 38.7 | 0.50 | 1.00 |
| 50 | 2.036 | 42.8 | 0.50 | 1.00 |
| 100 | 4.279 | 46.8 | 0.50 | 1.00 |
| 150 | 6.704 | 50.2 | 0.60 | 1.13 |
| 200 | 9.288 | 53.1 | 0.80 | 1.50 |
| 250 | 12.013 | 55.8 | 1.00 | 1.88 |
| 300 | 14.862 | 58.1 | 1.20 | 2.25 |
| 350 | 17.819 | 60.2 | 1.40 | 2.63 |
| 400 | 20.872 | 61.8 | — | — |
Type J: -210 to 1200 °C
Iron / constantan. High output, common on older plant; the iron leg oxidises, so it is not for long service in moist air above 500 °C.
| t (°C) | Emf (mV) | dE/dt (µV/°C) | Class 1 (±°C) | Class 2 (±°C) |
|---|---|---|---|---|
| -210 | -8.095 | 19.1 | — | — |
| -200 | -7.890 | 21.9 | — | — |
| -150 | -6.500 | 33.1 | — | — |
| -100 | -4.633 | 41.1 | — | — |
| -50 | -2.431 | 46.6 | — | — |
| 0 | 0.000 | 50.4 | 1.50 | 2.50 |
| 50 | 2.585 | 52.8 | 1.50 | 2.50 |
| 100 | 5.269 | 54.4 | 1.50 | 2.50 |
| 150 | 8.010 | 55.2 | 1.50 | 2.50 |
| 200 | 10.779 | 55.5 | 1.50 | 2.50 |
| 250 | 13.555 | 55.5 | 1.50 | 2.50 |
| 300 | 16.327 | 55.4 | 1.50 | 2.50 |
| 350 | 19.090 | 55.2 | 1.50 | 2.63 |
| 400 | 21.848 | 55.2 | 1.60 | 3.00 |
| 450 | 24.610 | 55.4 | 1.80 | 3.38 |
| 500 | 27.393 | 56.0 | 2.00 | 3.75 |
| 550 | 30.216 | 57.0 | 2.20 | 4.13 |
| 600 | 33.102 | 58.5 | 2.40 | 4.50 |
| 650 | 36.071 | 60.3 | 2.60 | 4.88 |
| 700 | 39.132 | 62.2 | 2.80 | 5.25 |
| 750 | 42.281 | 63.7 | 3.00 | 5.63 |
| 800 | 45.494 | 64.6 | — | — |
| 850 | 48.715 | 64.0 | — | — |
| 900 | 51.877 | 62.4 | — | — |
| 950 | 54.956 | 60.7 | — | — |
| 1000 | 57.953 | 59.3 | — | — |
| 1050 | 60.890 | 58.3 | — | — |
| 1100 | 63.792 | 57.8 | — | — |
| 1150 | 66.679 | 57.6 | — | — |
| 1200 | 69.553 | 57.2 | — | — |
Type N: -270 to 1300 °C
Nicrosil / nisil. Designed to replace K: the same sort of range with markedly better stability and less drift at high temperature.
| t (°C) | Emf (mV) | dE/dt (µV/°C) | Class 1 (±°C) | Class 2 (±°C) |
|---|---|---|---|---|
| -270 | -4.345 | 0.3 | — | — |
| -250 | -4.313 | 2.9 | — | — |
| -200 | -3.990 | 9.9 | — | — |
| -150 | -3.336 | 16.0 | — | — |
| -100 | -2.407 | 20.9 | — | — |
| -50 | -1.269 | 24.3 | — | — |
| 0 | 0.000 | 26.0 | 1.50 | 2.50 |
| 50 | 1.340 | 27.7 | 1.50 | 2.50 |
| 100 | 2.774 | 29.6 | 1.50 | 2.50 |
| 150 | 4.302 | 31.4 | 1.50 | 2.50 |
| 200 | 5.913 | 33.0 | 1.50 | 2.50 |
| 250 | 7.597 | 34.3 | 1.50 | 2.50 |
| 300 | 9.341 | 35.4 | 1.50 | 2.50 |
| 350 | 11.136 | 36.4 | 1.50 | 2.63 |
| 400 | 12.974 | 37.1 | 1.60 | 3.00 |
| 450 | 14.846 | 37.8 | 1.80 | 3.38 |
| 500 | 16.748 | 38.3 | 2.00 | 3.75 |
| 550 | 18.672 | 38.7 | 2.20 | 4.13 |
| 600 | 20.613 | 39.0 | 2.40 | 4.50 |
| 650 | 22.566 | 39.1 | 2.60 | 4.88 |
| 700 | 24.527 | 39.3 | 2.80 | 5.25 |
| 750 | 26.491 | 39.3 | 3.00 | 5.63 |
| 800 | 28.455 | 39.3 | 3.20 | 6.00 |
| 850 | 30.416 | 39.2 | 3.40 | 6.38 |
| 900 | 32.371 | 39.0 | 3.60 | 6.75 |
| 950 | 34.319 | 38.9 | 3.80 | 7.13 |
| 1000 | 36.256 | 38.6 | 4.00 | 7.50 |
| 1050 | 38.179 | 38.3 | — | 7.88 |
| 1100 | 40.087 | 38.0 | — | 8.25 |
| 1150 | 41.976 | 37.6 | — | 8.63 |
| 1200 | 43.846 | 37.2 | — | 9.00 |
| 1250 | 45.694 | 36.7 | — | — |
| 1300 | 47.513 | 36.0 | — | — |
Type E: -270 to 1000 °C
Nickel-chromium / constantan. The highest output of the standard types — about 68 µV/°C near room temperature — which makes it the choice when resolution matters most.
| t (°C) | Emf (mV) | dE/dt (µV/°C) | Class 1 (±°C) | Class 2 (±°C) |
|---|---|---|---|---|
| -270 | -9.835 | 1.6 | — | — |
| -250 | -9.718 | 9.7 | — | — |
| -200 | -8.825 | 25.1 | — | — |
| -150 | -7.279 | 36.2 | — | — |
| -100 | -5.237 | 45.2 | — | — |
| -50 | -2.787 | 52.6 | — | — |
| 0 | 0.000 | 58.7 | 1.50 | 2.50 |
| 50 | 3.048 | 63.2 | 1.50 | 2.50 |
| 100 | 6.319 | 67.5 | 1.50 | 2.50 |
| 150 | 9.789 | 71.1 | 1.50 | 2.50 |
| 200 | 13.421 | 74.0 | 1.50 | 2.50 |
| 250 | 17.181 | 76.2 | 1.50 | 2.50 |
| 300 | 21.036 | 77.9 | 1.50 | 2.50 |
| 350 | 24.964 | 79.2 | 1.50 | 2.63 |
| 400 | 28.946 | 80.1 | 1.60 | 3.00 |
| 450 | 32.965 | 80.6 | 1.80 | 3.38 |
| 500 | 37.005 | 80.9 | 2.00 | 3.75 |
| 550 | 41.053 | 80.9 | 2.20 | 4.13 |
| 600 | 45.093 | 80.7 | 2.40 | 4.50 |
| 650 | 49.116 | 80.2 | 2.60 | 4.88 |
| 700 | 53.112 | 79.7 | 2.80 | 5.25 |
| 750 | 57.080 | 79.1 | 3.00 | 5.63 |
| 800 | 61.017 | 78.4 | 3.20 | 6.00 |
| 850 | 64.922 | 77.7 | — | 6.38 |
| 900 | 68.787 | 76.8 | — | 6.75 |
| 950 | 72.603 | 75.8 | — | — |
| 1000 | 76.373 | 75.2 | — | — |
The five types side by side
Output near room temperature, and what a 1 µV instrument error is worth in degrees there — the number that decides whether your readout is good enough for the job.
| Type | Range (°C) | dE/dt at 20 °C | 1 µV is worth | Emf at 100 °C |
|---|---|---|---|---|
| K | -270 to 1372 | 40.3 µV/°C | 24.8 m°C | 4.096 mV |
| T | -270 to 400 | 40.3 µV/°C | 24.8 m°C | 4.279 mV |
| J | -210 to 1200 | 51.5 µV/°C | 19.4 m°C | 5.269 mV |
| N | -270 to 1300 | 26.6 µV/°C | 37.6 m°C | 2.774 mV |
| E | -270 to 1000 | 60.5 µV/°C | 16.5 m°C | 6.319 mV |
Frequently asked questions
- How many millivolts is 100 °C on a type K thermocouple?
- 4.096 mV, measured against a reference junction held at 0 °C. If your meter's terminals are the reference and they sit at 22 °C, the same junction reads about 3.216 mV instead — the difference is the emf of the reference junction, and forgetting it is the most common thermocouple error there is.
- What is the millivolt output of a thermocouple at 0 °C?
- Zero, for every type, because the tables are defined against a 0 °C reference junction. A thermocouple measures the difference between its two junctions, so 0 mV means the two ends are at the same temperature — not that the measuring junction is at 0 °C.
- Which thermocouple type has the highest output?
- Type E, at roughly 61 µV/°C near room temperature, against about 41 µV/°C for type K and 26 µV/°C for type N. Higher output means a given microvolt of instrument noise or offset is worth fewer millikelvin, which is why type E is chosen where resolution matters most.
- Which type is best below 0 °C?
- Type T. It is defined down to −270 °C, is the best behaved of the common types in that region, and is the one IEC 60584-1 gives the tightest sub-zero class to. Types K, N and E are defined down there too; type J is not, stopping at −210 °C.
- What do these tables actually describe?
- New, nominal wire of that letter type: the standard reference function every type-K thermocouple is meant to approximate. A specific thermocouple deviates from it, which is what a calibration certificate quantifies. Use the certificate for measurement and the table for everything else.
- Where do the numbers come from?
- The NIST ITS-90 Thermocouple Database (Monograph 175) polynomials, evaluated in this page by the same code the calculator uses. That code is unit-tested against the published tables at both ends and through the middle of each type's range, so a mistyped coefficient cannot pass unnoticed.
Values are computed in this page from the NIST ITS-90 Thermocouple Database (Monograph 175) reference functions, by code unit-tested against the published tables. Tolerance classes are from IEC 60584-1, published by the International Electrotechnical Commission; ValiTrac does not distribute either document.