Thermocouple Temperature Converter Guide: Millivolts, Types and Temperature Scales

Thermocouples are the workhorses of industrial temperature measurement. They are simple, rugged, and cheap: two wires of different metals joined at one end, producing a small voltage that changes with temperature. That simplicity hides a subtlety that catches many users. The voltage does not measure the temperature of the measuring junction alone. It measures the difference in temperature between the measuring junction and a reference junction, and converting millivolts to degrees requires knowing the thermocouple type, the reference temperature, and the correct conversion table or polynomial. A converter tool performs that conversion, but using it well means understanding what it is converting and what can go wrong.

This guide explains the principle, the common types, and the conversion process, with a worked temperature-scale example whose arithmetic can be checked by hand. It is general technical guidance, not a specification for any safety-critical measurement. Where temperature measurement protects people, equipment, or a process, follow the instrument manufacturer's instructions, the applicable standards, and the site's safety procedures, and have critical measurements verified by qualified personnel.

The Seebeck effect and the two junctions

When two different metals are joined and the two junctions of the resulting circuit are at different temperatures, a voltage appears. This is the Seebeck effect, named after Thomas Johann Seebeck, who observed it in the 1820s. In a practical thermocouple, one junction, the hot or measuring junction, is placed where the temperature is to be measured, and the other junction, the cold or reference junction, is kept at a known temperature, traditionally the ice point, 0 degrees Celsius, or compensated electronically to a known value. The measured voltage corresponds to the temperature difference between the two junctions, not to the absolute temperature of the measuring junction. If the reference junction is at 20 degrees Celsius and the measuring junction is at 120 degrees, the voltage corresponds to a difference of 100 degrees, and the converter must add the reference temperature to report 120 degrees.

Modern instruments usually handle reference compensation internally, measuring the temperature at the terminal block where the thermocouple wires connect and adding it automatically. That convenience creates a trap: if the instrument's compensation is disabled, set wrongly, or confused by a terminal block at a different temperature from the assumed reference, every reading will be offset by the error. A reading that is consistently wrong by roughly room temperature is the classic symptom of a reference problem. When a thermocouple reading disagrees with another thermometer by a plausible room-temperature amount, check the reference compensation before suspecting the sensor itself.

Thermocouple types and their ranges

Thermocouples are identified by type letters, each denoting a pair of alloys with a particular voltage-to-temperature relationship and a suitable temperature range. Type K, chromel and alumel, is the most common general-purpose type, usable roughly from minus 200 to over 1,200 degrees Celsius, though the practical range depends on wire size and sheath. Type J, iron and constantan, suits lower temperatures and reducing atmospheres but oxidises at high temperatures. Type T, copper and constantan, is accurate at low and cryogenic temperatures. Types R, S, and B use platinum alloys for high temperatures, above 1,000 degrees, in applications such as furnaces and kilns, where base-metal types would fail. Type E offers a high voltage output per degree, useful for sensitive low-temperature measurement, and Type N is a newer nickel-based type designed for stability at high temperatures.

The type matters because the millivolt-to-temperature relationship differs for each type, and it is not linear. A voltage of, say, 10 millivolts means a different temperature for Type K than for Type J, and the difference is not a fixed offset. Conversion uses standard tables or polynomials published in standards such as IEC 60584, which tabulate voltage against temperature for each type at a reference of 0 degrees Celsius. A converter that asks for the type is not being fussy; it is asking for the one piece of information without which conversion is impossible. Using the wrong type's table is a common and serious error, because the result will look plausible, with no warning, while being wrong by tens or hundreds of degrees.

Converting between temperature scales

Alongside millivolt conversion, a thermocouple converter often converts between temperature scales: Celsius, Fahrenheit, and Kelvin. Those conversions are exact formulas, unlike the empirical thermocouple tables, and they are worth knowing by hand as a check. Celsius to Fahrenheit: multiply Celsius by 9, divide by 5, and add 32. Fahrenheit to Celsius: subtract 32, multiply by 5, and divide by 9. Celsius to Kelvin: add 273.15. Kelvin has no degrees symbol in its name, and its zero is absolute zero, the lowest possible temperature, which is minus 273.15 degrees Celsius. Confusing scales is a classic error in industrial settings, especially when one instrument reports Fahrenheit and another reports Celsius, and a converter that shows all three scales side by side helps catch it.

Accuracy and resolution deserve attention. A thermocouple system, including sensor, extension wire, and instrument, typically achieves accuracy of a few degrees, limited by wire tolerances, reference compensation, and electrical noise, rather than the fractions of a degree that a digital display may imply. A display showing 500.0 degrees does not mean the temperature is known to a tenth of a degree; it means the instrument resolved its calculation to a tenth of a degree, which is not the same thing. For control purposes, repeatability, the ability to return to the same reading under the same conditions, often matters more than absolute accuracy, and thermocouples are valued partly because they repeat well even where absolute accuracy is modest.

Worked example with sample values

Suppose a sample temperature of 500 degrees Celsius must be expressed in Fahrenheit, using a clearly labelled sample value and the exact conversion formula. Multiply 500 by 9, which equals 4,500. Divide by 5, which equals 900, because 5 times 900 is 4,500. Add 32, which gives 932. So 500 degrees Celsius equals 932 degrees Fahrenheit. Check the result in reverse: subtract 32 from 932, which gives 900; multiply by 5, which gives 4,500; divide by 9, which gives 500, confirming the original value. The round trip closes exactly, as it should for an exact formula.

Now apply the thermocouple logic to a sample reading. Suppose a sample Type K thermocouple, with its reference junction correctly compensated to 0 degrees Celsius, produces a voltage that the standard Type K table associates with 500 degrees Celsius. The converter reports 500 degrees Celsius, which is 932 degrees Fahrenheit and 773.15 Kelvin, because 500 plus 273.15 equals 773.15. If instead the reference junction were actually at 20 degrees Celsius but the instrument assumed 0 degrees, the voltage would correspond to a difference of 480 degrees, and the instrument would report about 480 degrees Celsius, understating the true temperature by about 20 degrees. That single example shows why reference temperature is not a footnote: it shifts every reading by its own amount, and in a process controlled to a tight temperature, a 20-degree error can be the difference between a good product and a failed batch.

Common mistakes to avoid

The first mistake is using the wrong thermocouple type in the converter or instrument, which produces plausible but wrong temperatures. The second is forgetting the reference junction, especially when connecting a thermocouple to a data logger or controller whose compensation settings differ from the previous instrument. The third is using ordinary copper wire instead of matching thermocouple extension wire between the sensor and the instrument, which creates extra junctions at unknown temperatures and corrupts the reading. The fourth is confusing temperature scales, especially Celsius and Fahrenheit, when comparing instruments. The fifth is trusting display resolution as accuracy, and making fine control decisions on differences smaller than the system's real accuracy. The sixth is exposing a thermocouple to conditions beyond its type's range, such as using a base-metal type in a corrosive or very high temperature environment, which drifts or fails. Check the type, check the reference, use matching extension wire, confirm the scale, and treat the displayed precision with the scepticism it deserves.

Try it next

Try the Thermocouple Temperature Converter Guide tool with your own numbers, then explore the Pt100 RTD Temperature Converter and the Voltage Divider Calculator for related calculations. For a companion read, see /blog/pt100-rtd-temperature-converter-guide/, which applies the same measure-carefully habit to a different problem.