Electrical

Thermocouple

Two dissimilar metals generating a voltage proportional to temperature difference

A thermocouple is a temperature sensor made from two wires of dissimilar metals joined at one end. When the junction is at a different temperature than the open ends (the cold junction), a small voltage develops via the Seebeck effect. The voltage encodes the temperature difference through a calibrated, nonlinear function specific to the metal pair. Thermocouples cover huge ranges (down to -250 °C and up to 2,300 °C for some types), respond quickly, are rugged, and need no external excitation, but accuracy is limited and a separate cold-junction reference is required.

  • EffectSeebeck (1821)
  • OutputMicrovolts per degree
  • Common typesK, J, T, E, N, R, S, B
  • K-type range-200 to +1,260 °C
  • Cold junctionReference temp required
  • Accuracy±1 to ±2 °C typical

Interactive visualization

Press play, or step through manually. The visualization is yours to drive — try it before reading on.

Open visualization fullscreen ↗

Watch the 60-second explainer

A condensed visual walkthrough — narrated, captioned, under a minute.

Why thermocouples matter

  • Industrial furnaces. Steel, glass, ceramic processing.
  • Engines. Exhaust gas temperature in turbines and turbochargers.
  • HVAC. Boiler control, flame safety, duct temperatures.
  • Cryogenics. Type T probes in liquid nitrogen and helium systems.
  • Aerospace. Combustor and exhaust monitoring.
  • Cooking. Probe thermometers, oven controls.
  • Research. Wide-range, fast-response measurements.

Common misconceptions

  • Junction generates voltage. The temperature gradient along the wire does; the junction just connects pairs.
  • Longer wire means more voltage. Output depends only on temperature difference between junctions.
  • Any wire works for extensions. Must match the alloy or compensation breaks.
  • Calibration lasts forever. Aging and contamination shift output, especially at high temperature.
  • More accurate than RTDs. RTDs are more precise; thermocouples win on range and ruggedness.
  • Output is linear. All types are nonlinear; lookup tables or polynomial fits are required.

Frequently asked questions

How does it work?

When two dissimilar metals are joined and a temperature gradient exists along them, charge carriers diffuse at different rates in each metal, producing a net voltage across the open ends. This is the Seebeck effect. The voltage depends on the metal pair and the temperature difference between the hot and cold junctions, not on the wire length or junction size.

Why is a cold junction needed?

A thermocouple measures temperature differences, not absolute temperatures. The reading is the voltage between the hot junction and where the wires connect to the measuring instrument. The instrument must know the temperature of that connection point (the cold junction) and add it back. Modern systems integrate a separate temperature sensor at the terminal block for cold-junction compensation.

What's the difference between thermocouple types?

Each pair has a unique Seebeck coefficient, range, and accuracy. Type K (chromel/alumel): general-purpose, -200 to 1,260 °C. Type J (iron/constantan): vacuum, reducing atmospheres. Type T (copper/constantan): cryogenic. Types R, S, B (platinum-rhodium): very high temperature, expensive. Standard color codes and wire identification are critical to avoid mismatched extension wire errors.

How does it compare to RTDs and thermistors?

RTDs (platinum resistance) are more accurate (±0.1 °C) but slower, narrower range, and need excitation current. Thermistors are most sensitive at room temperatures, very nonlinear, and limited range. Thermocouples cover the widest range, respond fastest, are simplest, but have lowest accuracy. Use thermocouples for furnaces, RTDs for process control, thermistors for medical and consumer.

What's the response time?

Depends on junction size and sheath. An exposed-bead 0.13 mm K-type responds in tens of milliseconds. A grounded-junction sheathed probe (3 mm) takes a few seconds. An ungrounded probe is electrically isolated but slower. For furnace control, multi-second response is fine; for combustion diagnostics, exposed beads are needed.

What causes errors?

(1) Cold junction temperature errors. (2) Mismatched extension wire (using copper instead of matching alloy). (3) Thermal gradients along the wire near the junction. (4) Aging and oxidation, which shifts calibration. (5) Electrical noise picked up by the millivolt-level signal. (6) Mechanical strain on the wires, which alters Seebeck coefficient.

What's a thermopile?

Many thermocouples in series, with alternating junctions exposed to source and reference temperatures. The voltages add, multiplying sensitivity. Thermopiles are used in radiation pyrometers, gas-fired pilot light flame sensors, and IR sensors for non-contact temperature measurement (forehead thermometers, motion detectors). They produce enough voltage to drive small valves directly without amplification.