The phenomenon
If two conductors of different materials are joined in a circuit and the two junctions are at different temperatures, an electromotive force appears that can be measured with a voltmeter. This thermoelectric phenomenon is the Seebeck effect. The resulting voltage depends on the materials and on the temperature difference, and not on the size of the conductors or, in principle, on the shape of the junction.
Where it comes from
In a conductor subjected to a temperature gradient, the charge carriers on the hot side tend to diffuse toward the cold side. This redistribution generates an electric field that opposes the diffusion until an equilibrium is established. The magnitude of this effect per unit of temperature is the Seebeck coefficient, characteristic of each material. Two materials with different coefficients produce different voltages in each leg, and the net difference is the measurable signal.
An important nuance
It is common to think that the voltage arises at the junction. Physically, it originates along the conductors wherever there is a temperature gradient. This has practical consequences: if the wire is not homogeneous in a gradient zone, a spurious voltage is generated, and the reading depends on where the cable runs and not only on the temperatures at the ends.
Related effects
- Peltier effect. The inverse: when current flows through a junction of two materials, heat is absorbed or released. It is the basis of thermoelectric coolers.
- Thomson effect. Absorption or release of heat when current flows through a conductor with a temperature gradient.
Relationship to the thermocouple
A thermocouple takes advantage of the Seebeck effect: one junction is placed at the point to be measured and the other is held at a known reference. The measured voltage corresponds to the difference between the two, so the temperature of the reference junction must be known or compensated. The signal is small and nonlinear with temperature, so standardized tables or polynomials are used to convert it.
What the engineer sees
Unstable readings or unexplained deviations usually indicate parasitic thermoelectric voltages: connections of dissimilar metals, terminals at different temperatures or non-homogeneous conductors. In very low-voltage measurements, these effects also affect instruments that are not thermocouples.