Turn a temperature difference into a voltage — or work backwards from a measured voltage. Covers single thermocouples, multi-couple thermopiles and Bi₂Te₃ modules, with load power and thermoelectric conversion efficiency.
Preset values fill the coefficient field below — edit freely.
Signed. Positive → p-type dominant.
N = 1 single thermocouple; N > 1 thermopile / module.
—
—
Applies to both temperature fields.
DC resistance of legs + leads. Blank / 0 hides power results.
Set equal to Rint for maximum power transfer.
¯ZT at the mean temperature: ≈0.8–1.1 for cheap modules, 1.5–3 for good ones.
| Material / pair | S (µV/K) | Basis | Where you meet it |
|---|---|---|---|
| Type E — Chromel / Constantan | +68 | whole couple | Highest-output common base-metal thermocouple |
| Type J — Iron / Constantan | +52 | whole couple | Furnaces, older industrial PLCs |
| Type T — Copper / Constantan | +43 | whole couple | Cryogenic & low temperature, food and laboratory |
| Type K — Chromel / Alumel | +41 | whole couple | The workhorse: ovens, engines, kilns, 3D printers |
| Type N — Nicrosil / Nisil | +33 | whole couple | High temperature, lower drift than type K |
| Type S — Pt / Pt-Rh10% | +6.5 | whole couple | Precious metal, up to ~1600 °C, calibration labs |
| Type B — Pt-Rh30% / Pt-Rh6% | +15 | couple, ~1000 °C | Very high temperature; near-zero output at 0 °C |
| n-type Bi₂Te₃ leg | −215 | one leg, 300 K | Peltier coolers & near-room-temperature generators |
| p-type Bi₂Te₃ leg | +215 | one leg, 300 K | Peltier coolers & near-room-temperature generators |
| Bi₂Te₃ p–n couple | +420 | couple (2 legs) | One couple of a TEC / TEG module |
| n-type PbTe leg | −230 | one leg, ~600 K | Mid-temperature waste-heat recovery |
| p-type SiGe leg | +200 | one leg, 300–900 K | Space RTGs, high-temperature staged modules |
| Constantan vs. copper | −43 | relative to Cu | Strain gauges, thermal shunts |
| PEDOT:PSS (polymer) | −20 | one leg, 300 K | Flexible and wearable energy harvesting |
Indicative values only. The Seebeck coefficient is strongly temperature dependent (and composition dependent after doping or annealing), so treat these as starting points — use the manufacturer datasheet, or NIST / ITS-90 polynomial tables for anything you must calibrate against.
V = N · S · (T h − T c). The exact relation is the integral
V = N ∫T cT h S(T) dT; this page assumes S is constant across the range, so use the
value quoted at the mean temperature when the gradient is large.R int
should be the measured DC resistance of the whole device, not of a single leg.I = V / (R int + R load), P load = I² · R load,
and the matched-load maximum P max = V² / (4 · R int).η = ηCarnot · (√(1+ZT̄) − 1) / (√(1+ZT̄) − T c/T h),
with ZT̄ evaluated at (T h + T c)/2. This is the ideal single-pair result: it ignores contact
resistance and thermal shorting, so real modules land noticeably below it.T measured = T c + V / (N · S),
which is exactly what the ΔT mode reports.I · R int and simultaneously pumps heat back with the Peltier effect, which is why a generator's best
operating point is a compromise between voltage and power.