THE SEEBECK EFFECT ·

A temperature difference across a material drives a voltage: heat turned straight into electricity. The open-circuit signal is V = −S ΔT for Seebeck coefficient S. A real thermocouple needs two different materials — the net signal follows (SA−SB)ΔT — and the figure of merit ZT = S²σT/κ sets how much of that heat becomes work. Rendered, not quoted.

source Th. J. Seebeck, Magnetische Polarisation der Metalle und Erze durch Temperatur-Differenz, Abh. Königl. Akad. Wiss. Berlin 1822–23, pp. 265–373 — archive.org scan amber · historical proceedings, no DOI

Blue Teambuilds & defends the effect
3

THE MODEL

A thermal gradient pushes charge carriers from hot to cold; the pile-up sets up a back-field. At open circuit the two balance and leave a measured voltage.

V = −S · ΔT ΔT = T_hot − T_cold

Thermocouple of materials A, B: only the contrast survives —

V_net = −(S_A − S_B) · ΔT

Sign of S reads the carrier: n-type < 0, p-type > 0 — the same sign the Hall voltage exposes.

5

THE LINEAGE

Heat → voltage, discovered by Seebeck, 1821. The sign of S follows the carrier type just as the-hall-effect reads n and its sign from VH. And it is the reverse of the Peltier cooler: drive a current through the same junction and it pumps heat instead of making it.

7

THE WITNESS

Re-runs the engine's checks live. Confirms while the machine is honest; flips red the instant the Tamper (6) drops the (SA−SB) contrast.

witness idle
The Machinethe thermoelectric panel
4

DATA IN  in ↓

Two materials, a hot end and a cold end.

↓
0

THE PANEL  lit

Live engine — computes from pure logic, no stored numbers.

Single-leg V = −S ΔT
—
Thermocouple V_net
—
Carrier of leg A
—
Figure of merit ZT
—
↓
8

DATA OUT  out ↓

Proven: identical materials give zero; only contrast × ΔT makes a voltage.

booting…
Red Teamattacks & breaks the effect
1

THE ADVERSARY wall

"So one hot metal bar is a battery — free electricity from any warm rod."

No. A single material's leads are the same material; the two contact potentials cancel and the meter reads 0. Voltage appears only across different S. The idealized constant-S, linear-response picture is itself amber — real S(T) varies with temperature and the true signal is ∫S(T) dT.

2

THE GRAVEYARD

Any temperature gradient in one wire yields a usable EMF. → a homogeneous wire gives no net loop voltage (Magnus' law); you need a junction of two S.

Bigger S alone means a better generator. → efficiency is ZT = S²σT/κ — high κ shorts the gradient and kills it.

Sign of the voltage is arbitrary. → sign of S is set by carrier type; n-type and p-type push opposite ways.

6

THE TAMPER

Planted void: make a couple of two identical materials read nonzero — drop the (SA−SB) contrast so heat makes power from nothing. The Witness (7) catches it.