The sum of voltage times current over all the branches of a circuit is always zero
In any circuit with \(b\) branches, if the voltage \(v_k\) and the current \(i_k\) of each branch are taken with associated references (the current enters through the terminal marked as positive for the voltage), then:
It was published in 1952 by Bernard Tellegen, a Dutch engineer at the Philips laboratories (the inventor of the pentode and the gyrator). Its strength lies in how little it requires: it only uses Kirchhoff's laws. It does not matter what each branch contains: resistors, sources, diodes, transistors, time-varying elements… It holds for linear and nonlinear circuits, in DC, in AC and at any instant.
The theorem goes beyond energy conservation. If two circuits have the same graph (the same branches joined to the same nodes, with the same references), even if the elements are different, the voltages of one and the currents of the other also satisfy:
And the same with voltages and currents taken at different instants. These products are not physical powers at all: it is a purely topological property.
A reference node is chosen and the potential of each node is called \(e_n\). The voltage of a branch going from node \(m\) to node \(n\) is \(v_k = e_m - e_n\) (this already guarantees the voltage law). Substituting and grouping by nodes:
The bracket is zero at every node because of Kirchhoff's current law. At no step has the relation between the voltage and current of each element been used; that is why the \(v\) and the \(i\) can come from different circuits with the same graph.
The Millman's theorem circuit: 12 V and 6 V batteries with internal resistances of 2 Ω and 3 Ω and a 6 Ω load. The node voltage is 8 V; battery 1 delivers 2 A and battery 2 absorbs 0.667 A.
| Branch | \(v_k\) (V) | \(i_k\) (A), entering at + | \(v_k\, i_k\) (W) |
|---|---|---|---|
| E₁ | 12 | −2 | −24 (delivers) |
| R₁ | 4 | 2 | 8 |
| E₂ | 6 | 0.667 | 4 (charging) |
| R₂ | 2 | 0.667 | 1.333 |
| R₃ | 8 | 1.333 | 10.667 |
| Sum | 0 |
With a single circuit, Tellegen simply says that the power delivered by the sources equals the power absorbed by the other elements.
The load is changed to 8 Ω. The new circuit has the same graph and its currents are \(i_1'' = 1.826\) A, \(i_2'' = 0.783\) A (battery 2 is still charging) and \(i_3'' = 1.043\) A. Multiplying the voltages of the circuit with 6 Ω by the currents of the circuit with 8 Ω:
| Branch | \(v_k'\) (V), 6 Ω load | \(i_k''\) (A), 8 Ω load | \(v_k'\, i_k''\) |
|---|---|---|---|
| E₁ | 12 | −1.826 | −21.913 |
| R₁ | 4 | 1.826 | 7.304 |
| E₂ | 6 | 0.783 | 4.696 |
| R₂ | 2 | 0.783 | 1.565 |
| R₃ | 8 | 1.043 | 8.348 |
| Sum | 0 |
The sum is zero again, even though none of those products is the power of any real element.
Related: Kirchhoff's laws · Boucherot's theorem · compensation theorem