EN

Compensation theorem

How much the currents of a circuit change when a resistance or an impedance varies

1. What it is for

Often a circuit has already been solved and then the value of a component changes: a resistor heats up, it is replaced by another with a different tolerance, or an ammeter is inserted. Solving the whole circuit again is slow. The compensation theorem (also called the variation theorem) directly gives the change in every current and voltage caused by that change, by solving a much simpler circuit.

2. Statement

If, in a branch carrying a current \(I\), the resistance (or impedance) goes from \(R\) to \(R + \Delta R\), the changes in all the currents of the circuit are those produced, acting alone, by a compensation voltage source

\[ V_c = I\,\Delta R \]\[ \overline{V}_c = \overline{I}\,\Delta\overline{Z}\ \ \text{(AC)} \]

placed in series with the already modified branch (\(R + \Delta R\)) and with its polarity opposing the original current \(I\), while all the independent sources of the circuit are set to zero (voltage sources short-circuited and current sources open-circuited). The final current in each branch is the original one plus the change.

Original circuit: current \(I_3\) through \(R_3\).
Compensation circuit: sources set to zero, modified branch and \(V_c = I_3\,\Delta R\) opposing \(I_3\).

Seen from the branch that changes, the rest of the circuit is its Thévenin equivalent \(R_{th}\), so the change in the branch itself takes one line:

\[ \Delta I = -\frac{I\,\Delta R}{R_{th} + R + \Delta R} \]\[ I' = I\,\frac{R_{th} + R}{R_{th} + R + \Delta R} \]

Dual version: if the admittance of a branch with a voltage \(V\) across it changes by \(\Delta G\), the changes are those of a current source \(I_c = V\,\Delta G\) in parallel with the branch.

3. Why it works

If the original circuit is subtracted from the modified one (which superposition allows), the independent sources cancel out and each branch is left with only the change \(\Delta I = I' - I\). In the branch that changes, the voltage drop goes from \(R\,I\) to \((R + \Delta R)\,I'\), and the difference is

\[ (R + \Delta R)\,I' - R\,I = (R + \Delta R)\,\Delta I + \Delta R\,I \]

That is: the modified branch carrying \(\Delta I\) plus a fixed voltage \(\Delta R\,I\), which is exactly the compensation source. Since it relies on superposition, the theorem only holds for linear circuits.

Example 1: change of a resistance

In the circuit of the figure, with \(E_1 = 12\) V, \(R_1 = 2\,\Omega\), \(E_2 = 6\) V, \(R_2 = 3\,\Omega\) and \(R_3 = 6\,\Omega\), Millman's theorem gives \(V_a = 8\) V and \(I_1 = 2\) A, \(I_2 = -0.667\) A, \(I_3 = 1.333\) A. What happens if \(R_3\) becomes 8 Ω?

\[ V_c = I_3\,\Delta R = 1.333 \cdot 2 = 2.667\ \text{V} \]\[ R_1 \parallel R_2 = 1.2\,\Omega \]
\[ \Delta I_3 = -\frac{2.667}{8 + 1.2} = -0.290\ \text{A} \]\[ \Delta V_a = 0.290 \cdot 1.2 = 0.348\ \text{V} \]
\[ \Delta I_1 = -\frac{0.348}{2} = -0.174\ \text{A} \]\[ \Delta I_2 = -\frac{0.348}{3} = -0.116\ \text{A} \]

Final currents: \(I_1' = 1.826\) A, \(I_2' = -0.783\) A, \(I_3' = 1.043\) A, which satisfy \(I_1' + I_2' = I_3'\) and match those obtained by solving the circuit again with 8 Ω.

Example 2: ammeter insertion error

To measure \(I_3\), a 0.5 Ω ammeter is inserted. The ammeter is a \(\Delta R = 0.5\,\Omega\) in the branch, and the resistance it sees is \(R_{th} + R_3 = 1.2 + 6 = 7.2\,\Omega\):

\[ I_3' = 1.333 \cdot \frac{7.2}{7.2 + 0.5} = 1.247\ \text{A} \]\[ \text{error} = -\frac{0.5}{7.7} = -6.5\ \% \]

The ammeter reads 6.5 % less than what was flowing before it was inserted. The rule of thumb: its resistance must be very small compared with the one the branch sees.

4. Calculator

Two-source circuit of the figure: it calculates the currents, the compensation source and the change in each current when \(R_3\) changes by \(\Delta R\).

Circuit
V
Ω
V
Ω
Ω
Ω
Negative if it decreases
BranchBeforeChange (compensation)AfterSolving again

5. Applications

6. Precautions

Related: substitution theorem · reciprocity theorem · Kirchhoff's laws