EN

Superposition theorem

In a linear circuit, the effect of several sources is the sum of the effects of each one acting alone

1. Statement

In a linear circuit with several independent sources, any voltage or current is the sum of those that each source would produce acting alone, with the others turned off:

\[ x = x_1 + x_2 + \dots + x_n \]

Each term \(x_k\) is the contribution of source \(k\). They are added with their sign, relative to one reference direction. The theorem follows from linearity: doubling a source doubles its contribution, and the contributions do not interfere with each other.

2. How to apply it

  1. Choose the quantity you are looking for and its reference direction, and keep it in every step.
  2. Keep a single independent source and turn off the others as shown in the table.
  3. Dependent sources are never turned off: they stay in each subcircuit, controlled by the variable of that subcircuit.
  4. Calculate the contribution of that source and repeat with each of the others.
  5. Add the contributions with their sign. In AC, phasors can be added only if the sources have the same frequency; otherwise, add the time functions.
ElementHow it is turned off
Ideal voltage sourceShort circuit (0 V)
Ideal current sourceOpen circuit (0 A)
Real sourceThe ideal source is turned off and its internal resistance or impedance is kept
Dependent sourceIt is not turned off
Example 1: two independent DC sources

A 12 V source with R1 = 2 Ω feeds node a, from which R2 = 4 Ω goes to ground. A 3 A current source also injects current into a. Find the voltage \(V_a\) and the current through R2.

Complete circuit.

12 V source only: the current source is replaced by an open circuit, leaving a voltage divider.

\[ V_a' = 12 \cdot \frac{4}{2 + 4} = 8\ \text{V} \]

3 A source only: the 12 V source is replaced by a short circuit and R1 ends up in parallel with R2.

\[ R_1 \parallel R_2 = \frac{2 \cdot 4}{2 + 4} = 1.333\,\Omega \]\[ V_a'' = 3 \cdot 1.333 = 4\ \text{V} \]

Sum:

\[ V_a = 8 + 4 = 12\ \text{V} \]\[ I_2 = \frac{12}{4} = 3\ \text{A} \]

Check with the complete circuit: with \(V_a = 12\) V no current flows through R1, and all 3 A of the current source go through R2. However, power cannot be added:

\[ P_2 = \frac{12^2}{4} = 36\ \text{W} \]\[ \frac{8^2}{4} + \frac{4^2}{4} = 16 + 4 = 20\ \text{W} \]

16 W are missing: the cross term \(2 V_a' V_a'' / R_2\). Power depends on the square of the voltage and is not linear: first add voltages or currents, then calculate the power.

3. Dependent sources

A dependent (controlled) source does not generate energy by itself: its value depends on a voltage or current of the circuit itself. That is why it is not turned off when applying superposition. It stays in place in each subcircuit and takes the value given by its control variable in that subcircuit.

Example 2: circuit with a dependent voltage source

A 20 V source with 2 Ω feeds node a. Between a and ground there is a 4 Ω resistor in series with a dependent voltage source of value \(2 I_x\) (in volts, with \(I_x\) in amperes), where \(I_x\) is the current through the 2 Ω resistor. A 2 A current source injects current into a.

The diamond-shaped source is the dependent one: its voltage is 2·Iₓ.

20 V source only (the 2 A source open): the same current \(I_x'\) flows through the three branches of the loop.

\[ 20 = 2\,I_x' + 4\,I_x' + 2\,I_x' \;\Rightarrow\; I_x' = 2.5\ \text{A} \]\[ V_a' = 20 - 2 \cdot 2.5 = 15\ \text{V} \]

2 A source only (the 20 V source shorted): the dependent source is still there, now with \(I_x''\). The current in the 4 Ω branch is \(I_x'' + 2\), and \(I_x''\) comes out negative because it flows from ground towards a.

\[ V_a'' = 4\,(I_x'' + 2) + 2\,I_x'' \]\[ I_x'' = -\frac{V_a''}{2} \]\[ V_a'' = 2\ \text{V} \qquad I_x'' = -1\ \text{A} \]

Sum:

\[ I_x = 2.5 - 1 = 1.5\ \text{A} \]\[ V_a = 15 + 2 = 17\ \text{V} \]

Check: \(I_x = (20 - 17)/2 = 1.5\) A, 3.5 A flow through the 4 Ω branch and its voltage is \(4 \cdot 3.5 + 2 \cdot 1.5 = 17\) V. If the dependent source had also been turned off, the result would be \(V_a = 13.33 + 2.67 = 16\) V, which is wrong.

There is another way, useful in large circuits: treat the dependent source as if it were independent, with an unknown value, apply superposition with all the sources and close the equation of its control variable at the end. It gives the same result.

4. Circuits with DC and AC at the same time

This is the most practical use of the theorem. When a circuit has both DC and AC sources, it is solved twice: in DC, inductors are short circuits and capacitors are open circuits; in AC, phasors and impedances are used. The result is the sum of both responses in time:

\[ i(t) = I_0 + \sqrt{2}\,I_1 \sin(\omega t + \varphi) \]\[ I = \sqrt{I_0^2 + I_1^2} \]\[ P = P_0 + P_1 \]

Here \(I_0\) is the DC component and \(I_1\) the RMS value of the AC one. With sources of different frequencies powers do add: the cross term between DC and AC, or between two different frequencies, has zero average over a period. What can never be done is add phasors of different frequencies directly.

Example 3: inductor fed with DC and AC in series

A 12 V DC source and a 10 V RMS, 50 Hz source in series feed a 4 Ω resistor and a 9.55 mH inductor, whose reactance at 50 Hz is \(X_L = 2\pi \cdot 50 \cdot 0.00955 = 3\,\Omega\).

DC source and AC source in series with an RL circuit.

DC only (the AC source shorted): the inductor is a short circuit and the current is limited only by the resistor.

\[ I_0 = \frac{12}{4} = 3\ \text{A} \]

AC only (the DC source shorted):

\[ \underline{Z} = 4 + j3 = 5\angle 36.87^\circ\ \Omega \]\[ I_1 = \frac{10}{5} = 2\ \text{A} \]

Sum in time, with the current lagging the AC voltage by 36.87°:

\[ i(t) = 3 + 2\sqrt{2}\,\sin(\omega t - 36.87^\circ)\ \text{A} \]
\[ I = \sqrt{3^2 + 2^2} = 3.61\ \text{A} \]\[ P = 4 \cdot 3^2 + 4 \cdot 2^2 = 36 + 16 = 52\ \text{W} \]

The current oscillates between 0.17 A and 5.83 A without reversing. A DC ammeter (moving coil) reads 3 A, a true RMS one reads 3.61 A and an AC one with a blocking capacitor reads 2 A. The power in the resistor is the sum of both: 52 W.

Example 4: DC bias and AC signal with a coupling capacitor

A divider of two 1 kΩ resistors fed from 10 V sets the DC voltage of node a, as in the biasing of a transistor. A 1 V RMS signal comes in through a capacitor whose reactance at the signal frequency is 1 kΩ.

The capacitor lets the signal through but blocks DC.

DC only (the signal shorted): the capacitor is an open circuit and node a sits at half the supply.

\[ V_{a0} = 10 \cdot \frac{1}{1 + 1} = 5\ \text{V} \]

AC only (the 10 V source shorted): the two resistors are in parallel, 500 Ω, and form a divider with the capacitor.

\[ \underline{V}_{a1} = 1 \cdot \frac{500}{500 - j1000} = 0.447\angle 63.4^\circ\ \text{V} \]

Sum in time (amplitude \(0.447 \cdot \sqrt{2} = 0.632\) V):

\[ v_a(t) = 5 + 0.632\,\sin(\omega t + 63.4^\circ)\ \text{V} \]

The voltage of node a oscillates around 5 V, between 4.37 V and 5.63 V. The capacitor keeps DC away from the signal source, and the 10 V source, which is a short circuit in AC, keeps the signal away from the supply. Any amplifier is analysed this way: first the DC operating point, then the signal.

5. When it cannot be applied

6. What it is used for

Related: Kirchhoff's laws · Millman's theorem · DC circuit analyser · AC circuit analyser, which solves circuits with sources of several frequencies