Electrical Engineering › Section K · Methods of analysis and theorems (ac) › Topic K.3

Superposition (ac)

A linear network with several sources responds to all of them at once exactly as the sum of its responses to each one alone. In ac work that idea does double duty: it splits a problem with several sources, and it is the only correct way to handle sources at different frequencies, including a dc bias with an ac signal on top.

One source at a time

"The superposition theorem" showed that in a linear dc network any current or voltage is the algebraic sum of the contributions from each independent source acting alone. The same holds for ac networks of resistors, inductors, capacitors and linear dependent sources, because every element obeys a linear relationship between its phasor voltage and current. For sources at a single frequency, the contributions are phasors and add as complex numbers. Superposition is most useful when a network has two or three sources and you want one quantity: each sub-problem is often a simple series-parallel network, solvable with the divider rules. Bold letters are rms phasors.

Removing the other sources

To find the contribution of one independent source, set every other independent source to zero. A zeroed ideal voltage source becomes a short circuit, since its voltage is zero whatever the current. A zeroed ideal current source becomes an open circuit, since its current is zero whatever the voltage. Any internal impedance a real source has stays in the network: a generator with an internal impedance of \(50\,\Omega\) leaves \(50\,\Omega\) behind. Then solve the reduced network for the quantity you need, keeping track of its reference direction or polarity, and repeat for each source. The final answer is the phasor sum of the contributions.

Worked example: two sources, one frequency

Source \(\mathbf{E} = 10\angle 0^\circ\) V has its + terminal at node a and − at ground. A 5 Ω resistor joins a to node b, an inductive reactance of 5 Ω joins b to ground, and a current source \(\mathbf{I} = 2\angle 90^\circ\) A pushes current from ground into b. Find \(\mathbf{V}_b\). 1. E alone, current source open: \(\mathbf{V}' = 10\,(j5)/(5 + j5) = 7.07\angle 45^\circ\) V. 2. I alone, E shorted, so 5 Ω and j5 Ω are in parallel: \(\mathbf{V}'' = (2\angle 90^\circ)(3.54\angle 45^\circ) = 7.07\angle 135^\circ\) V. 3. Sum: \(\mathbf{V}_b = 7.07\angle 45^\circ + 7.07\angle 135^\circ = 10\angle 90^\circ\) V. 4. Check by nodal analysis: \(\mathbf{V}_b(0.2 - j0.2) = 2 + j2\) also gives \(10\angle 90^\circ\) V.

Dependent sources stay in

A dependent source is never switched off during superposition, because its value is not an independent input: it is set by a voltage or current inside the network. Leave it in every sub-circuit, and let its controlling variable take whatever value that sub-circuit gives it. In effect, a dependent source behaves as part of the network's structure, like an extra impedance, while only the independent sources take turns. If a network's only independent source is the one you are analysing, the dependent sources still act on its response. Forgetting this rule gives contributions that do not add up to the true answer.

Worked example: with a dependent source

Node b connects to a source \(\mathbf{E} = 4\angle 0^\circ\) V through 2 Ω, to ground through an inductive reactance of 2 Ω, and to ground through a dependent current source drawing \(0.25\mathbf{V}_b\) A out of b. An independent source \(\mathbf{I} = 1\angle 0^\circ\) A pushes current into b. 1. I alone, E shorted: Kirchhoff's current law gives \(\mathbf{V}'(0.5 - j0.5 + 0.25) = 1\), so \(\mathbf{V}' = 1.11\angle 33.7^\circ\) V. 2. E alone, I open: \(\mathbf{V}''(0.75 - j0.5) = 4/2 = 2\), so \(\mathbf{V}'' = 2.22\angle 33.7^\circ\) V. 3. Sum: \(\mathbf{V}_b = 3.33\angle 33.7^\circ\) V. 4. Check: solving with both sources at once gives \(3/(0.75 - j0.5)\), the same value. The 0.25 S term appears in both steps.

Sources at different frequencies

Phasors only make sense at one frequency, so contributions at different frequencies can never be added as phasors. Instead, solve the network separately at each frequency, using the reactances at that frequency, convert each result to a time function, and add the time functions. A dc source is simply a source at zero frequency: in its sub-circuit, inductors act as shorts and capacitors as opens, in the steady state. This is the standard way to analyse amplifiers and sensors, where a dc bias sets the operating point and a small ac signal rides on top. Each frequency can give a very different picture of the same network.

Worked example: dc bias plus an ac signal

A 12 V dc source in series with a 1 V, 1 kHz ac source feeds a 1 kΩ resistor to node b. From b to ground are a 1 kΩ resistor and a 1 µF capacitor. Find \(v_b(t)\). 1. Dc alone, capacitor open: \(V_b = 12 \times 1/(1 + 1) = 6\) V. 2. Ac alone, dc source shorted: \(X_C = 159\,\Omega\), so 1 kΩ in parallel with \(-j159\,\Omega\) is \(157\angle -81.0^\circ\,\Omega\), and the divider gives \(\mathbf{V}_b = 0.152\angle -72.3^\circ\) V. 3. Combine in time: with the ac source \(1.41\sin\omega t\) V,

v_b(t) = 6 + 0.215\sin(6283t - 1.26)\,\text{V}

where the phase is in radians (1.26 rad is 72.3°). 4. Check: without the capacitor the ac part would be 0.5 V; the capacitor, acting as a bypass capacitor, cuts it by about 70% while leaving the 6 V bias alone.

Power does not superpose

Superposition applies to currents and voltages, not to power, because power depends on the square of current. Suppose two 60 Hz sources drive in-phase contributions of 3 A and 4 A through a 2 Ω resistor. The true current is 7 A, so \(P = 7^2 \times 2 = 98\) W, not \(3^2 \times 2 + 4^2 \times 2 = 50\) W. Always add the currents first, then compute power. There is one exception: when the contributions are at different frequencies, or one is dc, their average powers do add, because the cross terms average to zero. That is why, as "Average and rms values" showed, the rms value of a dc level plus a sinusoid is \(\sqrt{V_{dc}^2 + V_{ac}^2}\).

Common mistakes to avoid

1. Opening a voltage source or shorting a current source; it is the other way round. 2. Removing a source's internal impedance along with the source. 3. Zeroing a dependent source. 4. Adding the magnitudes of contributions instead of the phasors: 7.07 V plus 7.07 V at 90° apart is 10 V, not 14.1 V. 5. Adding phasors found at different frequencies, or adding a dc value to an rms phasor, instead of adding time functions. 6. Superposing powers of contributions at the same frequency.

Key takeaways

In a linear ac network, any voltage or current is the sum of the contributions from each independent source acting alone. Zero the others: voltage sources become shorts, current sources become opens, internal impedances stay, and dependent sources always stay active. At one frequency, add the contributions as phasors. At different frequencies, including dc with ac, solve each frequency with its own reactances and add the time functions. Power is not superposable for contributions at the same frequency; find the total current first.

Flashcards

Flashcard

Superposition (ac)

A response equals the sum of the responses to each independent source acting alone

Valid for linear networks.

Test your understanding

Question 1 of 5

When you find the contribution of one source by superposition, what happens to the other sources?

Question 2 of 5

Two sources at the same frequency contribute 7.07∠45° V and 7.07∠135° V to a node voltage. What is the node voltage?

Question 3 of 5

A node has a 6 V dc component and a 1.5 V rms sinusoidal component. What is its peak instantaneous voltage?

Question 4 of 5

A linear network is driven by a 60 Hz source and a 180 Hz source. How should the total current in one branch be found?

Question 5 of 5

Two 60 Hz sources drive in-phase contributions of 3 A and 4 A through a 2 Ω resistor. What average power does the resistor absorb?

← Previous topicNext topic →