Electrical Engineering › Section K · Methods of analysis and theorems (ac) › Topic K.5

Maximum power transfer and other theorems (ac)

In dc, a load takes the most power from a source when its resistance matches the source's. In ac the answer has a twist: the load must also cancel the source's reactance. This topic finds the best load, the best you can do when the load is restricted, and three shorter theorems that round off ac network analysis.

From matched resistance to conjugate match

"Maximum power transfer" showed that a dc load receives the most power when \(R_L = R_{Th}\). For ac, replace the network driving the load by its Thevenin equivalent, \(\mathbf{E}_{Th}\) in series with \(Z_{Th} = R_{Th} + jX_{Th}\), as in "Thevenin and Norton (ac)". The load \(Z_L = R_L + jX_L\) receives \(P_L = I^2R_L\), where

I = \frac{E_{Th}}{\sqrt{(R_{Th} + R_L)^2 + (X_{Th} + X_L)^2}}

Only the load's resistance absorbs average power, so the question is how to make \(I^2R_L\) as large as possible. Two things help: making the total reactance zero, which maximises I, and then choosing \(R_L\) well.

The maximum power transfer theorem (ac)

First make \(X_L = -X_{Th}\), which removes the reactance term and leaves a purely resistive loop. The remaining choice is the dc problem, solved by \(R_L = R_{Th}\). Together:

Z_L = Z_{Th}^* = R_{Th} - jX_{Th}

The load impedance must be the complex conjugate of the Thevenin impedance: equal magnitude, opposite angle. The loop is then purely resistive, \(Z_T = 2R_{Th}\), so the source sees a power factor of 1, and the current is \(E_{Th}/(2R_{Th})\). The maximum power is

P_{max} = \frac{E_{Th}^2}{4R_{Th}}

where \(E_{Th}\) is the rms magnitude.

Worked example: a conjugate match

The network in "Thevenin and Norton (ac)" had \(\mathbf{E}_{Th} = 16\angle -36.9^\circ\) V and \(Z_{Th} = 3.84 + j1.12\,\Omega\). 1. Best load: \(Z_L = 3.84 - j1.12\,\Omega\), a resistor in series with a capacitor. 2. Maximum power: \(P_{max} = 16^2/(4 \times 3.84) = 16.7\) W. 3. Compare other loads. A plain 5 Ω resistor received 16.1 W, and a 4 Ω resistor, equal to \(|Z_{Th}|\), receives

P = \left(\frac{16}{|7.84 + j1.12|}\right)^2 \times 4 = 16.3\,\text{W}

4. Check: both resistive loads fall short of 16.7 W, as they must, though not by much here, because \(Z_{Th}\) has only a small reactance.

When the load is restricted

Often the load cannot be anything you like. Two cases are common. If the load must be a pure resistance, the best value is the magnitude of the Thevenin impedance:

R_L = |Z_{Th}| = \sqrt{R_{Th}^2 + X_{Th}^2}

If the load's reactance is fixed but its resistance can be adjusted, first bring the reactances as close to cancelling as allowed, then set

R_L = \sqrt{R_{Th}^2 + (X_{Th} + X_L)^2}

with each reactance positive if inductive and negative if capacitive. In both cases the power is less than \(P_{max}\), and you find it from \(I^2R_L\) with the actual total impedance.

Worked example: a resistor-only load

A source has \(\mathbf{E}_{Th} = 20\angle 0^\circ\) V and \(Z_{Th} = 8 + j6\,\Omega\). 1. A conjugate load, \(8 - j6\,\Omega\), would take \(P_{max} = 20^2/32 = 12.5\) W. 2. If the load must be a resistor, use \(R_L = |8 + j6| = 10\,\Omega\). Then \(I = 20/|18 + j6| = 1.05\) A and \(P = 1.054^2 \times 10 = 11.1\) W. 3. If instead the load has a fixed capacitive reactance of 2 Ω, set \(R_L = \sqrt{8^2 + (6 - 2)^2} = 8.94\,\Omega\), and the load receives 11.8 W. 4. Check: trying 8 Ω or 12 Ω as a resistor-only load gives 11.0 W, both below the 11.1 W at 10 Ω.

Matching, efficiency and where it matters

At a conjugate match the load resistance equals the Thevenin resistance, so the same current flows through both and they dissipate equal power: in the Thevenin model the efficiency is only 50%. Power systems therefore never operate at maximum power transfer; utilities want efficiency, with source impedances far smaller than load impedances. Matching matters where signals are weak and every milliwatt counts: radio receivers, antennas, transmission lines and audio stages. Radio-frequency equipment, for example, is commonly built around a 50 Ω standard so that sources, cables and loads match. The topic "Reflected impedance and impedance matching" shows how transformers achieve a match between unequal impedances.

Substitution and reciprocity

The substitution theorem says that if the voltage across a branch and the current through it are known, the branch can be replaced by any other combination of elements with the same voltage and current, without changing the rest of the network. A branch carrying \(2\angle 0^\circ\) A with \(10\angle 30^\circ\) V across it could become an impedance of \(5\angle 30^\circ\,\Omega\), or a source of \(10\angle 30^\circ\) V. The reciprocity theorem applies to linear, bilateral networks with a single independent voltage source and no dependent sources: if the source in branch A produces current \(\mathbf{I}\) in branch B, moving the same source to branch B, and replacing it in A by a short circuit, produces the same current \(\mathbf{I}\) in branch A.

Millman's theorem (ac)

Several voltage sources, each with a series impedance, connected in parallel between the same two nodes, combine into one equivalent source:

\mathbf{E}_{eq} = \frac{\sum \mathbf{E}_k Y_k}{\sum Y_k}, \quad Z_{eq} = \frac{1}{\sum Y_k}

where \(Y_k = 1/Z_k\) is each branch's admittance. A branch with an impedance but no source is included with \(\mathbf{E}_k = 0\). The theorem is really a source conversion of each branch to a current source, a sum of currents and admittances, and a conversion back. Signs and angles matter: taking node x as the + reference, a source whose + terminal points towards node y enters with a minus sign.

Worked example: Millman's theorem

Two branches join nodes x and y, each source with its + terminal towards x: \(\mathbf{E}_1 = 10\angle 0^\circ\) V in series with 2 Ω, and \(\mathbf{E}_2 = 10\angle 90^\circ\) V in series with an inductive reactance of 2 Ω. 1. Admittances: \(Y_1 = 0.5\) S and \(Y_2 = -j0.5\) S, so \(\sum Y = 0.5 - j0.5 = 0.707\angle -45^\circ\) S. 2. Numerator: \((10)(0.5) + (j10)(-j0.5) = 5 + 5 = 10\). 3. Result: \(\mathbf{E}_{eq} = 10/0.707\angle -45^\circ = 14.1\angle 45^\circ\) V, in series with \(Z_{eq} = 1 + j1\,\Omega\). 4. Check: the loop current is \((\mathbf{E}_1 - \mathbf{E}_2)/(2 + j2) = -j5\) A, and \(\mathbf{E}_1 - (-j5)(2) = 10 + j10\) V, which is \(14.1\angle 45^\circ\) V.

Common mistakes to avoid

1. Setting \(Z_L = Z_{Th}\) instead of its conjugate, which doubles the reactance rather than cancelling it. 2. Using \(|Z_{Th}|\) in \(E_{Th}^2/(4R_{Th})\): only the resistance belongs there. 3. Choosing \(R_L = R_{Th}\) for a resistor-only load; the best value is \(|Z_{Th}|\). 4. Assuming maximum power means maximum efficiency: it means 50%. 5. Applying reciprocity to a network with dependent sources. 6. Dropping the angles of the sources, or a source's sign, in Millman's theorem.

Key takeaways

A load receives maximum average power when \(Z_L = Z_{Th}^* = R_{Th} - jX_{Th}\), and then \(P_{max} = E_{Th}^2/(4R_{Th})\) at a power factor of 1 and 50% efficiency. A resistor-only load does best at \(R_L = |Z_{Th}|\); with a fixed load reactance, \(R_L = \sqrt{R_{Th}^2 + (X_{Th} + X_L)^2}\). Substitution swaps a branch for anything with the same voltage and current. Reciprocity lets a single source and its response trade places in a linear bilateral network. Millman's theorem merges parallel sources: \(\mathbf{E}_{eq} = \sum \mathbf{E}_kY_k/\sum Y_k\).

Flashcards

Flashcard

Maximum power transfer (ac)

Z_L = Z_Th*, the complex conjugate of the Thevenin impedance

R_L = R_Th and X_L = −X_Th.

Test your understanding

Question 1 of 5

A network has Z_Th = 6 − j8 Ω. What load impedance receives the maximum average power?

Question 2 of 5

A source has E_Th = 12 V rms and Z_Th = 3 + j4 Ω. What is the maximum average power it can deliver to a load?

Question 3 of 5

A network with Z_Th = 6 + j8 Ω must drive a purely resistive load. Which resistance takes the most power?

Question 4 of 5

Two branches are in parallel between nodes x and y: 12∠0° V with 4 Ω, and 12∠90° V with 4 Ω, both with + towards x. What is Millman's equivalent source?

Question 5 of 5

In a linear bilateral network with one voltage source and no dependent sources, the source in branch A drives 2 mA through branch B. The same source is moved to branch B. What flows in branch A?

← Previous topicNext topic →