Electrical Engineering › Section J · Series, parallel, and series-parallel ac circuits › Topic J.6

Series-parallel ac networks

Real ac networks mix series and parallel connections. No new laws are needed: you reduce the network block by block, find the source current, then work back out with Kirchhoff's laws and the divider rules, all in complex numbers.

Same method, complex numbers

In "Series-parallel and ladder networks" you solved dc networks by spotting which resistors were in series and which in parallel, collapsing them into one equivalent, and then expanding back out. The ac version is identical in structure. The only change is that every quantity is complex: impedances combine as complex numbers, Kirchhoff's laws hold for phasors, and the voltage and current divider rules carry angles. Because complex arithmetic is slower and easier to get wrong than real arithmetic, a tidy, step-by-step plan matters even more than it did in dc. Bold letters below are rms phasors.

A plan of attack

1. Redraw the network in your head as named blocks, each either a simple series string or a simple parallel group. 2. Combine the innermost blocks first, working outwards until one total impedance \(Z_T\) remains. 3. Find the source current \(\mathbf{I}_s = \mathbf{E}/Z_T\). 4. Work back inwards: use Ohm's law, the voltage divider rule or the current divider rule to find each block's voltage and current. 5. Check with a law you did not use, usually Kirchhoff's current law at a node or Kirchhoff's voltage law around a loop. Keep each intermediate value in both rectangular and polar form, since adding needs one and multiplying the other.

Worked example: reducing the network

A source \(\mathbf{E} = 100\angle 0^\circ\,\text{V}\) feeds a 12 Ω resistor in series with block B. Block B is an inductive reactance of 16 Ω (branch 2) in parallel with a branch made of a 20 Ω resistor and a 15 Ω capacitive reactance in series (branch 3). 1. Branch impedances: \(Z_2 = j16\,\Omega\) and \(Z_3 = 20 - j15 = 25\angle -36.87^\circ\,\Omega\). 2. Block B, by product over sum:

Z_B = \frac{(16\angle 90^\circ)(25\angle -36.87^\circ)}{20 + j1} = 12.8 + j15.4\,\Omega

3. Total: \(Z_T = 12 + Z_B = 24.8 + j15.4\,\Omega = 29.2\angle 31.8^\circ\,\Omega\). The network is inductive overall.

Worked example: working back out

Continue with the same network. 1. Source current: \(\mathbf{I}_s = 100\angle 0^\circ / 29.2\angle 31.8^\circ = 3.43\angle -31.8^\circ\) A. 2. Voltages: \(\mathbf{V}_{12} = 41.2\angle -31.8^\circ\) V across the resistor and \(\mathbf{V}_B = \mathbf{I}_s Z_B = 68.5\angle 18.5^\circ\) V across block B. 3. Branch currents:

\begin{aligned} \mathbf{I}_2 &= \mathbf{V}_B / Z_2 = 4.28\angle -71.5^\circ\,\text{A} \\ \mathbf{I}_3 &= \mathbf{V}_B / Z_3 = 2.74\angle 55.3^\circ\,\text{A} \end{aligned}

4. Checks: \(\mathbf{I}_2 + \mathbf{I}_3 = 3.43\angle -31.8^\circ\) A, satisfying Kirchhoff's current law, and \(\mathbf{V}_{12} + \mathbf{V}_B = 100\angle 0^\circ\) V, satisfying Kirchhoff's voltage law. Notice that the inductor branch carries 4.28 A, more than the source current.

Power in a series-parallel network

Total average power can be found two ways, which makes a good check. From the source, \(P_T = EI_s\cos\theta_T\), where \(\theta_T\) is the angle of \(Z_T\). From the elements, \(P_T\) is the sum of \(I^2R\) over every resistor, since inductors and capacitors absorb no average power. In the worked example, \(P_T = (100)(3.43)\cos 31.8^\circ = 292\) W, and the resistors give \(3.43^2 \times 12 + 2.74^2 \times 20 = 292\) W, an exact match when unrounded values are used. The power factor is \(\cos 31.8^\circ = 0.850\) lagging. Branch currents larger than the source current, as here, are normal wherever reactive currents partly cancel.

Ladder networks

A ladder alternates series impedances (the rails) and shunt impedances (the rungs), ending at a load. There are two good routes. The first reduces from the far end: combine the load with the last series impedance, put that in parallel with the last rung, and repeat back to the source, then work forwards with the divider rules. The second, the unit-current method, runs the other way: assume a convenient load current, such as \(1\angle 0^\circ\) A, and work back to the source voltage that current would need. Because the network is linear, every current and voltage then scales by the ratio of the actual source to that computed source.

Worked example: a ladder network

A source \(\mathbf{E} = 24\angle 0^\circ\,\text{V}\) feeds a 4 Ω series resistor, then a rung of \(-j8\,\Omega\), then a series \(6 + j3\,\Omega\), ending in a 10 Ω load. Find the load current. 1. Far end: \(Z_a = 6 + j3 + 10 = 16 + j3\,\Omega\). 2. With the rung: \(Z_b = (-j8)(16 + j3)/(16 - j5) = 3.64 - j6.86\,\Omega\). 3. Total: \(Z_T = 7.64 - j6.86 = 10.3\angle -41.9^\circ\,\Omega\), so \(\mathbf{I}_s = 2.34\angle 41.9^\circ\) A. 4. Rung voltage: \(\mathbf{V}_b = \mathbf{I}_s Z_b = 18.2\angle -20.1^\circ\) V, so \(\mathbf{I}_L = \mathbf{V}_b / Z_a = 1.12\angle -30.7^\circ\) A. 5. Check by unit current: 1 A in the load needs \(21.5\angle 30.7^\circ\) V at the source, and \(24/21.5\angle 30.7^\circ\) gives the same \(1.12\angle -30.7^\circ\) A.

Application: an R-C phase-shift ladder

Ladders are used on purpose in electronics. Consider three identical sections, each a series capacitor C followed by a resistor R to ground, with the output taken across the last resistor and nothing connected to it. Each section shifts the phase, and at one frequency,

f_0 = \frac{1}{2\pi\sqrt{6}\,RC}

the total shift is exactly 180° and the output is \(1/29\) of the input. With R = 10 kΩ and C = 10 nF, \(f_0 = 650\) Hz. An inverting amplifier with a gain of 29, and an input that does not load the ladder, can sustain an oscillation at \(f_0\) when connected around it: the idea behind the phase-shift oscillator. Note that the sections load each other, so the shift is not simply 60° per section.

Common mistakes to avoid

1. Adding a parallel block's impedances as if they were in series: 8 Ω in parallel with j8 Ω is \(4 + j4\,\Omega\), not \(8 + j8\,\Omega\). 2. Adding magnitudes instead of complex numbers when combining blocks. 3. Using the voltage divider across elements that do not carry the same current. 4. Rounding heavily early: complex steps compound errors, so carry four figures and round at the end. 5. Expecting no branch current to exceed the source current. 6. Forgetting to scale every quantity, magnitude and angle, after the unit-current method.

Key takeaways

Series-parallel ac networks are solved exactly like dc ones: combine blocks into \(Z_T\), find \(\mathbf{I}_s = \mathbf{E}/Z_T\), then work back with Ohm's law, the divider rules and Kirchhoff's laws, all with phasors. Check with a law you have not used. Total power is \(EI_s\cos\theta_T\) and also the sum of \(I^2R\) over the resistors. Ladders reduce from the far end, or by assuming a load current and scaling. An R-C phase-shift ladder of three sections shifts 180° at \(f_0 = 1/(2\pi\sqrt{6}RC)\), with an output of \(1/29\) of the input.

Flashcards

Flashcard

Series-parallel ac method

Reduce blocks to Z_T, find I_s = E/Z_T, then work back out

The dc method with complex numbers.

Test your understanding

Question 1 of 5

A 6 Ω resistor is in series with a block made of an 8 Ω resistor in parallel with an inductive reactance of 8 Ω. What is the total impedance?

Question 2 of 5

A source current of 3∠0° A divides between two branches. One branch carries 2∠−60° A. What does the other carry?

Question 3 of 5

In the unit-current method, a load current of 1∠0° A needs a source of 20∠30° V. The actual source is 50∠0° V. What is the actual load current?

Question 4 of 5

In a series-parallel network, the only resistors are a 5 Ω resistor carrying 2 A and an 8 Ω resistor carrying 1.5 A. What is the total average power from the source?

Question 5 of 5

A three-section R-C phase-shift ladder is driven at f₀ = 1/(2π√6 RC). What does its unloaded output look like?

← Previous topicNext topic →