Power Supplies and Power Electronics Fundamentals

Introduction and Context

Electronic systems require electrical energy in forms that are compatible with their circuits. A utility source may provide sinusoidal alternating current (AC), whereas digital, analog, communication, and control circuits generally require regulated direct-current (DC) voltages. A power supply therefore performs one or more of the following functions: transformation, rectification, filtering, regulation, isolation, protection, and energy conversion.

Power electronics is the study and application of semiconductor devices used to control and convert electrical power. Unlike small-signal electronics, where the principal concern is voltage or current amplification, power electronics emphasizes efficiency, thermal stress, switching behavior, electromagnetic compatibility, reliability, and safe energy transfer. The same converter may be analyzed simultaneously as an electrical circuit, a control system, a thermal system, and a reliability-critical product.

A useful distinction is that a linear power supply controls output voltage by dissipating excess voltage as heat, while a switched-mode power supply (SMPS) controls energy transfer using high-frequency switching and reactive components. Linear supplies can provide very low noise, but switched supplies normally offer substantially higher efficiency and smaller size.

Learning Objectives

After completing this chapter, the learner should be able to:

• explain the functional stages of practical power supplies;
• analyze half-wave, full-wave, and bridge rectifiers;
• calculate ripple, efficiency, regulation, and capacitor requirements;
• explain the operation of linear regulators and switching converters;
• distinguish buck, boost, and buck–boost conversion;
• evaluate semiconductor switching losses, thermal limits, isolation, and protection;
• select an appropriate supply architecture for a practical application;
• assess electromagnetic, safety, reliability, and environmental constraints.

1. Power-Supply Architecture

A conventional isolated DC supply commonly contains:

1. an input protection and electromagnetic-interference (EMI) filter;
2. a transformer, when voltage conversion or galvanic isolation is required;
3. a rectifier;
4. a smoothing filter;
5. a regulator or DC–DC converter;
6. feedback, control, and protection circuits;
7. an output filter and load interface.

The overall conversion efficiency is approximately

where P_out is output power, P_in is input power, and η_i is the efficiency of each conversion stage. This product relationship assumes that stage losses are represented by their individual efficiencies.

A supply rated at 80% efficiency does not deliver 80% of its voltage; it converts 80% of its input power into useful output power. The remaining power becomes heat, acoustic energy, leakage, or other losses.

Important terminology

• Line regulation: change in output voltage caused by a change in input voltage, with load held constant.
• Load regulation: change in output voltage caused by a change in load current, with input held constant.
• Ripple: periodic residual AC voltage superimposed on DC output.
• Dropout voltage: minimum input-to-output voltage difference required by a regulator.
• Hold-up time: time for which output remains within specification after input power is interrupted.
• Galvanic isolation: absence of a conductive DC path between input and output.
• Power factor: ratio of real input power to apparent input power.
• Crest factor: ratio of peak current to RMS current.

2. Rectification and Filtering

2.1 Half-wave rectifier

A single diode conducts during one polarity of the AC input. For an ideal sinusoidal source with peak voltage V_m and resistive load R_L, the average output voltage is

and the RMS output voltage is

These expressions assume an ideal diode and an unfiltered resistive load. A half-wave rectifier has poor transformer utilization and large ripple, so it is used mainly in low-power or noncritical applications.

2.2 Full-wave rectifier

A full-wave circuit uses both half-cycles. A center-tapped rectifier requires a center-tapped transformer; a bridge rectifier uses four diodes and does not require a center tap. For an ideal full-wave rectifier:

The ripple frequency is twice the AC supply frequency:

For a bridge rectifier, two diodes conduct in series during each half-cycle. If each conducting diode has forward drop V_F, the approximate capacitor-charged peak is

This approximation neglects transformer resistance, diode dynamic resistance, and capacitor discharge during conduction.

2.3 Capacitor-input filter

A capacitor connected after a rectifier charges near the waveform peak and discharges through the load between peaks. For sufficiently small ripple, the peak-to-peak ripple is approximately

where I_L is load current in amperes, C is capacitance in farads, and f_r is ripple frequency in hertz. For a full-wave rectifier supplied from 50 Hz, f_r = 100\ Hz. The approximation assumes nearly constant load current and relatively small ripple compared with DC voltage.

The approximate average DC output is

The capacitor must be rated for both voltage and ripple current. A practical design includes tolerance, temperature effects, aging, and a voltage margin.

Worked Example 1: Smoothing-capacitor selection

A bridge rectifier is supplied by a 12 V RMS, 50 Hz transformer secondary. Assume each conducting diode drops 0.8 V, the load current is 0.20 A, and the permitted ripple is 1.0 V peak-to-peak.

The secondary peak is

The capacitor-charged peak is approximately

The ripple frequency is 100 Hz. Therefore,

Thus C ≥ 2000\ μF. A standard 2200 μF capacitor would be a reasonable nominal choice, provided its voltage, ripple-current, temperature, and lifetime ratings are satisfactory. The approximate average voltage is 14.87 V before regulator losses and transformer regulation are considered.

3. Performance Measures

Ripple factor

Ripple factor is the ratio of RMS AC ripple to DC output:

For an unfiltered ideal half-wave rectifier, r ≈ 1.21, and for an unfiltered ideal full-wave rectifier, r ≈ 0.482. Filtering reduces these values.

Rectifier efficiency

Rectifier efficiency is

For ideal resistive rectifiers, the maximum theoretical efficiencies are approximately 40.6% for half-wave and 81.2% for full-wave operation. Real values are lower because of diode, transformer, wiring, and switching losses.

Regulation

One common load-regulation definition is

where both voltages are measured under the same input conditions. Some manufacturers use V_no-load in the denominator, so the definition must be stated when comparing specifications.

Line regulation may be expressed as

or in millivolts per volt. Datasheet conventions differ; engineering comparison requires a common definition.

4. Linear Voltage Regulators

A linear regulator operates a transistor in its active region as a controlled series element. A feedback amplifier compares a sampled output with a stable reference and adjusts transistor conduction.

For a simple regulator, the dissipated power is approximately

and ideal efficiency is

when regulator quiescent current is neglected. These relations assume a series regulator with approximately constant output voltage.

A low-dropout regulator (LDO) is designed to operate with a small input-output difference. Its minimum input must satisfy

A regulator may also have current limiting, thermal shutdown, short-circuit protection, reverse-current blocking, and soft start. Capacitor equivalent series resistance (ESR) and stability requirements are often critical; an arbitrary output capacitor can cause oscillation in some regulators.

Thermal design

If a regulator dissipates P_D, its junction temperature can be estimated by

where T_J is junction temperature in °C, T_A is ambient temperature in °C, and θ_JA is junction-to-ambient thermal resistance in °C/W. The estimate assumes steady-state operation and a specified mounting arrangement. Safe design requires T_J to remain below the manufacturer’s maximum, with derating for uncertainty and transient conditions.

5. Switched-Mode Power Supplies

An SMPS rapidly switches a semiconductor between low-voltage/high-current and high-voltage/low-current states. Ideally, power dissipation during both states is small, although real devices experience conduction, switching, gate-drive, magnetic, capacitor, diode, and control losses.

A typical isolated SMPS includes a rectifier, switching transistor, high-frequency transformer, secondary rectifier, output filter, feedback controller, and isolation barrier. Isolation feedback may use an optocoupler or an isolated sensing device.

5.1 Pulse-width modulation

In PWM control, the duty ratio is

where t_on is switch on-time, T_s is switching period, and f_s is switching frequency. In ideal converters, changing D controls the average energy transferred to the load.

Higher switching frequency can reduce magnetic and filter size, but increases switching loss, electromagnetic interference, and gate-drive demands. Frequency selection is therefore an optimization rather than a simple maximization problem.

5.2 Non-isolated converters

For an ideal continuous-conduction buck converter:

For an ideal continuous-conduction boost converter:

For an ideal inverting buck–boost converter:

These relationships assume ideal switches, ideal diodes, negligible inductor and capacitor losses, steady state, and continuous inductor current. They do not directly apply in discontinuous-conduction mode.

The inductor volt-second balance explains the buck relationship. In steady state, the average inductor voltage over one switching period is zero:

which gives V_out = D V_in. This derivation is valid for the ideal buck converter in continuous conduction.

Worked Example 2: Ideal buck converter

An ideal buck converter reduces 24 V to 12 V in continuous conduction. From V_out = D V_in,

The switch must therefore be on for 50% of each switching period. If the switching frequency is 100 kHz, the period is 10 μs and the on-time is 5 μs. Practical duty ratio must be increased slightly to compensate for switch and diode voltage drops.

6. Semiconductor Devices in Power Conversion

• Power diode: uncontrolled unidirectional device; important ratings include forward current, reverse voltage, reverse-recovery charge, and thermal resistance.
• Power MOSFET: voltage-controlled, fast-switching device, especially effective at low and medium voltage. Conduction loss is approximately P_cond = I_RMS² R_DS(on) when resistance is the dominant mechanism.
• IGBT: combines MOS-gate control with bipolar conduction; useful at higher voltage and power, generally with slower switching than a MOSFET.
• Thyristor/SCR: latches on after triggering and turns off when current falls below its holding condition or by commutation; valuable in controlled rectifiers and high-power AC applications.
• SiC MOSFET and Schottky diode: wide-bandgap devices enabling higher temperature, voltage, and frequency operation, with reduced reverse-recovery effects in suitable applications.
• GaN transistor: supports very fast switching and low charge, but demands careful layout, gate driving, and control of parasitic inductance.

Switch ratings must include voltage overshoot, current surge, avalanche energy, short-circuit withstand time, and safe operating area—not merely nominal RMS values.

7. Control, Feedback, and Stability

A regulated converter compares a measured output with a reference and modifies duty ratio, frequency, or phase. The feedback loop has a power stage, sensor, error amplifier, compensation network, modulator, and delay. Stability depends on loop gain and phase margin.

A simplified feedback relation is

where V_ref is reference voltage and k_s is the sensing fraction. The controller attempts to reduce error, but excessive gain or poorly designed compensation can produce oscillation or unacceptable transient response.

Design evaluation commonly includes:

• line and load transients;
• startup and shutdown behavior;
• control-loop stability;
• output impedance;
• overshoot and undershoot;
• current-limit response;
• operation in continuous and discontinuous conduction.

8. Protection, Safety, and Compliance

A professional supply must address abnormal as well as nominal operation. Essential protections include input fuse or circuit breaker, inrush-current limiting, overvoltage protection, overcurrent or short-circuit protection, thermal shutdown, undervoltage lockout, surge suppression, and reverse-polarity protection where applicable.

Mains-connected systems require reinforced insulation or appropriate basic and supplementary insulation, protective earthing where applicable, creepage and clearance distances, fire-resistant materials, and safe discharge of hazardous capacitors. Regulatory evaluation may involve standards for electrical safety, conducted and radiated emissions, immunity, energy efficiency, and environmental substances. Exact requirements depend on product category and jurisdiction; compliance should be verified against the current applicable standards rather than assumed from circuit similarity.

Power-factor correction (PFC) is important in higher-power AC-input supplies. A bridge and capacitor input draw narrow current pulses, producing poor displacement-corrected power factor and harmonic current. Passive or active PFC broadens the input current waveform and can reduce harmonic distortion. In an ideal sinusoidal system with unity displacement factor, real power is

where PF is power factor. This equation applies to sinusoidal RMS voltage and current definitions; with distorted waveforms, PF includes both displacement and distortion effects.

9. Applied Case: Isolated Supply for an Industrial Sensor Node

Background. An industrial sensor node requires an isolated 5 V, 1 A output from a nominal 24 V DC control bus. It must tolerate bus variation, electrical transients, continuous operation, and limited enclosure ventilation.

Problem. A linear regulator would dissipate approximately

at nominal input, before accounting for quiescent current. This is thermally unattractive. A non-isolated buck converter would be efficient but would not satisfy the isolation requirement.

Analysis. An isolated flyback converter is a suitable candidate at this power level. The design must establish transformer turns ratio, maximum duty ratio, switching frequency, primary MOSFET voltage stress, secondary rectifier stress, magnetizing inductance, leakage-energy clamp, control-loop compensation, and isolation construction. A nominal efficiency of 85% would imply approximately

and total loss of about 0.88 W. This is much easier to remove than 19 W, though local hot spots can still be significant.

Approach. A practical design would include an input fuse, transient suppressor, EMI filter, controlled startup, current-mode control, properly rated transformer insulation, an RCD or active clamp, synchronous or Schottky secondary rectification, output filtering, optically isolated feedback, and thermal derating. Validation would include no-load, full-load, short-circuit, input-transient, startup, conducted-emissions, temperature-rise, and isolation tests.

Outcome. The isolated switching architecture meets the power and isolation requirements with lower thermal loss, but introduces design complexity, high-frequency EMI, transformer safety constraints, and control-loop issues.

Learning points. Architecture must be selected from system requirements, not efficiency alone. Isolation, fault behavior, thermal performance, noise, cost, manufacturability, and certification are coupled design variables.

10. Contemporary Developments and Best Practices

Modern power supplies increasingly use digital control, synchronous rectification, wide-bandgap switches, planar magnetics, integrated power modules, and telemetry for predictive maintenance. Digital control can support adaptive dead time, programmable current limits, fault logging, and system-level energy management, but introduces sampling delay, quantization, software validation, and cybersecurity concerns in networked equipment. Artificial intelligence is relevant mainly to condition monitoring, lifetime prediction, and adaptive optimization; it does not remove the need for physically valid protection limits.

Best practices include:

• begin with a complete electrical, thermal, mechanical, safety, and regulatory specification;
• calculate worst-case rather than nominal stresses;
• minimize high-current loop area and use deliberate PCB return paths;
• separate noisy power grounds from sensitive signal paths while controlling their reference relationship;
• place decoupling components close to switching devices;
• provide test points for current, gate voltage, switching-node voltage, and temperatures;
• validate with differential probes and correctly rated current probes;
• derate capacitors, semiconductors, magnetics, and connectors;
• test fault conditions before optimization for efficiency;
• document assumptions, tolerances, revision history, and compliance evidence.

Challenges and Limitations

No converter is ideal. Rectifiers produce harmonics and ripple; linear regulators waste power at large voltage differences; switching converters generate EMI and require control-loop expertise; transformers and inductors introduce parasitic capacitance and leakage inductance; electrolytic capacitors age with temperature and ripple current; and high efficiency may conflict with low noise, low cost, or small volume. Simulation is valuable but cannot fully replace measurement because layout parasitics, magnetic behavior, thermal interfaces, and component tolerances strongly influence real performance.

Chapter Summary

Power supplies convert available electrical energy into safe, stable, and useful forms. Rectifiers convert AC to pulsating DC, filters reduce ripple, and regulators maintain output voltage despite input and load variation. Full-wave rectification provides better utilization and twice-line-frequency ripple compared with half-wave operation. Linear regulators are simple and low-noise but dissipate approximately (V_in − V_out)I_out as heat. Switched-mode converters transfer energy through controlled switching and reactive components, achieving high efficiency and compact size at the cost of EMI, control, and layout complexity. Ideal buck, boost, and buck–boost relationships depend on continuous conduction and ideal components. Sound design also requires thermal analysis, feedback stability, protection, isolation, power-factor consideration, standards compliance, and worst-case validation.

Key Terms

Rectifier: Circuit that converts AC into unidirectional or pulsating DC.

Ripple: Residual periodic AC component on a DC output.

Bridge rectifier: Full-wave rectifier using four diodes.

Load regulation: Output-voltage change caused by load-current variation.

Line regulation: Output-voltage change caused by input-voltage variation.

Linear regulator: Regulator that controls voltage through a dissipative series or shunt device.

LDO: Low-dropout linear regulator requiring a small input-output voltage difference.

SMPS: Switched-mode power supply using high-frequency switching.

PWM: Pulse-width modulation used to control average converter energy.

Duty ratio: Fraction of a switching period during which a switch is on.

Buck converter: Step-down DC–DC converter.

Boost converter: Step-up DC–DC converter.

Buck–boost converter: Converter capable of step-up or step-down operation, commonly with polarity inversion.

Galvanic isolation: Separation without a conductive DC path.

Power factor: Ratio of real power to apparent power.

PFC: Power-factor correction.

Dropout voltage: Minimum input-output difference required for regulation.

Thermal resistance: Temperature rise per unit dissipated power.

Creepage: Surface distance between conductive parts.

Clearance: Air distance between conductive parts.

Review Questions

Short Answer Questions

1. Why does a full-wave rectifier have twice the ripple frequency of a half-wave rectifier?
2. Define line regulation and load regulation.
3. State two limitations of a linear regulator.
4. What is meant by galvanic isolation?
5. Why are current and voltage derating important in power-semiconductor selection?

Long Answer Questions

1. Explain the stages of an isolated switched-mode power supply and the function of each stage.
2. Derive the ideal continuous-conduction buck-converter relationship using inductor volt-second balance.
3. Compare linear and switched-mode regulators in efficiency, noise, thermal design, complexity, and applications.
4. Explain capacitor-input filtering and derive the approximate ripple-voltage relation.
5. Discuss safety, EMI, thermal, and regulatory considerations in mains-connected power supplies.

Analytical / Application Questions

1. A 15 V RMS, 50 Hz transformer feeds a bridge rectifier with 0.7 V drop per conducting diode. A 0.5 A load permits 2 V peak-to-peak ripple. Estimate the minimum smoothing capacitance and capacitor peak voltage.
2. A 48 V input must supply 12 V at 2 A. Compare the ideal loss of a linear regulator with an 88% efficient buck converter. Discuss the thermal implications.
3. Select and justify an isolated or non-isolated architecture for a battery-powered medical measurement interface, considering patient safety, efficiency, noise, battery range, and certification.

Multiple Choice Questions with Answers

Question 1

For an ideal full-wave rectifier with input peak voltage V_m, the average output voltage is:

A. V_m/π

B. 2V_m/π

C. V_m/2

D. πV_m/2

Correct Answer: B. 2V_m/π

Question 2

A full-wave rectifier supplied from a 50 Hz source has an ideal ripple frequency of:

A. 25 Hz

B. 50 Hz

C. 100 Hz

D. 150 Hz

Correct Answer: C. 100 Hz

Question 3

For a capacitor-input filter, increasing the load current while keeping capacitance and ripple frequency constant will:

A. Decrease ripple voltage

B. Increase ripple voltage

C. Eliminate ripple

D. Reverse output polarity

Correct Answer: B. Increase ripple voltage

Question 4

A linear regulator receives 20 V and supplies 5 V at 0.4 A. Neglecting quiescent current, its approximate dissipation is:

A. 2 W

B. 4 W

C. 6 W

D. 8 W

Correct Answer: C. 6 W

Question 5

The ideal continuous-conduction relationship for a buck converter is:

A. V_out = V_in/(1−D)

B. V_out = DV_in

C. V_out = −DV_in/(1−D)

D. V_out = V_in/D

Correct Answer: B. V_out = DV_in

Question 6

A 24 V ideal buck converter operates with a duty ratio of 0.25. Its output voltage is:

A. 6 V

B. 12 V

C. 18 V

D. 96 V

Correct Answer: A. 6 V

Question 7

Which device is generally characterized by voltage-controlled operation and fast switching at low-to-medium voltage?

A. Power MOSFET

B. Relay coil

C. Fuse

D. Transformer core

Correct Answer: A. Power MOSFET

Question 8

Galvanic isolation means that:

A. Input and output have identical voltages

B. There is no conductive DC path between input and output

C. The circuit has no magnetic components

D. The output is always unregulated

Correct Answer: B. There is no conductive DC path between input and output

Question 9

If a converter delivers 10 W at 80% efficiency, its input power is:

A. 8 W

B. 10 W

C. 12.5 W

D. 18 W

Correct Answer: C. 12.5 W

Question 10

For a 100 kHz switching frequency, the switching period is:

A. 1 μs

B. 10 μs

C. 100 μs

D. 1 ms

Correct Answer: B. 10 μs