Power Electronics and Electric Drives

Introduction

Power electronics converts and controls electrical energy through semiconductor switching. Diodes, thyristors, MOSFETs, insulated-gate bipolar transistors (IGBTs), and wide-bandgap devices enable efficient rectifiers, choppers, inverters, chargers, renewable-energy interfaces, and motor drives. Unlike linear power control, switched conversion ideally dissipates little power because a device is either blocking voltage or conducting current with a small voltage drop. Real devices, however, have conduction losses, switching losses, thermal limits, parasitic effects, and electromagnetic compatibility constraints.

Laboratory work in this area is potentially hazardous. Converter circuits can contain lethal voltages, high fault currents, charged capacitors, rotating machinery, and unexpectedly energized motor terminals. Correct measurement technique, protective coordination, isolated gate driving, and disciplined test planning are therefore integral engineering competencies rather than administrative precautions.

Learning Objectives

After completing this chapter, students should be able to:

• distinguish controlled and uncontrolled rectifiers;
• analyze basic buck, boost, and bidirectional choppers;
• evaluate single-phase and three-phase inverter operation;
• apply pulse-width modulation principles;
• calculate conduction, switching, and thermal losses under stated assumptions;
• measure efficiency, harmonics, and total harmonic distortion correctly;
• explain scalar and vector motor-control methods;
• evaluate gate-drive isolation, protection, electromagnetic interference, and laboratory safety; and
• plan and interpret practical converter and electric-drive tests.

1. Power Conversion Structure

A practical drive contains more than a power converter. It includes command processing, control algorithms, isolated gate drivers, current and voltage measurement, protection, thermal management, and the electromechanical load.

Figure: Command, power, sensing-feedback, and protection paths in an electric drive.

Protection must not be represented as applying an uncontrolled command directly to the motor. Sensors observe converter or plant variables; protection logic then inhibits switching or commands the controller to enter a safe state.

2. Power Semiconductor Devices

2.1 Device selection

Device ratings must include voltage overshoot, current transients, junction temperature, short-circuit withstand time, safe operating area, and repetitive switching stress. A device rated at the nominal bus voltage is not adequately selected unless transient margins have also been established.

2.2 Switching states and commutation

In an ideal switch, the off-state current is zero and the on-state voltage is zero. Real transitions have simultaneous voltage and current, producing energy loss. Commutation is the transfer of current from one device or path to another. It may be line commutated, as in a thyristor rectifier, or forced by actively controlled switches.

A freewheeling diode provides a path for inductive current when the controlled switch turns off. Interrupting inductive current without a suitable path can create a large overvoltage because v_L = L\,di/dt.

3. Uncontrolled and Controlled Rectifiers

3.1 Single-phase diode bridge

An ideal full-wave diode bridge supplied by a sinusoidal source produces the absolute value of the input voltage. If the source is

then the average output voltage is

Here V_m is source peak voltage, V_s,rms is source RMS voltage, and the devices and source impedance are assumed ideal. This average applies to the rectified voltage waveform; load-current shape depends on resistance, inductance, back emf, and filtering.

A capacitor-input rectifier behaves differently. Its DC voltage approaches the source peak at light load, while short charging intervals create high crest-factor line current and substantial harmonics.

3.2 Thyristor bridge

For an ideal single-phase fully controlled bridge with continuous DC current and negligible source inductance,

where α is the firing delay measured from the natural commutation point. For 0° ≤ α < 90°, average power normally flows from AC to DC. At α = 90°, ideal average output voltage is zero. Operation with α > 90° can produce line-commutated inversion only when continuous current is maintained by a DC-side energy source and adequate commutation margin exists. The equation must not be applied indiscriminately to discontinuous current.

Source inductance causes commutation overlap and reduces average DC voltage. It also produces notches in the AC voltage and increases electromagnetic disturbance.

Worked Example 1: Controlled rectifier

A 230 V RMS single-phase source supplies an ideal fully controlled bridge with continuous current. For α = 45°,

The result assumes ideal devices, negligible overlap, and continuous load current.

3.3 Three-phase bridge

For a six-pulse three-phase diode bridge with continuous DC current, ideal devices, and negligible overlap,

where V_LL,rms is the line-to-line supply voltage. The dominant characteristic AC-side current harmonics are generally of orders h = 6k ± 1, although their exact magnitudes depend on the DC-side impedance and supply conditions.

4. DC–DC Choppers

4.1 Buck converter

An ideal buck converter reduces a DC voltage using a controlled switch and freewheeling path. Under continuous-conduction mode (CCM), periodic steady state, and negligible losses, the average inductor voltage is zero:

Therefore,

Here D = t_on/T_s is duty ratio, T_s = 1/f_s is switching period, f_s is switching frequency, V_s is input voltage, and V_o is average output voltage.

The peak-to-peak inductor ripple in CCM is

where L is inductance. The equality follows from V_o = DV_s and is valid for the ideal steady-state buck converter.

Worked Example 2: Buck chopper

A 100 V source feeds an ideal buck converter operating at 20 kHz with D = 0.40 and L = 2.0\,mH.

Since T_s = 50\,μs, the on-time is 20 μs. Thus,

This is peak-to-peak ripple, not RMS current.

4.2 Boost and bidirectional converters

For an ideal boost converter in CCM,

This relation assumes periodic steady state and lossless components. At large duty ratios, parasitic resistance and switching losses prevent unlimited voltage gain.

Four-quadrant choppers permit positive and negative motor voltage and current. They support motoring, reversal, and regenerative braking, but require dead-time control and careful prevention of DC-link shoot-through.

5. Inverters and Pulse-Width Modulation

An inverter converts DC to controlled AC. A voltage-source inverter uses a relatively stiff DC-link voltage, whereas a current-source inverter uses a controlled DC-link current. The two-level three-phase voltage-source inverter contains three complementary switching legs.

5.1 Square-wave and PWM operation

Square-wave switching provides high fundamental utilization but produces low-order harmonics. Pulse-width modulation shifts significant harmonic energy toward the switching frequency, making filtering easier and enabling amplitude control.

In sinusoidal PWM (SPWM), a sinusoidal reference is compared with a triangular carrier. The amplitude modulation ratio is

For linear SPWM, normally 0 ≤ m_a ≤ 1. For a two-level three-phase inverter, the fundamental phase-to-neutral voltage peak is

and the fundamental line-to-line RMS voltage is

These expressions apply to balanced linear SPWM. Overmodulation and space-vector modulation have different voltage-utilization behavior.

The frequency modulation ratio is m_f = f_c/f₁, where f_c is carrier frequency and f₁ is reference fundamental frequency. Higher switching frequency can improve waveform quality but increases switching and gate-drive losses.

5.2 Dead time

Complementary switches must never conduct simultaneously. A blanking interval called dead time is inserted between turn-off of one device and turn-on of the other. Excessive dead time distorts phase voltage, particularly near current zero, and can produce torque ripple. Insufficient dead time risks destructive shoot-through.

5.3 Space-vector modulation

Space-vector modulation treats inverter switching states as voltage vectors. It selects adjacent active vectors and zero vectors during each sampling interval. Compared with basic sinusoidal PWM, it can improve DC-bus utilization and provides systematic control of switching sequences.

6. Harmonics and Power Quality

A periodic current may contain a DC component, a fundamental AC component, and higher harmonics. Let I₀ denote the DC component, I₁ the RMS fundamental component, and I_h the RMS value of harmonic order h. The total RMS current is

Current total harmonic distortion is therefore

If an instrument instead reports AC-only RMS, defined by I_rms,ac² = I_rms,total²-I₀², then

The DC term must not be omitted when total RMS includes DC. Practical analyzers also have finite bandwidth, so a reported THD value is meaningful only with a stated frequency range and acquisition method.

For distorted voltage and current, apparent power remains S = V_rmsI_rms, but power factor is

where P is average real power. Displacement factor alone does not account for distortion.

7. Losses, Efficiency, and Thermal Constraints

7.1 Conduction loss

For a MOSFET represented by an on-state resistance,

Here I_D,rms,cycle is the RMS current through that specific device evaluated over the complete switching or fundamental analysis interval, including intervals when its current is zero. Thus the duty factor is already embedded in the RMS value. If only the RMS current during the conducting interval is known, the complete-cycle RMS must first be found; for constant duty D and otherwise zero current, I_D,rms,cycle = I_on,rms√D. The resistance must be evaluated at the expected junction temperature.

An approximate IGBT conduction model is

where the current averages and RMS values are taken for the individual device over the complete cycle, V_CE0 is the threshold-like voltage in the linearized model, and r_CE is differential on-state resistance.

7.2 Switching loss

If data-sheet or measured turn-on and turn-off energies correspond to the actual voltage, current, temperature, gate resistance, and commutation conditions, average switching loss is approximately

Diode reverse-recovery and output-capacitance losses may require separate inclusion. Linear scaling of tabulated switching energy is only an approximation and must be validated for design work.

Worked Example 3: Semiconductor loss and temperature

A MOSFET has hot R_DS(on) = 45\,mΩ. Its device-current RMS over the complete switching period is 8.0 A. At the operating condition, E_on+E_off = 120\,μJ, f_s = 20\,kHz, and other losses are 0.15 W.

If ambient temperature is 45 °C and effective steady-state junction-to-ambient thermal resistance is 8 °C/W,

This lumped calculation assumes steady state and a valid effective thermal resistance. Transient operation requires a thermal-impedance network or manufacturer transient curves.

7.3 Efficiency

Converter efficiency is

Input and output powers must be measured simultaneously under thermal steady state where possible. For an AC motor, shaft output power is

where T is torque in N·m, ω_m is mechanical angular speed in rad/s, and n is speed in r/min.

8. Electric Drives

8.1 DC motor drives

For a separately excited DC motor with approximately constant field flux,

Armature-voltage control regulates speed below base speed. Field weakening can extend speed above base speed but reduces available torque and requires mechanical-speed limits.

8.2 Induction-motor scalar control

The air-gap flux of an induction motor is approximately proportional to voltage divided by electrical frequency. Constant volts-per-hertz control therefore maintains approximately constant flux below base speed:

At low frequency, stator-resistance voltage drop becomes significant, so a voltage boost may be required. Above base frequency, the inverter voltage limit causes field weakening and declining torque capability.

8.3 Vector control

Field-oriented control transforms measured phase currents into rotating d–q components. In a properly oriented reference frame, one component controls flux and the other controls torque, producing DC-motor-like decoupling. The control chain commonly includes Clarke and Park transformations, current controllers, an inverse transformation, and PWM.

For a sinusoidal permanent-magnet synchronous motor,

where p is the number of pole pairs, λ_m is permanent-magnet flux linkage, and L_d, L_q, i_d, and i_q are rotating-frame inductances and currents. For a surface-mounted machine with L_d ≈ L_q, reluctance torque is small and torque is approximately proportional to i_q.

8.4 Regenerative braking

During regeneration, mechanical energy is converted into electrical energy. The DC link must transfer this energy to the supply, an energy-storage system, or a braking resistor. A diode front end cannot normally return energy to the AC supply; without another energy path, the DC-link voltage may rise dangerously.

9. Gate Drives, Isolation, Protection, and EMC

9.1 Gate-drive isolation

High-side devices require gate voltages referenced to moving switch nodes. Isolation can be provided by pulse transformers, optocouplers, capacitive isolators, or magnetic digital isolators. Selection criteria include:

• isolation working voltage and transient rating;
• common-mode transient immunity;
• propagation delay and channel matching;
• driver peak source and sink current;
• undervoltage lockout;
• Miller-clamp or negative-bias capability; and
• creepage and clearance requirements.

Isolation is not automatically safe merely because an isolator is present. PCB contamination, inadequate spacing, isolated-supply capacitance, and probe connections can defeat the intended barrier.

9.2 Protection functions

Common protections include desaturation or overcurrent detection, DC-link overvoltage, undervoltage lockout, overtemperature shutdown, motor overload, loss of sensor signal, ground-fault detection, and overspeed protection. A hardware overcurrent path is often required because software may respond too slowly.

Fault handling should define whether switching is immediately inhibited, ramped down, latched off, or automatically retried. Automatic restart can be dangerous around rotating machinery and should be controlled by a documented safety policy.

9.3 Electromagnetic compatibility

Fast voltage and current transitions excite parasitic capacitances and inductances. Differential-mode noise flows around the intended power loop; common-mode noise flows through parasitic capacitance to chassis, earth, motor frame, and cables.

Mitigation measures include minimizing commutation-loop area, using short gate loops, separating power and signal returns, applying snubbers, controlling gate resistance, using common-mode chokes, shielding motor cables, and terminating shields correctly. Lowering dv/dt or di/dt can reduce emissions but may increase switching loss.

Compliance work may involve IEC, IEEE, CISPR, UL, or national requirements depending on equipment category and jurisdiction. Engineers must use current editions and recognize that laboratory prototypes are not automatically compliant products.

10. Laboratory Methods

10.1 Pre-energization procedure

1. Inspect topology, ratings, polarity, grounding, and mechanical guards.
2. Calculate expected voltage, current, speed, and loss limits.
3. Set current-limited supplies and conservative controller limits.
4. Verify dead time and gate pulses using a low-energy supply.
5. Confirm isolated or appropriately rated voltage probes.
6. Check emergency stop, discharge resistors, and interlocks.
7. Begin at reduced DC-link voltage where the apparatus permits.

A conventional earth-referenced oscilloscope ground clip must never be connected casually to a switching node. Doing so can short that node to protective earth. Use a correctly rated differential probe or an approved isolated measurement architecture.

10.2 Rectifier and chopper tests

For a rectifier, record input voltage and current, output average voltage, ripple, firing angle, and power factor. Compare measurements with ideal equations while accounting for device drops, overlap, and discontinuous current.

For a chopper, vary duty ratio at constant input voltage. Record average output voltage, inductor ripple, device temperature, and efficiency. Capture switching-node and gate waveforms to assess overshoot, ringing, and dead time.

10.3 Inverter and drive tests

For an inverter-fed motor:

• verify phase sequence at low voltage;
• record DC-link voltage and current;
• measure line voltage, phase current, speed, torque, and temperature;
• obtain harmonic spectra using an adequate sample rate and anti-aliasing practice;
• evaluate acceleration, steady operation, reversal, and braking only within safe limits; and
• compare electrical input power with shaft output power.

Uncertainty should be reported. Current-probe offset, torque-sensor zero drift, bandwidth limitation, phase error, and non-simultaneous readings can materially distort calculated efficiency.

11. Illustrative Application Case: Pump-Drive Retrofit

Background

An illustrative water-treatment facility replaces a throttled induction-motor pump with an 11 kW variable-frequency drive. The three-phase supply is 400 V RMS line-to-line. The drive uses a six-pulse diode rectifier, DC-link capacitor, and two-level IGBT inverter.

Problem

Initial tests show excessive motor-cable common-mode current, visible switching overshoot, and a heat-sink temperature above the design target. The engineering team must verify voltage capability, efficiency, thermal margin, and EMC measures without compromising motor insulation.

Analysis

The ideal average rectifier output under continuous-current assumptions is

Suppose the required inverter fundamental line voltage is 230 V RMS. Under linear SPWM,

Thus the requested voltage lies within the linear modulation range. Measurements at the operating point give 9.80 kW electrical input and 9.20 kW shaft output:

The 0.60 kW difference includes rectifier, inverter, motor, cable, and mechanical losses; it must not all be assigned to semiconductor loss.

Oscilloscope measurements using a correctly rated differential probe reveal that long motor leads and rapid IGBT transitions cause reflected-wave overvoltage. A current probe around all three phase conductors indicates common-mode current returning through the cable shield and stray capacitance.

Approach and Solution

The team shortens the inverter commutation loop, adjusts gate resistance to reduce switching speed moderately, adds a properly rated output dv/dt filter, and bonds the motor-cable shield over a low-impedance 360° termination. Switching loss and heat-sink temperature are remeasured after the gate-resistance change. Protection thresholds are verified through controlled low-energy tests rather than destructive fault application.

The drive controller is configured for quadratic pump loading and reduced speed during low-flow periods. Because centrifugal-pump power is approximately proportional to the cube of speed under similar operating conditions, moderate speed reduction can yield substantial energy savings, although the exact saving depends on the hydraulic system curve.

Outcome

Motor-terminal overshoot and common-mode current decrease, measured temperatures remain within the derated design limit, and system efficiency is documented with uncertainty estimates. The installation passes the applicable site EMC assessment.

Learning Points

• Ideal voltage equations establish feasibility but do not establish compliance or reliability.
• EMC adjustments interact with switching loss and thermal design.
• System efficiency cannot be inferred from semiconductor loss alone.
• Feedback and protection must be validated under controlled, safe conditions.
• Variable-speed control can reduce both electrical consumption and mechanical throttling losses.

12. Contemporary Developments and Engineering Responsibility

SiC and GaN devices enable higher switching frequencies and power densities, but faster edges intensify layout, insulation, measurement, and EMC challenges. Integrated power modules increasingly combine switches, drivers, sensing, and protection.

Model-predictive control and data-driven condition monitoring are being introduced into drives. Machine-learning models can identify bearing, thermal, or switching anomalies, but their outputs must not replace independently validated protection. Training data, false alarms, cybersecurity, and explainability are relevant engineering-governance concerns.

Responsible design includes energy efficiency, repairability, material use, safe end-of-life handling, and accurate reporting of performance. Engineers must not disable safety interlocks to obtain favorable test results or omit adverse thermal and harmonic measurements.

Chapter Summary

Power converters control energy through semiconductor switching. Diode and thyristor rectifiers convert AC to DC, choppers regulate DC voltage, and PWM inverters synthesize variable-frequency AC. Their ideal equations are useful only when assumptions such as continuous current, negligible losses, and linear modulation are satisfied.

Converter assessment requires waveform measurement, harmonic analysis, efficiency testing, semiconductor-loss estimation, and thermal verification. THD calculations must distinguish total RMS, DC, fundamental, and harmonic components. MOSFET conduction loss requires the individual device-current RMS over the complete analysis cycle; switching loss depends on operating-condition-specific transition energy.

Electric drives combine converters with motor models, feedback, control, isolated gate driving, and protection. Scalar control provides economical induction-motor regulation, while field-oriented control offers dynamic torque and flux control. Safe laboratory practice, correct sensing direction, hardware protection, EMC-aware layout, and suitable measurement isolation are essential to reliable professional work.

Key Terms

Commutation: Transfer of current from one semiconductor path to another.

Controlled rectifier: AC–DC converter whose average output is regulated through controllable switches.

Dead time: Delay inserted between complementary switch commands to prevent shoot-through.

Duty ratio: Fraction of a switching period for which a controlled switch is commanded on.

Field-oriented control: Rotating-frame motor control that approximately decouples flux and torque.

Gate driver: Circuit that supplies the voltage and current required to switch a power device.

Harmonic: Sinusoidal waveform component at an integer multiple of the fundamental frequency.

Inverter: Converter that produces controlled AC from a DC source.

Modulation index: Ratio governing the commanded fundamental voltage in a PWM inverter.

Regenerative braking: Conversion of mechanical energy into recoverable electrical energy.

Reverse recovery: Transient reverse current that occurs when a conducting diode turns off.

Shoot-through: Simultaneous conduction of devices across a DC bus.

Snubber: Circuit used to limit switching transient voltage, current, or oscillation.

Space-vector modulation: PWM method based on selecting inverter voltage vectors.

Thermal resistance: Steady-state temperature rise per unit dissipated power, expressed in °C/W or K/W.

Total harmonic distortion: RMS magnitude of higher harmonics relative to the RMS fundamental.

Voltage-source inverter: Inverter supplied by a comparatively stiff DC-link voltage.

Wide-bandgap semiconductor: Device material, such as SiC or GaN, supporting fast and high-temperature switching.

Review Questions

Short Answer Questions

1. Distinguish an uncontrolled rectifier from a controlled rectifier.
2. Why is dead time required in a voltage-source inverter?
3. What assumptions are required for the ideal buck relation V_o = DV_s?
4. Why can increasing switching frequency increase converter temperature?
5. What is the purpose of galvanic isolation in a gate-drive system?

Long Answer Questions

1. Explain the operation of a single-phase fully controlled bridge, including rectification and inversion conditions.
2. Derive the ideal CCM voltage-conversion ratio and inductor-ripple expression for a buck converter.
3. Compare sinusoidal PWM, square-wave operation, and space-vector modulation.
4. Discuss conduction loss, switching loss, thermal resistance, and transient thermal impedance in converter design.
5. Explain constant-volts-per-hertz and field-oriented control of AC motors, including their limitations.

Analytical / Application Questions

1. A 300 V DC bus supplies a three-phase inverter using linear SPWM with m_a = 0.80. Calculate the fundamental line-to-line RMS voltage and identify the assumptions used.
2. A converter current has total RMS value 15 A, DC component 9 A, and fundamental RMS component 10 A. Calculate current THD and explain why the DC component must be removed.
3. Design a laboratory test plan for measuring inverter efficiency, harmonics, device temperature, and motor shaft power while addressing isolation, uncertainty, and emergency shutdown.

Multiple Choice Questions with Answers

Question 1

An ideal single-phase diode bridge is supplied from 230 V RMS. What is its average rectified output voltage?

A. 146 V

B. 207 V

C. 230 V

D. 325 V

Correct Answer: B. 207 V

Question 2

For an ideal fully controlled single-phase bridge with continuous current, line-commutated inversion may occur when which condition is satisfied?

A. α < 45° with a purely resistive load

B. α = 0° with discontinuous current

C. α > 90° with sustained DC current, a DC-side energy source, and adequate commutation margin

D. α = 90° without a DC-side energy source

Correct Answer: C. α > 90° with sustained DC current, a DC-side energy source, and adequate commutation margin

Question 3

An ideal buck converter has V_s = 100\,V, D = 0.4, f_s = 20\,kHz, and L = 2\,mH. What is the CCM peak-to-peak inductor-current ripple?

A. 0.24 A

B. 0.40 A

C. 0.60 A

D. 1.20 A

Correct Answer: C. 0.60 A

Question 4

A three-phase two-level inverter uses linear SPWM with V_dc=400\,V and m_a=0.80. Approximately what is its fundamental line-to-line RMS voltage?

A. 113 V

B. 196 V

C. 245 V

D. 320 V

Correct Answer: B. 196 V

Question 5

A MOSFET has R_DS(on)=0.05\,Ω. Its device-current RMS, evaluated over the complete switching period including zero-current intervals, is 6 A. What is its conduction loss?

A. 0.30 W

B. 1.20 W

C. 1.80 W

D. 3.60 W

Correct Answer: C. 1.80 W

Question 6

A periodic current has total RMS value 10 A, DC component 6 A, and fundamental RMS component 7 A. What is its THD?

A. 38.7%

B. 55.3%

C. 71.4%

D. 87.1%

Correct Answer: B. 55.3%

Question 7

A switch has E_on+E_off=150\,μJ at its operating condition and switches at 20 kHz. What is its approximate switching loss, excluding other losses?

A. 0.30 W

B. 1.33 W

C. 3.00 W

D. 7.50 W

Correct Answer: C. 3.00 W

Question 8

Which statement correctly describes sensing and protection in a motor drive?

A. Protection should apply uncontrolled electrical power directly to the motor.

B. Sensor feedback should originate at the converter or plant, while protection should inhibit the gate drivers or command the controller.

C. Motor feedback should flow only from the controller toward the motor.

D. Gate-driver isolation eliminates the need for overcurrent protection.

Correct Answer: B. Sensor feedback should originate at the converter or plant, while protection should inhibit the gate drivers or command the controller.

Question 9

A device dissipates 4 W at an ambient temperature of 40 °C. If its effective steady-state junction-to-ambient thermal resistance is 10 °C/W, what is the estimated junction temperature?

A. 44 °C

B. 50 °C

C. 80 °C

D. 400 °C

Correct Answer: C. 80 °C

Question 10

Why is approximately constant V/f used below base speed in an induction-motor drive?

A. To maintain approximately constant air-gap flux

B. To eliminate all rotor losses

C. To hold switching frequency equal to rotor speed

D. To make motor torque independent of current

Correct Answer: A. To maintain approximately constant air-gap flux