Electric Drives

What is an Electric Drive?

Electric drives dominate motion control because they are highly efficient (85–98%), span an enormous power range from milliwatts to megawatts, allow four-quadrant operation, are clean and quiet, are easy to control electronically, and can regenerate energy back to the source.

Historical Evolution of Electric Drives

Drive technology evolved in distinct eras: group (line-shaft) drives and early DC motors (1880s–1920); Ward–Leonard systems and mercury-arc rectifiers (1930s–1960); the thyristor era with current- and voltage-source inverters and analog control (1960s–80s); IGBT/MOSFET power stages with field-oriented control (FOC), direct torque control (DTC) and DSP control (1980s–2000); and today's wide-bandgap (SiC, GaN) devices, sensorless and model-predictive control, digital twins and AI.

Components of an Electric Drive

Every drive is built from five essential subsystems plus the mechanical transmission:

1. Power source — AC mains, a DC bus, battery, fuel cell or photovoltaic array.
2. Power modulator (converter) — adjusts voltage, current, frequency and waveform; rectifiers, choppers, inverters, cycloconverters and AC voltage controllers.
3. Electrical motor — DC, induction (IM), synchronous (SM), PMSM, BLDC, switched-reluctance (SRM) or stepper.
4. Sensing unit — voltage, current, speed (encoder/resolver), position and temperature.
5. Control unit — analog or digital controller (DSP, FPGA, microcontroller) implementing speed, torque and position loops.

Classification of Electric Drives

Drives are classified along four independent axes, and by physical configuration:

• By source: DC drives, AC drives, hybrid.
• By motor: induction, synchronous, DC/BLDC.
• By control: scalar (V/f), vector/FOC, DTC/MPC.
• By application: servo, traction, industrial.

Selection of an Electric Drive — Engineering Trade-offs

Drive selection balances eight criteria: steady-state requirements (rated power, torque, speed, duty cycle S1–S8); transient demands (starting/braking torque, acceleration); speed range and regulation; number of quadrants required (1Q, 2Q or 4Q); environment (ambient temperature, IP rating, hazardous area); power-quality and harmonic limits (IEEE 519, IEC 61000); capital plus life-cycle cost with energy-saving payback; and reliability and maintainability (MTBF, brushed versus brushless).

Fundamental Torque Equation

The dynamics of every drive follow Newton's law for rotation:

Here \(T_m\) is the motor developed torque (N·m), \(T_L\) is the load torque referred to the shaft (N·m), \(J\) is the polar moment of inertia (kg·m\(^2\)) and \(\omega_m\) is the angular velocity (rad/s). Three operating states follow directly: \(T_m \gt T_L\) gives acceleration, \(T_m \lt T_L\) gives deceleration, and \(T_m = T_L\) is steady state.

When the inertia varies (a robotic arm, reeling drum or winch) an extra term appears:

The associated power balance is \(P_m = T_m\omega_m\), \(P_L = T_L\omega_m\), and the accelerating power

Referring Load Parameters to the Motor Shaft

A real drive train contains gears and translating masses. Defining the gear ratio \(a_i = \omega_i/\omega_m\), all inertias and torques are referred to the motor shaft so the single equation of motion still applies.

These follow from energy conservation, \( \tfrac{1}{2}J_{eq}\omega_m^2 = \tfrac{1}{2}J_0\omega_m^2 + \sum_i \tfrac{1}{2}J_i\omega_i^2 + \tfrac{1}{2}M v^2\), with efficiencies \(\eta_i\) accounting for transmission losses.

Load Torque Characteristics

The total load torque combines several components:

where \(T_{Lf}\) is friction \((C + B\omega_m)\), \(T_{Lw}\) is windage \((C_w\omega_m^2)\), \(T_{Ls}\) is the useful shaft work and \(T_{Lg}\) is a gravity term (hoists). Four canonical speed–torque shapes recur: constant torque (hoist), torque linear in speed (viscous friction), torque proportional to \(\omega^2\) (fans and pumps) and constant-power loads where \(T_L \propto 1/\omega\). Loads are also classified as active (hoist, crane — torque retains its sign and can drive the motor) or passive (drill, mill, press — torque always opposes motion).

Steady-State Stability

Applying a small perturbation about an operating point and linearising the equation of motion gives

whose solution \(\Delta\omega_m = (\Delta\omega_m)_0\,e^{t/\tau}\) has time constant

Four-Quadrant Operation

With \(P = T\omega_m\), the sign of power distinguishes motoring \((P\gt0)\) from regeneration \((P\lt0)\). Plotting torque against speed yields four quadrants: Q-I forward motoring, Q-II forward braking, Q-III reverse motoring and Q-IV reverse braking.

Braking Methods — Comparison

Three electrical braking methods are used in drives. Regenerative braking occurs when \(E \gt V_s\) (or \(\omega \gt \omega_s\)): energy flows back to the source with \(P_{reg} = E I_a\), making it the most efficient method. Dynamic (rheostatic) braking disconnects the supply and connects a braking resistor, dissipating kinetic energy as heat with \(P_{diss} = E^2/(R_a + R_{br})\). Plugging reverses the supply voltage, producing a very large current \(I_a = (V+E)/(R_a + R_{br})\) for fast but inefficient stopping.

DC Motor — Fundamental Equations

The DC machine is governed by an EMF and a torque relation that share the same machine constant \(K\phi\):

Applying Kirchhoff's voltage law to the armature gives the circuit and speed equations:

There are three control variables — armature voltage \(V_a\), field flux \(\phi\) and series resistance \(R_a\) — and each defines a distinct speed-control region. The mechanical time constant is \( \tau_m = JR_a/(K\phi)^2 \).

DC Machine — Construction and Cross-Section

The key parts are the yoke/frame (mechanical support and flux return path), the field poles carrying shunt or series windings that establish \(\phi\), the laminated armature whose slots hold the lap or wave winding, the commutator and brushes (a mechanical inverter that rectifies the internal AC EMF to DC at the terminals), and interpoles and compensating windings that cancel armature reaction and reduce sparking.

where \(P\) is the number of poles, \(Z\) the total conductors and \(A\) the number of parallel paths (\(A=2\) for wave winding, \(A=P\) for lap winding).

Types of DC Motors — Characteristics

The four excitation arrangements give markedly different speed–torque characteristics, which determine where each is used.

Speed Control — Three Regions

A separately excited DC motor offers two control regions. Below base speed, the flux is held at its rated value and the armature voltage is varied, giving a constant-torque region:

Above base speed, the voltage is capped and the field is weakened, so \( \omega_m \propto 1/\phi \). Power remains constant while \(T_{max} \propto \phi \propto 1/\omega_m\) — the constant-power region.

Starting of a DC Motor

Three solutions are used: a stepped rheostatic starter (traditional), a reduced-voltage start through a converter (modern), or the Ward–Leonard system (legacy, smooth). For an \(n\)-step starter the current swings between a maximum \(I_1\) and minimum \(I_2\):

Power Semiconductor Devices — the Drive Toolbox

The converter that feeds the motor is built from power semiconductors. The choice of device fixes the achievable voltage, current and switching frequency.

Performance Indices for Phase-Controlled Converters

Phase-controlled converters are characterised by DC-side and AC-side indices. On the DC side,

On the AC side, the efficiency \(\eta = P_{dc}/P_{ac}\), transformer utilisation factor \(\text{TUF} = P_{dc}/(V_sI_s)_{rated}\), displacement power factor \(DPF = \cos\phi_1\), harmonic factor \(HF = \sqrt{I_s^2-I_{s1}^2}/I_{s1}\) and overall power factor \(PF = (I_{s1}/I_s)\cos\phi_1\).

These matter directly for DC drives: a low ripple factor gives smooth armature current and low torque ripple; a high TUF means better transformer utilisation; a high power factor reduces line current and \(I^2R\) loss; and a low harmonic factor secures compliance with IEEE 519 and IEC 61000-3-2.

Single-Phase Phase-Controlled Converters

For the full bridge with inductive load, the key design relations are \(V_a = \tfrac{2V_m}{\pi}\cos\alpha\), \(I_a = (V_a - E)/R_a\), \(V_{a,rms} = V_m/\sqrt{2}\), \(FF = \pi/(2\sqrt{2}\cos\alpha)\), fundamental source current \(I_{s1} = 4I_a/(\pi\sqrt 2)\) and \(PF \approx 0.9\cos\alpha\). With \(\alpha \lt 90^\circ\) the converter rectifies; with \(\alpha \gt 90^\circ\) it inverts and returns energy (regeneration).

Single-Phase Full Converter with R-L-E (DC Motor) Load

Solving \(L_a\dot i + R_a i = v_o(t) - E\) gives the steady-state armature current in continuous conduction:

with \(Z = \sqrt{R_a^2 + (\omega L_a)^2}\), \(\phi = \tan^{-1}(\omega L_a/R_a)\) and \(\tau_a = L_a/R_a\). The average current is \(I_a^{avg} = (V_a - E)/R_a\). The critical inductance for the CCM boundary is approximately \(L_{a,crit} \approx V_m\sin\alpha/(\omega I_{a,min})\); below this the drive enters DCM and \(V_a\) rises.

Single-Phase Converter-Fed DC Drives — Quadrant Map

For a single full converter, \(V_a = \tfrac{2V_m}{\pi}\cos\alpha\): with \(0 \le \alpha \lt 90^\circ\) the drive rectifies in Q-I, and with \(90^\circ \lt \alpha \le 180^\circ\) it inverts in Q-IV (regeneration). A semi-converter, \(V_a = \tfrac{V_m}{\pi}(1+\cos\alpha)\), is single-quadrant (no inversion). A dual converter (two anti-parallel full bridges) achieves four-quadrant operation: in non-circulating mode \(\alpha_1 + \alpha_2 = 180^\circ\) with one bridge active; in circulating mode both are active and a reactor limits the circulating current \(i_{circ} = \tfrac{1}{\omega L_r}\!\int (v_{o1}-v_{o2})\,dt\).

Three-Phase Phase-Controlled Converters

For the three-phase full bridge with inductive load, \(V_a = \tfrac{3\sqrt 2 V_{LL}}{\pi}\cos\alpha\), \(I_{s,rms} = \sqrt{2/3}\,I_a\), \(I_{s1} = \tfrac{\sqrt 6}{\pi}I_a\), \(HF = 0.311\), \(DPF = \cos\alpha\), \(PF = \tfrac{3}{\pi}\cos\alpha\) and the ripple frequency is \(6f\). Source inductance modifies the output to \(V_a = \tfrac{3\sqrt 2 V_{LL}}{\pi}\cos\alpha - \tfrac{3\omega L_s I_a}{\pi}\).

Three-Phase Full Converter-Fed DC Drive — Operating Modes

The speed–torque relation in continuous conduction is a family of parallel lines, one per firing angle:

At light load the converter enters DCM where \(V_a = \tfrac{3V_{mL}}{\pi}[\cos\alpha - \cos(\alpha+\beta)]\) with \(\beta \lt 60^\circ\), and the output becomes load-dependent. Four-quadrant reversing drives (rolling mills) use a dual converter with \(\alpha_1 + \alpha_2 = 180^\circ\) so that \(V_{a1} = -V_{a2}\), a reactor limiting the circulating current.

Discontinuous conduction, source inductance and filtering all degrade performance. A critical inductance keeps the three-phase full converter in CCM:

Phase control gives a poor power factor at low speed (\(PF = 0.955\cos\alpha\)). It is improved by asymmetrical or sequence control of multi-stage SCR banks, by a PWM active front-end (giving unity power factor and sinusoidal line current) or by 12-pulse operation for harmonic mitigation.

Chopper-Fed DC Drives — Five Classes

DC–DC choppers feed DC motors from a DC bus and are grouped into five classes by the polarity of voltage and current they allow. The duty ratio is \(\delta = t_{on}/T\).

For the Class A buck chopper the ripple is \(\Delta I_a \approx V_s\delta(1-\delta)/(f_c L_a)\), maximal at \(\delta = 0.5\), with critical inductance \(L_{a,crit} = V_s\delta(1-\delta)/(f_c I_{a,min})\). The Class B boost chopper regenerates by boosting the motor EMF onto the bus, with \(\delta_{regen} = 1 - (E - I_aR_a)/V_s\) and \(P_{reg} = E|I_a|\). Class E (the H-bridge) gives full four-quadrant operation, and unipolar PWM offers four times lower ripple than bipolar PWM. Control is usually fixed-frequency PWM (variable \(\delta\)) or hysteretic, where \(i_a\) is bounded between limits and the frequency varies.

Closed-Loop Control — Cascade Structure

Modern DC drives use a cascade of two loops: a fast inner current loop and a slower outer speed loop. The armature current and mechanical responses are first-order:

The cascade arrangement is preferred because saturation of the speed-loop PI sets a natural current limit, the inner loop gives fast current (and hence torque) control, and the outer loop gives good speed regulation. A bandwidth ratio of about 10:1 between the loops keeps them decoupled.

Ward–Leonard System versus Modern Solid-State Drive

The classical Ward–Leonard system varies DC-motor speed through the field of a motor-generator set — an induction motor drives a DC generator whose field rheostat sets the armature voltage of the DC motor. It is smooth and inherently four-quadrant, but uses three rotating machines, is bulky and only 60–70% efficient. A modern solid-state drive replaces the generator set with a four-quadrant dual converter under closed current and speed loops — compact, 92–97% efficient and with millisecond response.

Induction Motor — Fundamentals

The induction motor runs below the rotating field. Its key speeds and slip are

The rotor quantities scale with slip: rotor frequency \(f_r = sf\), rotor EMF \(E_{2s} = sE_2\), rotor reactance \(X_{2s} = sX_2\), so \(Z_{2s} = R_2 + jsX_2\). The per-phase equivalent circuit places \(R_1\) and \(X_1\) in the stator, a magnetising branch \(X_m\), and the referred rotor branch \(X_2'\) in series with \(R_2'/s\).

Parameter Estimation — No-Load and Blocked-Rotor Tests

The equivalent-circuit parameters are found from two tests. In the no-load test (rated voltage and frequency, rotor free, \(s \to 0\)):

In the blocked-rotor test (rotor locked, \(s = 1\), reduced voltage to circulate rated current):

For a NEMA design-B machine, \(X_1 \approx X_2' \approx X_{eq}/2\) and \(R_2' \approx R_{eq} - R_1\). In practice \(R_1\) is measured separately by a DC test (with an AC/DC ratio of about 1.05–1.2), the blocked-rotor test is run at reduced frequency (about 25%) for high-efficiency designs to avoid skin-effect bias, and the no-load friction-and-windage loss is separated by extrapolating to zero voltage.

Power Flow and Efficiency

Power passes from the input through stator copper loss, the air gap, rotor copper loss and friction-and-windage loss to the shaft. The core relations are \(P_{ag} = 3I_2'^2 R_2'/s\), \(P_{cu2} = sP_{ag}\), \(P_{mech} = (1-s)P_{ag}\), \(P_{out} = P_{mech} - P_{f+w}\) and \(T_e = P_{ag}/\omega_s\).

Torque–Slip Characteristic

Torque rises with slip to a maximum (the breakdown torque) and then falls. The slip at maximum torque and the Kloss approximation are

The same characteristic describes motoring (\(0 \lt s \lt 1\)), plugging/braking (\(s \gt 1\)) and generating (\(s \lt 0\)).

Starting Methods for Induction Motors

For star–delta starting, \(I_Y = I_{DOL}/3\) and \(T_{Y/\Delta} = \tfrac{1}{3}T_{DOL}\). For an auto-transformer at tap \(k\), the line current is \(k^2 I_{DOL}\) and the starting torque is \(k^2 T_{DOL}\).

Braking of Induction Motors

Regenerative braking requires \(\omega_m \gt \omega_s\) (so \(s \lt 0\)) — from an active load (a descending hoist or downhill EV) or by reducing the supply frequency below the rotor speed in a VVVF drive; torque reverses and power flows to the source. Plugging swaps two stator phases to reverse \(n_s\); the slip becomes \(s = 2 - s_{old}\), torque opposes rotation and the motor decelerates rapidly, but it must be disconnected near zero speed to prevent reversal. DC dynamic braking disconnects the AC supply and injects DC into two stator phases, creating a stationary field in which the rotor acts as a loaded synchronous generator; the equivalent AC current for a star connection is \(I_{ac,eq} = \sqrt{2/3}\,I_{dc}\), and the braking torque is adjustable through \(I_{dc}\) and the rotor resistance.

Speed Control Methods — Summary

Induction-motor speed control divides into three families. Stator-side methods change the stator voltage (\(T \propto V^2\)), the frequency (which moves \(n_s\)), the V/f ratio (which preserves flux) or the pole count (discrete speeds). Rotor-side methods, available only on slip-ring machines, add rotor resistance (inefficient) or recover the slip power (static Kramer, Scherbius/DFIG, rotor-EMF injection). High-performance methods give DC-like control: field-oriented control (FOC), direct torque control (DTC) and model-predictive control (MPC).

Stator Voltage Control — AC Voltage Controller

Varying the stator voltage at fixed frequency through an anti-parallel SCR pair (a TRIAC for small motors) changes the torque without holding the rotor flux constant. Because torque scales with the square of voltage at any given slip,

the slip at maximum torque is unchanged while \(T_{max}\) falls as \(V_1^2\). For an R-load with firing angle \(\alpha\), \(V_{1,rms} = V_{ph}\sqrt{1 - \alpha/\pi + \sin 2\alpha/2\pi}\).

Constant V/f Control — the Workhorse

Below base frequency the drive operates in the constant-torque region; above base frequency it enters the constant-power (field-weakening) region. At low frequency the stator resistance drop becomes significant and the flux collapses, so a voltage boost is added: \(V_1 = V_{boost} + kf\).

VSI versus CSI — Two Inverter Philosophies

A voltage-source inverter (VSI) presents a stiff DC voltage through a large capacitor and outputs a six-step or PWM voltage; its fundamental in six-step mode is \(V_1 = 2V_d/\pi\), PWM uses modulation index \(m_a = V_{ctrl}/V_{tri}\), and four-quadrant operation needs anti-parallel diodes plus an active front-end. A current-source inverter (CSI) presents a stiff DC current through a large inductor and outputs a quasi-square current; the motor sets the voltage, four-quadrant operation is inherent, and it suits very large machines.

Six-Step VSI — Voltage and Harmonic Analysis

With \(180^\circ\) conduction, each switch conducts for half a cycle and three conduct at once, producing six unique switch combinations and a six-step phase voltage:

The fundamental is \(V_{an,1} = 2V_d/\pi\) with harmonics \(V_{an,n} = V_{an,1}/n\) for \(n = 6k\pm1\); the line-voltage fundamental is \(V_{ab,1} = 2\sqrt3 V_d/\pi\). The RMS values are \(V_{an,rms} = \sqrt{2/3}\,V_d \approx 0.471V_d\) and \(V_{ab,rms} \approx 0.816V_d\), with a voltage THD of about 31%.

PWM Techniques for VSI

Sinusoidal PWM compares a sinusoidal reference with a high-frequency triangular carrier. The modulation index \(m_a = \hat V_{ref}/\hat V_{tri}\) (with \(0 \le m_a \le 1\) in the linear range) sets the output: \(\hat V_{1,line} = m_a \tfrac{\sqrt3}{2}V_d\). Overmodulation (\(m_a \gt 1\)) adds low-order harmonics.

Advanced schemes improve on basic SPWM: space-vector PWM (SVPWM) raises bus utilisation by 15% (\(V_{1,max} = V_d/\sqrt3\)); third-harmonic injection gives a similar gain; discontinuous PWM lowers switching loss; and random PWM spreads the EMI spectrum.

Inverter Dead-Time and Compensation

Dead-time produces low-order distortion (5th, 7th harmonics), torque ripple and audible noise, worst at low fundamental amplitude. Compensation uses sign-of-current correction (add or subtract \(\Delta V\)), adaptive observers, or wide-bandgap devices with much smaller \(t_d\) (50 ns versus 2 µs). PWM also produces a common-mode voltage \(V_{cm} = (v_a+v_b+v_c)/3\) that drives bearing currents and EMI, mitigated by common-mode chokes, shielding and CM-aware modulation.

Multilevel Inverters for Medium-Voltage Drives

The common topologies are the neutral-point-clamped (NPC) inverter with clamping diodes, the flying-capacitor (FC) inverter with floating capacitors, the cascaded H-bridge (CHB) with isolated DC sources (modular), and the modular multilevel converter (MMC) used for HVDC and very-high-voltage drives.

Cycloconverter and Slip-Power Recovery

A cycloconverter performs direct AC–AC frequency conversion (output frequency typically \(\le f_{in}/3\)), used for very large low-speed drives such as ball mills and ship propulsion. On slip-ring machines, the slip power can be recovered rather than wasted.

The static Kramer drive rectifies the rotor slip power and inverts it back to the mains through a line-commutated inverter; the converter is rated at only about 30% of the motor rating, but it allows sub-synchronous operation only. The slip-power balance gives \(P_{ag} = T\omega_s\), \(P_{slip} = sP_{ag}\), and \(\omega_m = \omega_s\left(1 - V_{dc,inv}/V_{dc,rect}\right)\).

The Scherbius drive uses a bidirectional rotor converter (cycloconverter or back-to-back), enabling both sub- and super-synchronous operation. This is the basis of the doubly-fed induction generator (DFIG) used in wind turbines.

Vector Control (Field-Oriented Control)

There are two variants: direct FOC, where the flux is measured (Hall/search coil) or estimated from a voltage model, and indirect FOC, where the slip frequency is added to the rotor speed — simpler and more popular. FOC needs an accurate rotor resistance (it drifts with temperature), so online adaptation or a model-reference adaptive system (MRAS) is used, along with an encoder or a sensorless observer.

Direct Torque Control (DTC)

DTC controls torque and stator flux directly, without coordinate transformations or current loops. The torque is

where \(\delta\) is the angle between the stator and rotor flux and \(\sigma = 1 - L_m^2/(L_sL_r)\). Hysteresis controllers on \(|\psi_s|\) and \(T_e\), combined with the flux sector, select an optimal voltage vector from a switching table.

Synchronous Motor — Fundamentals

The per-phase relation is \(\mathbf{V} = \mathbf{E}_f + \mathbf{I}_a(R_a + jX_s)\). The developed power for a cylindrical rotor is

and for a salient-pole machine the two-reaction theory adds a reluctance term:

By varying the field current the power factor is controlled: an under-excited machine draws a lagging (inductive) current, unity excitation gives the lowest armature current, and an over-excited machine draws a leading (capacitive) current. Plotting \(I_a\) against \(I_f\) at constant power gives the characteristic V-curves.

Power-Angle Characteristic and Stability

The torque follows the power, \(T = P/\omega_s\). The salient-pole reluctance term peaks earlier in \(\delta\), giving a higher maximum power than the cylindrical machine. Pull-out occurs where \(dP/d\delta = 0\).

Synchronous Motor Drives — Types

Four drive arrangements are used. A cycloconverter performs direct AC–AC conversion (\(f_o \le f_{in}/3\)) for ball mills and ship propulsion. A VSI/PWM-fed drive converts constant-frequency input to variable-frequency PWM for PMSMs and general drives. A self-controlled (LCI) drive feeds commutating pulses from rotor position, behaving as a "commutatorless DC motor" for large rugged machines. A true synchronous drive uses an independent frequency source and needs a damper, suited to multi-motor textile mills.

In self-controlled mode the over-excited (leading power factor) motor supplies the reactive power needed for inverter commutation, with no hunting or loss of synchronism — making it popular for large salient-pole machines in the megawatt range. Synchronous motors are started by a damper winding (asynchronous start and pull-in), a pony motor, a variable-frequency inverter ramp, or reduced frequency for an LCI.

Load-Commutated Inverter (LCI) Drive

The lead angle \(\gamma\) between the motor current and back-EMF is set to about \(30^\circ\) for safe commutation. The inverter SCRs commutate from the motor EMF only above about 5% speed; below this a pulsed-mode start is used until natural commutation takes over. Rugged thyristors make the LCI the standard for very large pumps, compressors and ID fans up to about 100 MW.

VSI/CSI-Fed Synchronous Motor Drives

A VSI-fed synchronous motor uses sinusoidal PWM, with \(V_{1,LL} = m_a\tfrac{\sqrt3}{2}V_{dc}\) for SPWM or \(V_{1,LL} = m_a V_{dc}/\sqrt2\) for SVPWM (a 15% gain). Scalar V/f suits open-loop operation while FOC is used for high-performance PMSM drives. A CSI-fed machine (auto-sequentially commutated) produces quasi-square current, commutated by capacitors before pickup and then by the motor EMF, with inherent four-quadrant operation. Because there is no slip, the inverter output frequency directly sets the speed: \(n_s = 120f_1/P\).

PMSM — Permanent Magnet Synchronous Motor

In the rotor \(d\)-\(q\) frame the PMSM voltage and torque equations are

The torque has a magnet (alignment) component and a reluctance component. Two constructions exist: surface PM (SPM), where \(L_d = L_q\), non-salient, with no reluctance torque; and interior PM (IPM), where \(L_d \lt L_q\) and reluctance torque is available.

PMSM Control Strategies

The voltage-limit ellipse shrinks with speed, driving the operating trajectory toward \(i_d = -\psi_f/L_d\) in deep field weakening.

BLDC Motor — the Rugged Sibling

The BLDC machine differs from the PMSM in its waveforms and control: the back-EMF is trapezoidal (with a \(120^\circ\) flat top), the phase current is rectangular (\(120^\circ\) blocks), commutation is electronic at \(60^\circ\) intervals, and position is sensed by three Hall sensors (or estimated sensorlessly). With two phases conducting at any instant,

Stepper Motor Drives

The step angle is \( \beta = 360^\circ/(N_r\cdot m)\), where \(N_r\) is the number of rotor teeth and \(m\) the number of phases. Three types exist: variable-reluctance (soft-iron rotor, high speed, no detent), permanent-magnet (PM rotor, detent torque, low speed) and hybrid (VR + PM, typically \(\beta = 1.8^\circ\), the most common). Driving modes trade resolution for torque: full-step with one phase on, full-step with two phases on (more torque), half-step (\(\beta/2\)) and microstepping (\(\beta/n\), up to \(n = 256\)). The performance limits are the pull-in rate (maximum starting rate without losing steps), the pull-out rate (maximum running rate) and the holding torque (static torque at rated current).

Stepper Drive Electronics

An \(L\)-\(R\) winding driven by a step voltage rises with time constant \(\tau_e = L/R\):

The step time \(T_s = 1/f_{step}\) must satisfy \(T_s \gtrsim 4\tau_e\), which limits the maximum step rate. Drive techniques extend speed: an L/R drive adds series resistance to raise the voltage (with high losses); a bi-level or chopper drive uses a high voltage for fast current rise with a chopper limiting the current; and a constant-current chopper uses PWM hysteresis tracking. Microstepping feeds two sinusoidal currents \(90^\circ\) apart, \(i_a = I_p\cos(N_r\theta)\) and \(i_b = I_p\sin(N_r\theta)\), giving \(\beta_{micro} = \beta/n\).

Switched Reluctance Motor (SRM)

Torque comes from the co-energy: \(T_e = \left.\partial W'(i,\theta)/\partial\theta\right|_{i}\), which in the linear region is \(T_e = \tfrac{1}{2}i^2\,dL(\theta)/d\theta\). Each phase is fed by an asymmetric half-bridge with two switches and two diodes.

The per-phase voltage equation includes a motional back-EMF term:

Two control modes are used. At low speed, chopping limits the current by chopping the bus, controlling duty and turn-on angle. At high speed, single-pulse operation switches on at \(\theta_{on}\) and off at \(\theta_{off}\). The average torque and base speed are

Above \(\omega_b\) the machine enters its constant-power region. Acoustic noise is reduced by profiled current waveforms (torque-sharing functions), randomised PWM and skewed laminations.

Heating and Cooling of Motors

A motor is modelled as a first-order thermal body: the loss power either raises the temperature or is conducted away, \(P_{loss}\,dt = G_h c\,d\theta + A\lambda\,\theta\,dt\). Heating and cooling then follow exponentials:

The cooling time constant often exceeds the heating one because there is no forced cooling at standstill (\(\tau_c \gt \tau_h\)).

Classes of Duty (IEC 60034-1)

Eight standard duty types describe how a motor is loaded over time: S1 continuous (steady-state reached); S2 short-time (stops to cool to ambient); S3 intermittent periodic (cyclic duration factor below 100%); S4 intermittent with heavy starts; S5 intermittent with starting and braking; S6 continuous with intermittent load (never stops); S7 continuous with starting and braking; and S8 continuous with speed changes.

where \(t_N\) is the load (working) time and \(t_R\) the rest time. Most catalogue ratings assume S1; non-S1 duties permit a higher peak at a reduced cycle, verified by the equivalent-current method.

Motor Rating Selection Methods

For loads that vary over a cycle, an equivalent steady value is computed so the catalogue rating can be chosen as \(I_{rated} \ge I_{eq}\):

For short-time duty of length \(t_r\), the allowable overload factor is

so a shorter loading time permits a larger overload. For self-cooled motors that stop during idle, the effective idle (cooling) time is reduced by a factor \(\beta \approx 0.5\): \(t_{idle,eff} = \beta\,t_{idle}\).

Traction Drives — Physics

The tractive effort balances acceleration, gradient and resistance forces:

where \(M_e\) accounts for rotational inertia and \(r\) lumps rolling and wind resistance. A typical trip has four phases — acceleration, free run, coast and brake — and energy use is quoted as specific energy \(W_{sp}\) in kWh per tonne-kilometre.

Electric Vehicle (EV) Drive Architecture

An EV powertrain runs from the battery (48–800 V) through an optional DC/DC boost stage, an inverter (VSI with SVPWM), the traction motor (PMSM or IM), a single-speed gearbox and the wheels. A motor control unit implements FOC/MTPA/field-weakening, a battery management system handles cell balancing and state-of-charge, and regenerative braking returns energy from the wheels to the battery.

Wind & Renewable Energy Drives

The doubly-fed induction generator (DFIG) connects its stator directly to the grid and its rotor through a bidirectional (back-to-back) converter, giving variable speed over about \(\pm30\%\) of synchronous speed with a converter rated at only 30% of the nominal power — the standard for 1–5 MW turbines. The direct-drive PMSG is a low-speed, high-torque, multi-pole machine with a full-rated converter and no gearbox (used by Siemens, Enercon and GE). The captured power is

with \(C_p\) optimal at a tip-speed ratio \(\lambda_{opt} \approx 7\); maximum-power-point tracking drives the pitch and rotor speed to stay there. Grid codes require fault ride-through (LVRT), reactive support (\(\pm0.95\) power factor), frequency response and harmonic compliance.

Industrial Applications — Motor Selection Matrix

Power Quality Issues in Drives

The key metrics are the current total harmonic distortion and the true power factor:

Mitigation uses multi-pulse converters (12, 18 or 24-pulse), an active front-end (PWM rectifier), passive L or LCL filters, active power filters, and phase-shifting transformers. The IEEE 519-2014 limits depend on the short-circuit ratio \(I_{sc}/I_L\): for a low ratio (< 20) the total demand distortion must be \(\le 5\%\), while for a high ratio (> 1000) up to 20% is permitted.

Energy Efficiency in Drives

The system efficiency is the product of the converter, motor and transmission efficiencies, typically \(0.97 \times 0.93 \times 0.97 \approx 87\%\). The biggest savings come from variable-speed operation of fan and pump loads, where the affinity law makes power scale with the cube of speed:

so a 20% speed reduction saves about 49% of the power. Motor efficiency classes are defined by IEC 60034-30-1 as IE1 < IE2 < IE3 < IE4 < IE5 (ultra-premium).

Drive and Motor Protection Schemes

Protection is layered by speed of response: hardware (gate driver, fuse) acts in microseconds, firmware (DSP trip logic) in milliseconds, the application supervisor (drive/PLC) in tens of milliseconds, and plant relays and breakers in hundreds of milliseconds. Modern drives implement Safe Torque Off (STO) per IEC 61800-5-2, removing the inverter gate signals and certified to SIL3/PLe.

State-Space Model — DC Machine

With states \(\mathbf{x} = [i_a,\ \omega_m]^T\) and inputs \(\mathbf{u} = [V_a,\ T_L]^T\), the DC machine is a linear second-order system:

The transfer functions follow as

Induction Motor — d-q Model (Synchronous Frame)

In the synchronously rotating frame the stator and rotor voltage equations are

with flux linkages \(\psi_{ds} = L_s i_{ds} + L_m i_{dr}\) and \(\psi_{dr} = L_r i_{dr} + L_m i_{ds}\) (the \(q\)-axis is analogous). The electromagnetic torque and mechanical equation close the model:

Park Transformation

The Park transformation maps three-phase quantities to the rotating \(dq0\) frame:

with the inverse

The reference-frame choice depends on the application: \(\theta = 0\) gives the stationary \(\alpha\)-\(\beta\) (Clarke) frame; \(\theta = \omega_e t\) gives the synchronous frame (where balanced AC quantities become DC at steady state); and \(\theta = \theta_r\) gives the rotor frame used in PMSM FOC.

Feedback Devices — Position and Speed Sensors

Selection is driven by the required resolution and bandwidth, the environment (temperature, dust, oil, vibration), cost and cabling, and whether absolute position is needed at power-up. Resolver-to-digital conversion excites the resolver at 5–10 kHz and demodulates the sin/cos envelopes through a tracking converter — robust to noise, shock and temperature, hence standard for EV traction motors.

Sensorless Control Techniques

At medium and high speed the rotor position is estimated from the back-EMF: EMF integration (voltage model), flux observers, model-reference adaptive systems (MRAS) and Kalman or Luenberger observers. At zero and low speed, where the back-EMF vanishes, saliency-based methods are used — high-frequency signal injection, the INFORM method and rotating HF carriers.

Model Predictive Control (MPC)

In finite-control-set MPC (FCS-MPC) the cost is evaluated for all seven inverter voltage vectors and the minimum chosen:

Continuous-control-set MPC (CCS-MPC) instead solves a constrained optimisation over a prediction horizon:

solved by quadratic programming at each step (with explicit solutions possible). MPC handles constraints natively, is inherently multivariable and gives a very fast torque response, at the cost of a high computational burden and (for FCS) a variable switching frequency.

The following fully solved problems span the core competencies of the course — DC and AC drive sizing, converter firing-angle selection, induction-motor torque, thermal rating and traction dynamics. Each is worked in the style expected in university examinations and GATE.

Example 1 — DC Motor Speed Control

First establish the machine constant from rated data. The rated back-EMF and angular speed are

Because the load is constant-torque, the armature current stays at 10 A. At \( V = 150 \) V,

Example 2 — Three-Phase Full-Converter Drive

The mean converter output voltage for a three-phase full converter is

Example 3 — Induction-Motor Torque

The synchronous speed is \( n_s = 1500 \) rpm so \( \omega_s = 157.08 \) rad/s, and the per-phase voltage is \( V_1 = 400/\sqrt{3} = 230.9 \) V.

The starting-to-maximum ratio is \( T_{st}/T_{max} = 0.28 \), and the full-load torque at \( s = 0.05 \) is approximately 80 N·m.

Example 4 — V/f Control at Reduced Frequency

Scaling voltage and reactance with frequency:

Example 5 — Motor Rating for Intermittent Load

Using the equivalent-torque (RMS) method:

Example 6 — Thermal Overload

Example 7 — Chopper-Fed DC Drive

The armature current ripple, for \( L_a = 10 \) mH, \( f_c = 1 \) kHz and \( \delta = 0.435 \), is

Example 8 — Traction Speed–Time Dynamics

Wide-Bandgap Devices: SiC and GaN

Silicon power devices are approaching their physical limits. Wide-bandgap (WBG) semiconductors — silicon carbide (SiC) and gallium nitride (GaN) — break through those limits with higher blocking voltage (up to 10 kV for SiC), switching frequencies in the hundreds of kilohertz, operation above 200 °C, roughly half the switching loss of silicon, and consequently much smaller passive components.

WBG devices are already standard in EV traction inverters (Tesla, Lucid), wind and solar inverters, high-frequency DC–DC converters, and aerospace and rail systems.

Emerging Topics in Electric Drives

Several research and product directions are reshaping the field:

Smart Drives and Industry 4.0

Connected drives now embed intelligence and communication far beyond simple speed regulation:

Motor Current Signature Analysis (MCSA) is a powerful predictive-maintenance tool: a broken rotor bar produces sidebands around the line frequency in the current spectrum at

Mental Map of the Subject

The entire course radiates from a single object — the electric drive — through six branches: drive dynamics and stability, DC drives (rectifier- and chopper-fed), induction-motor drives (V/f, FOC and DTC), synchronous and permanent-magnet drives (LCI, cycloconverter, PMSM and BLDC), advanced control (sensorless estimation and model-predictive control), and applications (electric vehicles, rail, industry and wind).

Key Formulas — Quick Reference Card

Drive Dynamics

DC Motor

Converters

Induction Motor

Synchronous Motor

Field-Oriented Control and PMSM

Thermal and Rating

Traction and Wind

Park Transform (field angle)

References and Further Reading

Foundational Textbooks

1. G. K. Dubey, Fundamentals of Electrical Drives, 2nd ed., Narosa, 2001.
2. G. K. Dubey, Power Semiconductor Controlled Drives, Prentice-Hall, 1989.
3. R. Krishnan, Electric Motor Drives: Modeling, Analysis, and Control, Prentice-Hall, 2001.
4. R. Krishnan, Permanent Magnet Synchronous and Brushless DC Motor Drives, CRC Press, 2009.
5. R. Krishnan, Switched Reluctance Motor Drives, CRC Press, 2001.

Complementary Texts

6. B. K. Bose, Modern Power Electronics and AC Drives, Prentice-Hall, 2002.
7. W. Leonhard, Control of Electrical Drives, 3rd ed., Springer, 2001.
8. N. Mohan, Electric Machines and Drives: A First Course, Wiley, 2012.
9. P. C. Sen, Thyristor DC Drives, Wiley, 1981.
10. J. M. D. Murphy and F. G. Turnbull, Power Electronic Control of AC Motors, Pergamon, 1988.

Standards

• IEEE 519-2014 — harmonic limits for industrial drives
• IEC 60034 series — rotating electrical machines
• IEC 61800 series — adjustable-speed electrical power drive systems