Electronics Guide

H-Bridge Drive Topology

The H-bridge represents the fundamental power stage topology for brushed DC motor control, consisting of four switching elements arranged to allow current flow through the motor in either direction. When diagonal pairs of switches conduct, current flows through the motor, with the direction determined by which pair is active. This bidirectional capability enables both forward and reverse motor operation as well as regenerative braking.

Modern H-bridge implementations use MOSFETs or IGBTs as switching elements, selected based on voltage and current requirements. Low-voltage applications typically employ MOSFETs for their low on-resistance and fast switching capability. Higher voltage applications may use IGBTs, which offer better voltage blocking capability at the expense of higher switching losses. Integrated H-bridge driver ICs simplify design for lower-power applications by combining power switches, gate drivers, and protection circuits in a single package.

Protection features are essential in H-bridge designs. Shoot-through, where both switches in a half-bridge conduct simultaneously, creates a short circuit that can destroy the power stage. Dead-time insertion ensures that one switch is fully off before the complementary switch turns on. Overcurrent protection monitors motor current and shuts down the drive if safe limits are exceeded. Thermal monitoring prevents device damage from excessive heating.

PWM Control Methods

Pulse-width modulation provides the primary mechanism for controlling motor speed and torque in brushed DC drives. By rapidly switching the power devices on and off, PWM effectively controls the average voltage applied to the motor. The motor's inductance acts as a low-pass filter, smoothing the switched voltage into a nearly constant current that produces smooth torque.

Sign-magnitude PWM applies PWM to control motor speed while separate direction signals determine rotation direction. One half-bridge pair controls the positive direction while the other pair controls the negative direction. This approach is straightforward to implement but produces discontinuous current at direction reversals, potentially causing torque ripple during zero-speed crossings.

Locked antiphase PWM, also known as bipolar PWM, continuously switches the motor between positive and negative voltage at the PWM frequency. A 50% duty cycle produces zero average voltage and zero speed. Duty cycles above 50% produce forward rotation while duty cycles below 50% produce reverse rotation. This approach eliminates direction-reversal discontinuities and provides smooth operation through zero speed, making it preferred for servo applications.

PWM frequency selection involves trade-offs between several factors. Higher frequencies reduce audible noise and current ripple but increase switching losses in the power devices. Lower frequencies improve efficiency but may produce audible noise and increased current ripple that causes additional motor heating. Frequencies between 10 kHz and 25 kHz are common, staying above the audible range while maintaining reasonable efficiency.

Current Control and Limiting

Current control is fundamental to brushed DC motor drives, as motor torque is directly proportional to armature current. Open-loop voltage control provides only indirect speed control, with actual speed depending on load torque. Closed-loop current control enables precise torque regulation and protects both the motor and drive from overcurrent conditions during starting, stall, and rapid acceleration.

Peak current limiting prevents damage during transient conditions such as motor starting or sudden load changes. When sensed current exceeds the limit, the controller reduces PWM duty cycle or temporarily disables switching to allow current to decay. Cycle-by-cycle current limiting responds within a single PWM period, providing fast protection. Average current limiting responds to the mean current over multiple PWM cycles, allowing brief transient overcurrents that the motor can safely tolerate.

Current sensing techniques include resistive shunts, Hall-effect sensors, and transformer-isolated current sensors. Low-side shunt sensing measures current in the ground return path, providing a simple and inexpensive solution but requiring careful common-mode noise management. High-side sensing measures current in the supply path, avoiding ground disturbances but requiring level-shifting circuitry. In-line sensing with isolated sensors offers the best signal quality but at higher cost.

Six-Step Commutation

Six-step commutation, also called trapezoidal commutation, represents the most straightforward approach to brushless DC motor control. The three-phase motor is driven by sequentially energizing pairs of windings, creating a rotating magnetic field that the rotor follows. At any instant, current flows through two of the three phases while the third phase remains unenergized. Six distinct commutation states complete one electrical cycle.

The timing of commutation state transitions is critical for optimal motor performance. Commutating too early or too late relative to rotor position reduces torque production and increases losses. Hall-effect position sensors mounted in the motor provide rotor position feedback for commutation timing. Three sensors spaced 120 electrical degrees apart generate a three-bit code that directly indicates which commutation state to apply.

Sensorless commutation eliminates Hall sensors by detecting the back-EMF generated in the unenergized winding. As the rotor magnets pass the unenergized stator windings, they induce a voltage that crosses zero at the optimal commutation point. The controller monitors this zero-crossing and commutates a fixed delay later. This approach requires the motor to be spinning to generate detectable back-EMF, necessitating special starting techniques such as open-loop forced commutation or high-frequency injection.

Six-step commutation produces torque ripple because the stator magnetic field advances in discrete 60-degree steps rather than rotating smoothly. This ripple manifests as vibration and acoustic noise that may be objectionable in some applications. Field-oriented control techniques can eliminate this ripple but require more sophisticated drives.

Field-Oriented Control

Field-oriented control, also known as vector control, transforms the inherently coupled AC quantities in a brushless motor into decoupled DC quantities that can be controlled independently. By mathematically transforming measured currents from the stationary stator reference frame to a reference frame aligned with the rotor magnetic field, the controller can independently regulate the flux-producing and torque-producing current components.

The coordinate transformation process begins with the Clarke transformation, which converts three-phase stator currents into a two-axis stationary reference frame. The Park transformation then rotates this reference frame to align with the rotor magnetic field, producing the direct-axis and quadrature-axis current components. The direct-axis current controls field strength while the quadrature-axis current controls torque.

Precise rotor position knowledge is essential for field-oriented control. Encoder or resolver feedback provides accurate position information for the coordinate transformations. Sensorless field-oriented control estimates rotor position from measured voltages and currents using observers or model-based estimators. These techniques work well at higher speeds but face challenges at low speeds where back-EMF signals are small.

Field-oriented control eliminates the torque ripple inherent in six-step commutation by maintaining optimal current vector orientation continuously. The resulting smooth torque production enables precision motion control applications. The computational requirements are higher than for six-step commutation, typically requiring a digital signal processor or high-performance microcontroller, but modern integrated motor control devices make field-oriented control practical for a wide range of applications.

Inverter Topologies

The standard three-phase inverter topology for brushless DC motors consists of three half-bridge power stages, one for each motor phase. Each half-bridge contains a high-side and low-side switch that connect the phase terminal to either the positive supply rail or ground. The six switches enable the drive to apply voltage vectors in any direction within the hexagonal voltage space.

Space vector modulation optimizes the use of the available DC bus voltage by selecting combinations of active voltage vectors that synthesize the desired output voltage. Compared to simple sinusoidal PWM, space vector modulation can achieve approximately 15% higher output voltage for the same DC bus voltage. The technique also offers flexibility in distributing null vector time to minimize switching losses or reduce common-mode voltage.

Dead-time effects in the inverter distort the output voltage from its commanded value. During dead time, when both switches in a half-bridge are off, the phase current determines which diode conducts and thus what voltage appears at the output. The net effect is a voltage error that depends on current polarity. Dead-time compensation algorithms measure or estimate current polarity and adjust the PWM commands to cancel the distortion.

Drive Modes and Excitation Sequences

Full-step drive energizes one or two phases at a time, producing the motor's basic step angle. Single-phase excitation, also called wave drive, energizes only one phase at a time, producing the lowest torque but also the lowest power consumption. Two-phase excitation energizes two phases simultaneously, producing approximately 40% more torque than single-phase excitation at the cost of higher current consumption.

Half-step drive alternates between single-phase and two-phase excitation, doubling the number of steps per revolution and reducing step angle by half. This technique improves positioning resolution without motor modifications but produces unequal torque at alternate step positions. The torque variation can cause resonance problems at certain speeds.

Microstepping further divides each full step by precisely controlling the current ratio between phases. By gradually shifting current from one phase to another rather than switching abruptly, the motor's rotor moves in much smaller increments. Microstepping factors of 8, 16, 32, or even 256 are common, enabling step angles of a fraction of a degree. The smooth motion also dramatically reduces acoustic noise and mechanical resonance compared to full-step operation.

The sinusoidal current profiles required for microstepping demand more sophisticated current control than simple full-step operation. Current-mode drive with pulse-width modulation maintains the programmed current waveforms regardless of motor speed and back-EMF. The current loop bandwidth must be high enough to accurately reproduce the desired current profiles at the maximum stepping rate.

Current Control Techniques

Voltage-mode drive, the simplest stepper drive approach, applies fixed voltages to the motor windings. Performance is limited because the winding inductance restricts how quickly current can change, limiting high-speed torque. The motor's rated current flows only at low speeds; at higher speeds, current and torque fall rapidly. Voltage-mode drives are adequate only for low-performance applications.

Chopper current regulation maintains constant current regardless of motor speed by rapidly switching the supply voltage. When the current reaches the target value, the drive turns off until current decays to a lower threshold, then turns on again. This cycle repeats at a rate determined by the motor inductance and the hysteresis band between upper and lower thresholds. Current-mode drives dramatically extend the useful speed range compared to voltage-mode drives.

PWM current control provides more precise current regulation than simple chopper control by using a fixed switching frequency with variable duty cycle. A current feedback loop adjusts the duty cycle to maintain the desired current despite changes in motor back-EMF and winding resistance. The fixed switching frequency simplifies EMI filtering and produces more predictable acoustic characteristics.

Decay mode selection affects current regulation dynamics when the drive switches off. Fast decay connects the winding to the opposite supply polarity, causing current to decrease rapidly. Slow decay simply shorts the winding, allowing current to circulate and decay slowly through resistive losses. Mixed decay combines both approaches, using fast decay initially to reach zero current quickly, then switching to slow decay to limit negative current buildup. The optimal decay strategy depends on motor parameters and operating conditions.

Anti-Resonance Techniques

Stepper motors exhibit mechanical resonance at certain speeds where the step frequency matches the natural frequency of the rotor and load. At resonance, the motor may lose steps, stall, or produce excessive vibration. The resonant frequency depends on motor torque, rotor inertia, and load characteristics, typically falling in the range of 100 to 400 full steps per second.

Microstepping reduces resonance amplitude by producing smoother torque pulses with less energy at the fundamental step frequency. The smaller steps excite the resonance less strongly, though they do not eliminate it entirely. Higher microstepping factors generally provide better resonance suppression.

Electronic damping techniques actively suppress resonance by modifying the drive currents based on detected or estimated rotor oscillation. Back-EMF sensing can detect rotor velocity variations that indicate resonance. The controller then adjusts phase currents to apply damping torque that opposes the oscillation. Some advanced stepper drives incorporate model-based observers that estimate rotor position and velocity for closed-loop damping without external sensors.

Acceleration profiling avoids extended operation at resonant speeds by accelerating quickly through the resonance region. S-curve or jerk-limited acceleration profiles are smoother than trapezoidal profiles and couple less energy into mechanical resonances. The control system must know where resonances occur and plan motion profiles accordingly.

Variable Frequency Drives

Variable frequency drives control induction motor speed by adjusting the frequency and voltage of the applied AC power. The motor speed is determined primarily by the supply frequency and the number of poles, with slip providing a small speed reduction that increases with load torque. By varying the frequency, the drive can adjust motor speed over a wide range while maintaining optimal flux levels and efficient operation.

Voltage-to-frequency ratio control, often called V/f control or scalar control, maintains constant flux by scaling voltage proportionally with frequency. At low frequencies, an additional voltage boost compensates for the increased significance of stator resistance voltage drop. V/f control provides adequate performance for many applications, including fans, pumps, and conveyors, where precise speed and torque control are not critical.

The rectifier front-end of a typical variable frequency drive converts AC line power to DC, creating a stable DC bus voltage that supplies the inverter. Simple drives use diode rectifiers, while more sophisticated drives employ active front-end rectifiers that provide regeneration capability and power factor correction. The DC link capacitor bank smooths the rectified voltage and provides energy storage for brief transients.

The inverter section synthesizes variable-frequency AC from the DC bus using the same three-phase topology employed in brushless DC drives. PWM techniques control both the fundamental frequency and voltage magnitude. The motor's inductance filters the PWM switching, producing nearly sinusoidal currents despite the switched voltage waveforms.

Direct Torque Control

Direct torque control provides high-performance induction motor control without the complexity of coordinate transformations used in field-oriented control. The technique directly controls stator flux magnitude and electromagnetic torque using hysteresis controllers that select inverter voltage vectors to keep flux and torque within tolerance bands around their commanded values.

Flux estimation integrates the stator voltage minus resistive drop to determine stator flux magnitude and angle. Torque estimation uses the cross product of stator flux and stator current. These estimates update at the control sampling rate, enabling rapid response to flux and torque errors. The accuracy of flux estimation, particularly at low speeds where integration errors accumulate, remains a challenge for sensorless direct torque control.

The switching table maps combinations of flux error, torque error, and flux position to optimal voltage vector selections. When both flux and torque are within tolerance, a zero vector is applied to minimize switching. When errors exist, active vectors are selected to correct the errors while advancing the stator flux vector appropriately. This approach produces variable switching frequency that depends on operating conditions.

Direct torque control achieves fast torque response because voltage vector selection occurs at every sampling period without the filtering inherent in current regulators. Torque response times of a few milliseconds are typical, matching or exceeding field-oriented control performance. The variable switching frequency produces a broader spectrum of acoustic noise compared to fixed-frequency PWM.

Indirect Field-Oriented Control

Indirect field-oriented control for induction motors requires knowledge of the slip frequency to properly orient current commands in the rotor flux reference frame. Unlike synchronous machines where the rotor field angle equals the rotor mechanical angle times pole pairs, the induction motor's rotor flux angle depends on both rotor position and the slip required to produce the commanded torque.

The slip frequency calculation uses motor parameters and commanded torque current to determine how much the rotor flux angle leads or lags the rotor mechanical angle. This calculated slip frequency, added to measured rotor speed, produces the reference frame angle for current control. Parameter accuracy directly affects control performance, making rotor resistance estimation particularly important as this parameter varies significantly with temperature.

Current regulators in the rotating reference frame control the flux-producing and torque-producing current components independently. The flux current command typically remains constant in normal operation, changing only during field weakening at high speeds. The torque current command varies to meet demanded torque, with speed regulators providing the command in closed-loop speed control applications.

Flux weakening enables operation above base speed by reducing the flux current command as speed increases beyond the point where full voltage is required. Reduced flux produces proportionally reduced back-EMF, allowing higher speeds with the available voltage. Torque capability decreases inversely with speed in the field-weakening region because the same current produces less force with reduced flux.

Basic Operating Principles

Torque in a switched reluctance motor depends on the rate of change of inductance with rotor position and the square of phase current. Positive torque results when current flows during rising inductance as rotor poles approach stator poles. Negative torque results when current flows during falling inductance as rotor poles depart. The drive must precisely control when each phase conducts to produce the desired motoring or generating action.

The highly nonlinear relationship between torque, current, and rotor position complicates switched reluctance motor control. Inductance varies by a factor of six or more between aligned and unaligned positions. Magnetic saturation further distorts the torque-current relationship. Accurate torque control requires either extensive characterization data or sophisticated real-time modeling.

Position sensing for switched reluctance motors can use encoders, resolvers, or sensorless techniques. The large inductance variation with position enables sensorless operation by injecting test pulses and measuring the current response. High inductance indicates that the rotor pole is approaching alignment while low inductance indicates misalignment. These measurements, combined with speed estimation from commutation timing, can provide sufficient position information for closed-loop control.

Asymmetric Bridge Topology

The asymmetric half-bridge topology is the standard power stage for switched reluctance motor drives. Each phase winding connects between two switches arranged with one high-side and one low-side device. Diodes provide freewheeling paths for current when the switches open. This topology requires two switches and two diodes per phase but provides full control flexibility including regenerative operation.

Three operating states are available for each phase. In the magnetization state, both switches conduct, applying full bus voltage to build phase current. In the freewheeling state, one switch conducts while current circulates through the other diode, maintaining current with minimal voltage. In the demagnetization state, both switches open, forcing current through both diodes and applying negative bus voltage to quickly extinguish the current.

Current regulation uses hysteresis or PWM techniques similar to those in other motor types. The choice between hard chopping (alternating magnetization and demagnetization) and soft chopping (alternating magnetization and freewheeling) affects efficiency and acoustic noise. Soft chopping produces lower current ripple and less acoustic noise at the cost of slower current decay when the phase must be turned off.

Torque Ripple Minimization

Torque ripple is the primary challenge in switched reluctance motor control, arising from the discontinuous nature of torque production as phases sequentially conduct. Each phase produces a pulse of torque during its conduction interval, and the sum of overlapping pulses from multiple phases determines instantaneous motor torque. Minimizing variations in this sum requires careful control of turn-on angle, turn-off angle, and current profiles.

Current profiling techniques shape the phase current waveform to produce flatter total torque. Rather than simple square-wave currents, optimized profiles may include ramped or sinusoidal shapes that compensate for the varying torque-per-ampere characteristic across the conduction interval. Generating these profiles requires accurate motor characterization and sufficient controller bandwidth to reproduce the commanded shapes.

Torque sharing functions explicitly define how total demanded torque should be distributed among phases based on rotor position. Each phase's contribution varies smoothly with position, ramping up as the phase enters its torque-producing region and ramping down as it exits. The sum of all phase contributions equals the demanded torque at every rotor position. Implementing torque sharing requires fast current regulation and accurate position information.

Iterative learning control and adaptive algorithms can optimize commutation parameters based on measured torque ripple or vibration. The controller experiments with small adjustments to turn-on timing, turn-off timing, or current profiles and retains changes that reduce ripple. Over time, these techniques can customize control to individual motors and operating conditions.

Servo System Architecture

The classical servo architecture uses cascaded control loops with position on the outer loop, velocity on the middle loop, and current or torque on the inner loop. Each loop operates at progressively higher bandwidth, with the current loop typically running at tens of kilohertz, the velocity loop at several kilohertz, and the position loop at hundreds of hertz to a few kilohertz. This cascade structure simplifies tuning and provides clear separation of concerns.

Position feedback typically uses encoders, resolvers, or linear scales depending on accuracy requirements and environmental conditions. Incremental encoders provide relative position changes that must be referenced at startup, while absolute encoders maintain position knowledge through power cycles. Resolution requirements vary from hundreds of counts per revolution for simple applications to millions of counts for precision positioning.

Velocity feedback may derive from differentiating position measurements or from dedicated velocity sensors such as tachometers. Differentiation amplifies noise and quantization errors, particularly with lower-resolution encoders. Filtering reduces noise but adds phase lag that limits control bandwidth. Higher-resolution encoders and encoder interpolation improve velocity estimation quality without filtering penalties.

Modern servo systems often implement additional feedback processing, including disturbance observers that estimate and compensate for load variations, friction observers that improve low-speed smoothness, and notch filters that attenuate mechanical resonances. These techniques extend achievable performance beyond what simple PID control can provide.

Trajectory Generation

Trajectory generation creates motion profiles that define how position, velocity, and acceleration should evolve over time to move from one point to another. The profile must respect the physical limits of the motor and load while achieving the shortest move time, smoothest motion, or other optimization objectives. The servo drive's trajectory generator computes these profiles in real time based on move commands and constraint parameters.

Trapezoidal velocity profiles accelerate at constant rate to a cruise velocity, maintain that velocity, then decelerate at constant rate to the target position. These profiles are simple to compute and achieve moves in near-minimum time. However, the instantaneous changes in acceleration at profile corners excite mechanical resonances and produce jerky motion in some applications.

S-curve profiles smooth the acceleration transitions by limiting jerk, the rate of change of acceleration. The acceleration ramps gradually rather than changing instantaneously, reducing excitation of mechanical resonances and producing smoother motion. The trade-off is slightly longer move times compared to trapezoidal profiles with the same peak velocity and acceleration limits.

Coordinated motion of multiple axes requires synchronized trajectory generation across all participating axes. Each axis must begin and end its motion simultaneously, matching velocity and acceleration profiles throughout the move. Path interpolation in Cartesian space followed by inverse kinematics enables straight-line tool paths even when individual axis motions are complex curves.

Position Control Algorithms

Proportional-integral-derivative control remains the foundation of most servo position loops. The proportional term provides stiffness that resists position errors. The integral term eliminates steady-state errors but can cause overshoot and hunting if gain is set too high. The derivative term provides damping that reduces overshoot and improves stability margins. Proper tuning balances these terms for the specific application requirements.

Feedforward control improves tracking performance by commanding a portion of the required control effort based on the desired trajectory rather than waiting for position errors to develop. Velocity feedforward adds a term proportional to commanded velocity, compensating for the lag inherent in feedback-only control. Acceleration feedforward adds a term proportional to commanded acceleration, improving tracking during rapid acceleration and deceleration.

Model-based control techniques use mathematical models of the motor and load to compute optimal control actions. Model predictive control looks ahead over a prediction horizon, optimizing the control sequence that will produce the best future trajectory. These approaches can explicitly handle constraints such as current limits and predict interactions between axes in multi-axis systems.

Adaptive control adjusts controller parameters in real time based on measured system response. Self-tuning algorithms can compensate for load changes, friction variations, and other disturbances that would degrade performance with fixed gains. Auto-tuning procedures characterize the system during commissioning, automatically selecting appropriate gains without manual adjustment.

Current Sensing and Regulation

Current feedback is fundamental to motor control, whether for protection, torque control, or commutation timing. The sensing method must provide adequate bandwidth for the control loop while rejecting common-mode noise and handling the large voltage swings present in switching power stages. Sampling synchronization with PWM timing ensures that measurements capture representative current values rather than switching transients.

Three-phase motor drives can reconstruct all three phase currents from measurements of only two phases, since the three currents must sum to zero in a balanced system. This reconstruction enables lower-cost designs with fewer current sensors. Alternatively, DC link current sensing measures the current returning to the bus capacitor, which equals one of the phase currents during each active vector. Careful timing of current samples relative to PWM states enables reconstruction of all phase currents from a single sensor.

Current loop bandwidth requirements vary with application. Six-step commutated brushless motors need only moderate bandwidth for basic overcurrent protection. Field-oriented control systems require higher bandwidth to track sinusoidal current commands without excessive phase lag. Servo applications demand the highest bandwidth to achieve fast torque response for tight position control.

Position and Speed Sensing

Incremental encoders generate pulses as the shaft rotates, with two channels in quadrature providing direction information along with position increments. The controller counts pulses and tracks direction to maintain position. An index pulse once per revolution enables position referencing. Resolution ranges from hundreds to millions of counts per revolution depending on the encoder design and any electronic interpolation.

Absolute encoders directly output position within a single revolution or over multiple revolutions without requiring a reference operation. Single-turn absolute encoders use optical or magnetic code patterns to indicate position. Multi-turn versions add gear trains or battery-backed counters to track complete rotations. Communication interfaces such as EnDat, BiSS, or SSI transfer position data digitally to the drive.

Resolvers provide robust position sensing using electromagnetic induction principles without sensitive optical components. The drive excites a reference winding with a high-frequency carrier and measures the induced voltages in sine and cosine output windings. The ratio of these voltages indicates rotor angle. Resolver-to-digital converters process the signals and output position and velocity data. Resolvers excel in harsh environments including high temperatures, vibration, and contamination.

Sensorless position estimation extracts rotor position from electrical measurements without mechanical sensors. Back-EMF based methods work well at moderate to high speeds where back-EMF signals are strong. High-frequency injection methods modulate the drive voltages with test signals and detect position-dependent impedance variations, enabling sensorless operation at low speeds and standstill. The computational requirements and commissioning complexity of sensorless methods must be weighed against sensor cost and reliability considerations.

Protection Systems

Overcurrent protection prevents damage to power devices and motor windings from excessive current. Hardware comparators can shut down the gate drives within microseconds of detecting an overcurrent condition. Software-based protection operates more slowly but can implement more sophisticated algorithms such as I-squared-t thermal models that account for short-term overloads the motor can safely tolerate.

Overvoltage protection prevents damage from regenerative energy when the motor decelerates rapidly or the load drives the motor. When bus voltage rises toward dangerous levels, the drive can activate braking resistors to dissipate excess energy, reduce deceleration rate, or transition to a freewheeling state that prevents further energy transfer to the bus. Line-regenerative drives can return energy to the supply, eliminating overvoltage concerns during regeneration.

Thermal protection monitors power device and motor temperatures to prevent overheating. Power device temperature sensing uses integrated temperature sensors or external thermistors mounted near the devices. Motor temperature sensing typically uses embedded thermistors or thermal models that estimate winding temperature from current and ambient conditions. The drive reduces output capability or shuts down when temperatures approach limits.

Ground fault detection identifies insulation failures that could present shock hazards or damage equipment. Differential current measurement detects imbalances indicating current leaking to ground. Ground fault interruption opens the circuit before dangerous current levels can develop. Requirements vary by application, with stringent requirements in equipment that may contact operators.