Dual Active Bridge Control: Topology, Modulation, and Real-Time Validation

The dual active bridge has become the default choice wherever a system needs isolated, two-way DC-DC conversion. But the topology is only as good as the dual active bridge control strategy that runs it. The same hardware can deliver near-99% efficiency or bleed power into circulating currents depending entirely on how its phase shifts are commanded, modeled, and validated.

 the dual active bridge topology, the modulation hierarchy that defines its control, the soft-switching and backflow-power physics every controller must respect, and—critically—how to validate that control in real time through dual active bridge simulation on FPGA-based hardware-in-the-loop platforms.

What a Dual Active Bridge Is

The dual active bridge, almost always shortened to DAB, is a high-frequency, galvanically isolated DC-DC converter designed to move power in either direction between two DC buses. It shows up across an enormous power range, from roughly a hundred watts in small auxiliary supplies up to several megawatts in grid-scale equipment.

Structurally the dual active bridge converter is elegant: two full bridges, built from controllable power switches, sit on either side of a high-frequency transformer. Because both bridges are actively switched rather than one acting as a passive rectifier, energy can travel from primary to secondary or from secondary to primary using exactly the same hardware. That single property—symmetry between the two sides—is what gives the DAB its name and its defining behavior.

The Physical Structure of the Dual Active Bridge Topology

Three components do essentially all of the work in a DAB.

The first is the pair of active bridges. Each bridge contains four controllable switches—commonly silicon MOSFETs, silicon carbide (SiC) MOSFETs, gallium nitride (GaN) devices, or IGBTs depending on the voltage and power class—so a standard single-phase DAB uses eight switches in total. Each bridge converts DC into a high-frequency AC square wave, and because both sides do this simultaneously under active control, power flows either way without any change in topology.

The second component is the high-frequency transformer. Its job is twofold: it provides galvanic isolation between the two DC buses, a requirement mandated by safety standards in applications such as electric vehicle charging, and it sets the voltage step-up or step-down ratio through its turns ratio. Running the transformer at high frequency rather than line frequency is what allows it, along with the associated filter components, to shrink dramatically in size and weight. This is the underlying reason DAB converters achieve such high power density.

The third component, often overlooked, is the series inductor that carries energy from one side to the other during each switching cycle. In many designs this inductor is not a separate physical part; it is realized using the transformer’s own leakage inductance, sometimes supplemented with an external inductor. Using leakage inductance this way reduces parts count and losses, but it means the transformer must be designed with a tightly controlled leakage value, which complicates the magnetics design considerably.

This three-part dual active bridge topology—two H-bridges, one transformer, one energy-transfer inductor—is the basis of every variant, from single-phase DABs to three-phase, multilevel, and multiport (multi-active-bridge) configurations.

Steady-State Principle of Operation and Inductor Current Dynamics

The core mechanism of dual active bridge control relies on modulating the phase displacement between the two AC voltage square waves generated at the terminals of the high-frequency isolation transformer. This control method, known as Single-Phase Shift (SPS) modulation, maintains a constant fifty percent duty cycle for both the primary and secondary H-bridges.

The diagonal switch pairs in the primary bridge are switched on and off in a complementary manner with appropriate dead-time intervals to prevent short circuits. This generates a two-level high-frequency AC square-wave voltage with a peak amplitude equal to the primary DC bus voltage. Similarly, the secondary-side switches are modulated to generate a high-frequency square-wave voltage with a peak amplitude equal to the secondary DC bus voltage.

By introducing a controlled phase shift angle between the primary and secondary square waves, a voltage differential develops across the series inductor. This voltage differential drives an AC current through the inductor and transformer windings, enabling controlled bidirectional power transfer.

To model the continuous current dynamics within a switching half-period, the operating states are analyzed under forward power transfer, where the primary voltage waveform leads the referred secondary voltage waveform. The half-period divides into two switching intervals.

Interval 1. At the beginning of the half-period, the primary-side switches turn on, applying a positive voltage across the primary terminals of the transformer. Meanwhile, the secondary-side switches continue to apply a negative voltage to the secondary terminals. Referred to the primary side, the voltage across the series energy-transfer inductor equals the sum of the primary DC bus voltage and the referred secondary DC voltage. As a result, the inductor current increases linearly from its initial negative peak value.

Interval 2. At the phase-shift transition point, the secondary-side switches commute, reversing the polarity of the referred secondary terminal voltage to positive, while the primary voltage remains positive. The voltage across the series inductor changes to the difference between the primary DC bus voltage and the referred secondary DC voltage. The slope of the inductor current alters accordingly, and the current continues to evolve linearly toward its next transition value.

By leveraging the steady-state half-wave symmetry of the AC waveforms, the inductor current at the end of the switching half-cycle must be equal in magnitude but opposite in sign to the current at the beginning. This boundary condition allows for analytical determination of the transition currents as functions of the primary voltage, referred secondary voltage, series inductance, and switching frequency.

The Power-Transfer Equation

The active power transferred from primary to secondary is calculated by averaging the instantaneous input power over a switching half-cycle. Under single-phase shift control, this power-transfer characteristic is parabolic. The active power is directly proportional to the product of the primary voltage, the referred secondary voltage, and a phase shift factor, and inversely proportional to the series inductance and the switching frequency.

This relationship demonstrates that maximum active power transfer occurs when the normalized phase shift is exactly one-half, corresponding to a ninety-degree phase displacement between the primary and secondary square waves. Operating with a phase shift greater than ninety degrees is avoided in practical systems because the derivative of power with respect to phase shift becomes negative, which degrades control stability and generates excessive circulating currents.

These three levers—primary and secondary voltages, series inductance, and switching frequency—define the design envelope, while the phase shift is the real-time control handle the digital controller commands cycle by cycle.

Zero-Voltage Switching Commutation Physics

One of the most valuable features of the dual active bridge topology is its ability to naturally achieve Zero-Voltage Switching (ZVS) for all primary and secondary semiconductor devices. Achieving soft-switching transitions eliminates turn-on losses, minimizes electromagnetic interference, and allows operation at high switching frequencies.

The physical mechanism of ZVS relies on utilizing the energy stored in the series inductor to charge and discharge the parasitic output capacitances of the semiconductor devices during the programmed dead-time intervals. Consider a single leg in the primary H-bridge consisting of an upper switch and a lower switch, each with an associated parallel parasitic capacitance and anti-parallel body diode. During a switching transition, the commutation sequence proceeds in three phases.

Phase 1: Turn-off transition and capacitive charging/discharging. Prior to the transition, the lower switch is conducting positive current, holding the switch node voltage at ground. When the gate signal for the lower switch is forced low, the switch turns off. Because the current through the series inductor cannot change instantaneously, the turn-off current is redirected from the semiconductor channel into the parasitic output capacitances of both switches. The current charges the parasitic capacitance of the lower switch toward the DC bus voltage while simultaneously discharging the parasitic capacitance of the upper switch toward zero.

Phase 2: Body-diode conduction and voltage clamping. When the voltage across the parasitic capacitance of the upper switch reaches zero, the negative voltage across its anti-parallel body diode forward-biases it. This clamps the switch node voltage to the primary DC bus voltage, and the inductor current freewheels through the body diode. Because the voltage across the upper switch is clamped to a diode drop, its drain-to-source voltage is effectively zero.

Phase 3: Zero-voltage turn-on. The gate-drive signal for the upper switch is applied while its anti-parallel body diode is actively conducting. Since the voltage across the channel is zero, the switch turns on with zero-voltage switching, eliminating turn-on losses and preventing capacitive discharge dissipation. After this turn-on event, the inductor current crosses zero and changes direction, transferring current flow smoothly from the body diode to the semiconductor channel.

For a successful ZVS transition, the inductor current must have the correct polarity and sufficient energy at the switching instant to fully charge and discharge the parasitic output capacitances. If the inductor energy is too low, the switch node voltage will not reach the opposite rail during the dead-time interval, resulting in hard-switching transitions and capacitive losses.

The soft-switching boundaries are highly dependent on the voltage matching ratio, defined as the ratio of the referred secondary voltage to the primary voltage. When the voltage matching ratio is greater than unity, the phase shift must exceed a minimum threshold to ensure the primary switches achieve ZVS. Conversely, when the ratio is less than unity, the phase shift must be large enough to guarantee soft-switching for the secondary-side switches.

When the converter operates outside these boundaries—typically during light-load conditions where the phase shift ratio approaches zero—the inductor current is insufficient to discharge the parasitic capacitances, resulting in hard switching. Active hardware enhancements, such as auxiliary commutation inductors, can supply additional reactive current, though they increase complexity and RMS currents. Advanced multi-phase-shift modulation schemes are a highly effective alternative to extend the ZVS range without increasing physical component count.

Backflow Power and Circulating Currents

A major performance challenge of dual active bridge control under SPS is backflow power, also called reactive circulating power: the electrical power that flows back to the transmitting source during a portion of each switching half-cycle, contrary to the overall direction of active power conversion.

This phenomenon occurs when the instantaneous voltage at the primary terminals of the transformer and the inductor current have opposite signs. Under these conditions the instantaneous power is negative, indicating energy is being returned to the primary DC link rather than transferred to the load.

During a forward power-transfer half-cycle, the primary bridge voltage is positive while the inductor current starts negative. The current rises linearly and crosses zero at a specific instant during the half-period. During the initial interval before the zero-crossing, power is returned to the source. The duration of this reverse flow is determined by the switching period, the voltage matching ratio, and the phase shift ratio.

The magnitude of backflow power is highly dependent on voltage mismatch and phase shift ratio. Under perfectly matched voltage conditions, backflow power is minimized. However, when the voltage matching ratio deviates significantly from unity, backflow power increases rapidly. Under light-load conditions where the phase shift is small, backflow power remains high.

The physical consequence is a significant increase in circulating currents. These currents do not contribute to net power delivered to the load but flow continuously through the semiconductor channels, body diodes, transformer windings, and series inductors. This leads to high RMS currents and substantial conduction losses. Consequently, the efficiency of a DAB operating under SPS degrades severely at light loads or high voltage mismatches, making efficiency optimization a key challenge for systems that operate across wide voltage and load ranges.

This is the central tension every advanced dual active bridge control scheme exists to resolve: ZVS demands a negative switching-instant current, but that same current is what creates backflow. The two objectives pull in opposite directions, and the modulation strategy is how an engineer negotiates between them.

Advanced Phase-Shift Modulation Strategies

To mitigate backflow power, minimize RMS currents, and extend the soft-switching range across wide voltage limits, advanced multi-phase-shift modulation schemes introduce additional control degrees of freedom by using inner phase shifts within the individual H-bridges, transforming the transformer terminal voltages from two-level square waves into three-level square waves.

Single-Phase Shift (SPS)

SPS is the baseline: a single control variable—the phase shift between the two bridges. It is the simplest scheme to implement and offers fast dynamics and inherent buck-boost capability, but it has a limited ZVS region, higher current stress, and significant backflow power when the load is light or the voltage ratio departs from unity.

Extended-Phase-Shift (EPS) Modulation

EPS introduces an inner phase-shift ratio between the diagonally operating switches of the primary-side H-bridge. The secondary-side H-bridge continues to operate with a fifty percent duty cycle, producing a three-level voltage waveform with a zero-voltage interval at the transformer primary winding. The external phase shift controls the magnitude and direction of active power transfer, while the inner phase shift lets the zero-voltage interval be matched dynamically to the operating voltage ratio—suppressing peak current stress and backflow power.

Dual-Phase-Shift (DPS) Modulation

DPS introduces an identical inner phase-shift ratio to both the primary and secondary H-bridges, generating matching three-level voltage waveforms with equal zero-voltage intervals at both transformer terminals. The external phase shift regulates active power. Because the inner phase shifts are identical, the control algorithm is less complex than other three-variable strategies while remaining highly effective at reducing current stress and eliminating backflow power under wide voltage gains.

Under DPS, the average output current, RMS inductor current, active power, and apparent power are all expressed as functions of the inner phase shifts, the external phase shift, and the voltage ratio. This lets the digital controller select optimal phase-shift combinations that minimize current stress for a given power demand.

Triple-Phase-Shift (TPS) Modulation

TPS is the most generalized control scheme, providing three independent control variables:

• Das primary H-bridge inner phase shift, controlling the primary zero-voltage interval.
• Das secondary H-bridge inner phase shift, controlling the secondary zero-voltage interval.
• Das external phase shift between the rising edges of the primary and secondary voltage waveforms.

Under TPS there are twelve distinct operating modes defined by the relationships among the three phase shifts. This versatility allows the control system to optimize performance across the entire load range, maintain ZVS for all devices from zero to full load, and guarantee the minimum-RMS-current trajectory for any arbitrary voltage ratio. Notably, SPS, EPS, and DPS can all be treated as special cases of TPS.

Evaluating the mode constraints and executing the corresponding piece-wise algorithms in real time is where computational platform capability becomes decisive—matching the transient demands of high-power grids requires the controller to recalculate optimal operating parameters every cycle.