Access Official Technical PDF

Share Summary:LinkedInEmail

The Vienna rectifier is a three-phase power-factor-correction (PFC) topology used to convert AC input power into a controlled DC output while shaping the input current. For engineers selecting a front end for a power supply or industrial converter, its appeal lies in a multilevel switching structure and the ability to achieve low current ripple. Its limits matter just as much: the basic topology is boost-type, non-isolated and unidirectional. Understanding both sides helps teams decide whether it fits their voltage range, power-flow requirements and system architecture.

How the Vienna rectifier shapes three-phase current

A conventional three-phase diode bridge conducts according to the line-to-line mains voltages. Its conduction pattern can create intervals of zero current and low-frequency harmonics, and the bridge does not provide active output-voltage control. A Vienna rectifier adds controlled switching to influence the bridge’s input voltages and conduction state.

The topology is boost-type: the minimum DC output voltage is constrained by the amplitude of the mains line-to-line voltage. Input-current shaping depends on the difference between the mains voltage and the voltage formed at the diode bridge input. Because that voltage formation depends on current direction, the available modulation range is also constrained. This interaction is important when designing the control system and selecting the operating voltage range.

The Vienna rectifier has a three-level bridge-leg characteristic. Its phase voltage can produce multiple voltage levels, which supports lower input-current ripple and can reduce the required input inductance. The control structure typically combines an outer DC-output-voltage loop with inner mains-current control. An additional loop balances the DC-link midpoint, using redundant switching states to manage the voltage balance.

Design benefits and engineering trade-offs

A multilevel waveform can offer an advantage over a standard two-level PWM rectifier. The comparison presented in the source material reports nine versus five voltage levels, with two to three times lower switching losses and a four- to six-fold reduction in inductance for the evaluated designs. These are topology and design comparisons, not guarantees for every implementation; component selection, switching frequency, control and thermal design all affect the final result.

The Vienna rectifier’s common-mode behavior also influences filter design. Its internal capacitor connection to a virtual star point can be used for common-mode EMI filtering without the same maximum-leakage-current constraint described for conventional common-mode capacitance. Designers may alternatively place common-mode filter stages on the DC side. The required number of filter stages depends in part on switching frequency, so EMI performance must be assessed alongside magnetics, layout and system requirements.

Demonstrators in the material illustrate the range of possible implementations. A 10 kW system operating from a 400 V AC input and producing an 800 V DC output reported 1.6% input-current THD at an 800 Hz mains frequency, with a 250 kHz switching frequency and 96.8% efficiency. The design used CoolMOS devices, SiC diodes and coldplate cooling. Such figures describe that specific demonstrator and operating point; they should not be treated as a general performance specification for the topology.

When another PFC rectifier may be a better fit

The basic Vienna rectifier has three notable boundaries: it does not provide isolation, its operation is boost-type, and it is unidirectional. Applications requiring a lower output voltage, a wide buck-boost range, galvanic isolation or bidirectional power flow may need a different approach.

Alternative concepts explored in the source include integrated active-filter rectifiers, one-out-of-three PWM rectifiers, Swiss rectifiers, buck-boost Y rectifiers and isolated matrix-type PFC rectifiers. They address different requirements rather than offering one universal replacement. For example, the Swiss rectifier is presented as providing controlled output voltage and sinusoidal mains current, while an isolated matrix-type approach builds on a dual-active-bridge concept and can support buck-boost operation. The appropriate choice depends on the intended power path and operating range, as well as circuit complexity and component stresses.

Reported demonstrator results also vary across these alternatives. An 8 kW Swiss rectifier demonstrator reported 99.26% efficiency at 60% rated load and approximately 0.5% mains-current THD at rated load. An isolated matrix-type PFC demonstrator reported 98.9% efficiency at 60% rated load, using zero-voltage switching, and approximately 4% mains-current THD at rated load. These results are not directly interchangeable: the designs differ in topology and operating conditions.

Designing beyond a single performance metric

Converter selection is a multi-variable exercise. Efficiency and power density are important, but they must be considered with cost, reliability, EMI behavior, thermal management, operating range and manufacturability. Increasing the number of voltage levels can improve some performance measures, yet it may also increase component count and design complexity. The source material emphasizes that only fault-tolerant approaches can make higher-complexity topologies viable where reliability is critical.

Switching faster can reduce the size of some passive components, but its impact on power density is not linear. In the Vienna rectifier examples, a tenfold increase in switching frequency corresponded to a twofold increase in power density. This illustrates why performance should be assessed at system level rather than inferred from switching frequency alone. Simulation, measurement and trade-off analysis can help teams understand how semiconductor technology, modulation and packaging combine to shape a design.

Conclusion

The Vienna rectifier remains a useful option when a three-phase, boost-type, non-isolated and unidirectional front end matches the application. Its multilevel operation can support controlled input current and low ripple, while midpoint balancing and EMI filtering require deliberate attention. Where isolation, buck operation or bidirectional power flow is necessary, other PFC architectures may be more suitable. For engineering teams, the practical decision is to compare complete converter systems against operating requirements—not to select a topology on efficiency or power density alone.

Access Official Technical PDF

About The Author