A buck-boost converter is a DC-DC switching power supply topology that can produce an output voltage either lower or higher than its input voltage, depending on the duty cycle of its switching element(s). The term covers several related topologies, including inverting and non-inverting variants. It is used when the input voltage range overlaps the desired output voltage, making a pure buck or pure boost topology insufficient.
In practice
Buck-boost converters appear frequently in battery-powered embedded systems where the supply rail can swing above and below the required system voltage across the discharge curve. A single-cell Li-ion battery, for example, ranges from roughly 4.2 V fully charged to 2.8-3.0 V at cutoff; a 3.3 V rail sits inside that window, so neither a pure buck nor a pure boost can regulate across the full discharge range without the other topology taking over at the crossover point.
Common practical topologies include the inverting buck-boost (single switch, single inductor, output polarity is opposite to input), the four-switch synchronous buck-boost (two high-side and two low-side FETs around a single inductor, output polarity same as input), and the SEPIC and Cuk topologies (which are related in usage but are generally considered distinct converter families rather than buck-boost subtypes). Integrated controller ICs such as the TI TPS63xxx family and the Analog Devices (formerly Linear Technology) LTC3780 implement non-inverting buck-boost regulation in small packages suited to embedded designs.
A key challenge in buck-boost operation is efficiency in the transition region where input and output voltages are nearly equal. Four-switch topologies typically cycle through buck mode, a transition mode, and boost mode; the exact switching behavior during transition varies by controller and modulation scheme, and if the control loop is poorly tuned or the transition band is wide, efficiency dips and inductor ripple current spikes during transition. Inductor selection is important: the peak current in boost mode is higher than in buck mode for the same load, so worst-case inductor saturation current must be sized for the boost phase. The blog posts "Lost Secrets of the H-Bridge, Part III: Practical Issues of Inductor and Capacitor Ripple Current" and "Lost Secrets of the H-Bridge, Part II: Ripple Current in the DC Link Capacitor" discuss ripple current sizing principles that apply directly to inductor and capacitor selection in switching converters.
EMI considerations are more demanding than for a pure buck: the inductor current waveform and switching node voltage can be more complex, especially in four-switch designs during mode transitions. Output and input capacitor selection, layout, and snubbing all follow the same principles as other hard-switched converters, but the wide operating range means the designer must verify stability and ripple across the full input-voltage span, not just at a single operating point.
Frequently asked
When should I use a buck-boost instead of separate buck and boost converters?
Use a buck-boost when the input voltage range spans the target output voltage and component count or PCB area makes two converters impractical. Separate converters can be more efficient at the extremes of the input range, but a well-designed four-switch buck-boost with a narrow transition band often achieves comparable efficiency in a single-inductor solution.
What is the difference between an inverting buck-boost and a non-inverting (four-switch) buck-boost?
An inverting buck-boost uses a single switch and single
inductor but produces an output whose polarity is opposite to the input (e.g., a positive input yields a negative output). A four-switch non-inverting buck-boost uses two half-bridges around one inductor to keep the output polarity the same as the input, which is required in most embedded power rails. The four-switch topology is more complex and can have higher conduction losses due to four FETs in the current path, though actual losses depend on FET Rds(on), gate-drive strategy, and operating point. It is nonetheless the more common choice in battery-powered designs where polarity inversion is unacceptable.
Why does efficiency often dip when input and output voltages are nearly equal?
In a four-switch design, when Vin is close to Vout, the converter must operate in a transition mode where all four switches are active simultaneously. This increases switching losses and
inductor ripple current. Some controllers minimize this by making the transition band as narrow as possible or by using a dedicated transition-mode duty cycle; others use analog control schemes that blend buck and boost duty cycles smoothly.
How do I size the inductor for a buck-boost converter?
Size the saturation current rating for the worst-case peak current, which typically occurs in boost mode at minimum input voltage and maximum load. The RMS current rating must also account for the higher duty cycles seen in boost operation. Inductance value sets the ripple current magnitude; higher inductance reduces ripple but increases physical size. Always verify that the
inductor remains out of saturation across the full Vin range, not just at the nominal operating point.
Can a microcontroller implement buck-boost control directly using PWM outputs?
In principle, yes: a
microcontroller with two complementary
PWM channels and a fast
ADC can implement a digital buck-boost control loop, and some MCUs include dedicated high-resolution PWM peripherals (such as the STM32 HRTIM or the dsPIC33 PWM module) that are well-suited to power-conversion control tasks including buck-boost regulation. In practice, the control loop bandwidth required for stable buck-boost regulation -- especially through the transition region -- often demands loop update rates of tens to hundreds of kilohertz, which can challenge general-purpose MCU peripherals and leaves little CPU headroom. Dedicated analog or mixed-signal controller ICs are more common in production designs.
Differentiators vs similar concepts
A
buck converter only steps voltage down (Vout less than Vin); a
boost converter only steps voltage up (Vout greater than Vin). A buck-boost handles both directions but at the cost of added complexity, additional switches or passive components, and a potential efficiency penalty near the crossover point. The SEPIC (Single-Ended Primary Inductance Converter) and Cuk topologies are related in application -- both can produce non-inverting step-up/step-down conversion -- but are distinct converter families with different current and output characteristics; each uses two
inductors and a coupling
capacitor, avoiding the four-switch complexity but requiring more magnetic components and being less common in compact, high-current embedded designs. The term "buck-boost" is sometimes loosely applied to any converter that spans both step-up and step-down operation; context is needed to determine whether an inverting single-switch, a non-inverting four-switch, or a SEPIC/Cuk topology is intended.
