How Does a Boost Converter Work?

Detailed Explanation

A boost converter (also called a step-up converter) is a switched-mode power supply topology that produces a DC output voltage higher than its input. The same principle of energy storage in an inductor that makes a buck converter work also makes a boost converter work — the difference is in how the switch and diode are arranged, which determines whether energy flows to step voltage down (buck) or step it up (boost).

Operating Principle

A boost converter has four key components: an inductor (L), a switch (usually a MOSFET), a catch diode (D), and an output capacitor (Cout). The switch alternates between two phases at a high frequency (typically 100 kHz to several MHz):

Phase 1 — Switch ON: The MOSFET closes, connecting the inductor directly from the input supply to ground. Current through the inductor ramps upward linearly, storing energy in its magnetic field. The output capacitor, charged from a previous cycle, supplies the load during this phase. The diode is reverse-biased during Phase 1 — no current flows from the inductor to the output.

Vin ─── L ─── (switch to GND)
                     ↑
              Cout supplies load

Phase 2 — Switch OFF: The MOSFET opens. The inductor resists the sudden interruption of current by generating a voltage spike — its polarity reverses (Lenz's law), so the inductor's output-facing terminal is now at Vin + VL. This sum voltage exceeds Vin, forward-biasing the diode and pushing current through it to the output capacitor and the load. The inductor current ramps downward as it delivers its stored energy.

Vin ─── L ──→ D ──→ Cout ──→ Load
              ↑
       VL adds to Vin

By adjusting the fraction of each period the switch is on, the controller sets the amount of energy stored per cycle and therefore the average output voltage.

Output Voltage and Duty Cycle

In continuous conduction mode (CCM), where inductor current never falls to zero between cycles, the relationship between input voltage, output voltage, and duty cycle D is:

Vout = Vin ÷ (1 − D)

where D = ton ÷ T (the fraction of the switching period the switch is on).

This formula assumes 100% efficiency. Real converters account for MOSFET on-resistance (RDSon), inductor winding resistance (DCR), and diode forward voltage drop. A controller's feedback loop adjusts D continuously to maintain the target output voltage as these losses and the input voltage vary.

Common Applications

The most widespread use of boost converters in embedded electronics is generating 5V from a single Li-ion cell (which discharges from 4.2V to 3.0V). A fixed 5V output is needed for:

• USB power output (USB-A or USB-C ports on portable battery banks)
• 5V-referenced sensors or displays that cannot run at 3.3V
• Driving LEDs that need a regulated current source above the battery voltage

Other common uses:

• 1.2–1.5V primary cell → 3.3V supply: boosts a single alkaline or NiMH cell to the 3.3V rail needed for a microcontroller.
• LED driver boost: white LEDs have a forward voltage of 3.0–3.5V; driving them from a 3.3V or lower supply requires a boost topology.
• Analog supply rail: some mixed-signal designs need a 5V or higher analog reference supply from a 3.3V digital rail.

For help choosing between a boost converter, a buck converter, and buck-boost topologies based on your input and output voltage ranges — including the single Li-ion cell → 3.3V scenario — see Buck, Boost, or Buck-Boost? How to Choose a DC-DC Converter Topology. For the higher-level decision between switching and linear regulators, see linear vs switching regulator: which should you use?.

Continuous vs Discontinuous Conduction Mode

Continuous Conduction Mode (CCM): inductor current flows continuously throughout both switch phases and never falls to zero. The duty cycle formula Vout = Vin ÷ (1 − D) applies. CCM has lower inductor current ripple, lower RMS current in the output capacitor, and better efficiency at high load. Most boost converters designed for moderate to heavy load operate in CCM.

Discontinuous Conduction Mode (DCM): at light loads, inductor current falls to zero during Phase 2 before the next Phase 1 begins. The converter enters a third idle phase where no current flows. DCM makes the Vout/Vin relationship nonlinear and load-dependent; it also causes the peak inductor current to be higher for the same average output current than CCM. Some controllers deliberately operate in DCM at light load (pulse-frequency modulation) to maintain efficiency — at very low load, CCM controllers can lose efficiency because switching losses are fixed per cycle regardless of output power.

Key Component Selection

Inductor (L): the inductor must handle the peak current without saturating and must have sufficiently low DCR (winding resistance). For a boost converter at switching frequency f, input voltage Vin, output voltage Vout, and load current Iout, the minimum inductance to stay in CCM is:

L_min = (Vin × D) ÷ (2 × f × ΔI)

where ΔI is the acceptable inductor current ripple (typically 20–40% of average inductor current). In practice, use the controller IC's datasheet design tool or reference design to select L — most TI, Diodes Inc., or Maxim boost controller datasheets include a design example with the exact calculation.

Output capacitor (Cout): the output capacitor filters the discontinuous diode current into a steady DC voltage. Use low-ESR ceramic capacitors (X5R or X7R, not X5R for high-temperature designs). The required capacitance depends on the acceptable output voltage ripple.

Catch diode (D): in non-synchronous boost converters, the catch diode must have a low forward voltage drop (to maintain efficiency) and fast reverse recovery (to avoid losses from shoot-through during Phase 1). Schottky diodes (SB140, SS3150, B340A) are the standard choice. In synchronous boost converters, a second MOSFET replaces the diode, reducing conduction losses and improving efficiency — particularly important at output voltages above 10V where diode drop is a significant fraction of total loss.

PCB Layout for Boost Converters

Boost converters generate high-frequency switching currents in a small loop (the switch, inductor, and catch diode). This switching loop is the primary source of conducted and radiated noise in a boost design. The layout principles are the same as for buck converters: keep the switching loop physically small, use a solid unbroken ground plane, and place input and output decoupling capacitors directly at the IC pins with short traces to ground.

For detailed layout guidance covering switch node copper area, ground plane continuity, and input/output capacitor placement, see how should you lay out a buck converter PCB? — the same principles apply directly to boost converter layout.

For power supply designs that require custom boost converter stages, PCB layout, and compliance with conducted emissions requirements, Zeus Design's circuit board design team covers topology selection through production-ready PCB layout — contact Zeus Design to discuss your power supply design.

Design Considerations

• Input capacitor is critical in boost converters: the boost converter's input current is continuous (unlike a buck, where input current is pulsed). However, the input capacitor still needs to handle the high-frequency ripple from the switching current. Place a low-ESR ceramic capacitor (100 nF to 10 µF, depending on input current) directly at the Vin and GND pins of the controller IC. See how should you place decoupling capacitors on a PCB? for placement rules.
• Maximum duty cycle limits boost ratio: real controllers cap D at approximately 85–95% to ensure adequate inductor reset time. This limits the maximum Vout/Vin ratio: at D_max = 0.85, Vout_max ≈ 6.7 × Vin. For higher boost ratios, consider a cascaded boost or an isolated topology (flyback, SEPIC).
• Current sensing for protection: boost converters can experience high peak inductor currents at startup or during output short-circuit conditions. Good controllers implement peak current limiting or hiccup-mode short-circuit protection. Verify the controller's current limit threshold is set below the inductor's saturation current rating — see inductor types and saturation current for how to select a power inductor with an appropriate Isat margin for switching converter use.
• Efficiency vs frequency tradeoff: higher switching frequency allows smaller inductors and capacitors (lower L and C values for the same ripple), reducing BOM cost and board space. However, switching losses (proportional to frequency) grow with frequency, reducing efficiency. Most compact boost converters for battery products operate at 1–2 MHz to balance these constraints.

Common Mistakes

• Using the Buck formula for Boost: the buck relationship Vout = Vin × D and the boost relationship Vout = Vin ÷ (1 − D) are frequently confused. At D = 0.5, a buck gives 50% of Vin; a boost gives 200% of Vin. Using the wrong formula produces an output voltage far from what was intended.
• Choosing an inductor based only on inductance value: inductors for switching converters must be rated for the peak current (not just the average), and their DCR must be low enough that the I²R loss in the winding doesn't dominate efficiency. A 10 µH inductor rated at 1A saturation current in a 2A boost design saturates immediately, causing the output voltage to collapse and potentially destroying the MOSFET.
• Ignoring right-half-plane zero in the feedback loop: boost converters have a right-half-plane (RHP) zero in their control-to-output transfer function that does not appear in buck converters. This zero limits the achievable closed-loop bandwidth — the feedback loop cannot be made as fast as a comparable buck. Attempting to close the loop faster than the RHP zero frequency causes oscillation. Controller datasheets specify the compensation design approach for the boost topology specifically; do not reuse buck compensation designs.
• Expecting the same PCB layout to work as a buck: the switching loop geometry and the position of the switch node differ between buck and boost topologies. Reusing a buck layout footprint for a boost results in suboptimal switching loop area and increased EMI. Lay out each topology following its specific reference design.