Boost Converter Design Guide
Learn how to design a step-up DC-DC converter from scratch — including inductor selection, MOSFET ratings, compensation, and real-world PCB layout best practices.
Skip the math — use our interactive designer:
Quick Answer
A boost converter steps up a lower DC input voltage to a higher DC output voltage. The key equation is D = 1 – (Vin / Vout) for the ideal duty cycle. For a 5V to 12V converter at 100kHz with 1A output, a typical design uses a 22µH inductor, 47µF output capacitor, and a switching MOSFET rated for at least 30V Vds. Boost converters are widely used in battery-powered devices, LED drivers, and USB PD applications.
How a Boost Converter Works
When the MOSFET switches ON, the inductor is connected directly across the input voltage. Current ramps up linearly, storing energy in the magnetic field. When the MOSFET switches OFF, the inductor voltage reverses polarity and adds to the input voltage, pushing current through the diode into the output capacitor — creating an output voltage higher than the input.
The output capacitor holds the voltage steady between switching cycles. The feedback loop adjusts the duty cycle to regulate Vout regardless of load or input variations.
Key Design Equations
1. Duty Cycle
For 5V to 12V at η=90%: D = 1 – (5×0.9)/12 = 0.625 (62.5%).
2. Inductor Selection
ΔIL is ripple current, typically 20–40% of Iin_max. Iin_max = Iout × Vout / (Vin × η).
3. Output Capacitor
ΔVout is the allowable output ripple (typically 50–100mV).
4. Peak Inductor Current
Critical for inductor saturation rating and MOSFET current rating.
Worked Example: 5V to 12V @ 1A
| Parameter | Value | Calculation |
|---|---|---|
| Input Voltage (Vin) | 5V | — |
| Output Voltage (Vout) | 12V | — |
| Output Current (Iout) | 1A | — |
| Switching Frequency (fsw) | 100kHz | — |
| Efficiency (η) | 90% | — |
| Duty Cycle (D) | 0.625 (62.5%) | D = 1-(5×0.9)/12 |
| Input Current (Iin_avg) | 2.67A | Iin = 1×12/(5×0.9) |
| Ripple Current (ΔIL) | 0.8A | 30% of 2.67A |
| Inductor (L) | 39.1µH → 33µH | L = 5×0.625/(0.8×100k) |
| Peak Current (IL_peak) | 3.07A | IL_peak = 2.67+0.8/2 |
| Output Capacitor (Cout) | 12.5µF → 22µF | ΔVout=50mV, D=0.625 |
Component Selection Guide
MOSFET Selection
| Parameter | Requirement | Explanation |
|---|---|---|
| Vds(max) | ≥ 1.2 × Vout (≥ 15V for 12V output) | Boost switch sees Vout + diode drop |
| Id(max) | ≥ 1.5 × IL_peak (≥ 5A) | Margin for startup inrush |
| Rds(on) | < 30mΩ | Lower Rds(on) = lower conduction loss |
| Qg | < 15nC for 100kHz | Gate charge determines switching loss |
Diode Selection
- Schottky diode recommended for low Vf (0.3–0.5V vs 0.7V for silicon)
- Reverse voltage rating: ≥ 1.3 × Vout
- Average forward current: ≥ Iout
- Peak surge current: ≥ IL_peak
Inductor Requirements
- Saturation current Isat ≥ 1.3 × IL_peak (at least 4A for this example)
- DCR < 50mΩ to minimize copper loss
- Shielded construction preferred to reduce radiated EMI
- Core material: Ferrite for 100kHz–1MHz; Powdered iron for < 100kHz
Output Capacitor
- MLCC (X7R) for low ESR — minimum 22µF, 25V rating
- Add 100–220µF electrolytic in parallel if load steps exceed 500mA
- Output ripple current equals Iout during switch-off time
Common Mistakes & How to Avoid Them
| Mistake | Fix |
|---|---|
| MOSFET Vds rating too low — avalanche breakdown at startup | Always rate Vds for at least 1.2 × Vout_max; consider startup overshoot |
| Right-half-plane zero causes poor transient response | Keep inductor small and crossover frequency below f_RHPZ = (1-D)²×Rload/(2π×L) |
| Input current underestimated — traces and connectors overheat | Iin = Iout × Vout/(Vin×η); at 5V to 12V/1A, input current is 2.67A, not 1A |
| No soft-start — massive inrush current charges Cout instantly | Implement soft-start: limit duty cycle ramp to 1–5ms |
| Output disconnected while switching — voltage soars uncontrolled | Always include overvoltage protection (OVP) — boost cannot run open-loop |
Right-Half-Plane Zero (RHPZ)
The boost converter’s most challenging control characteristic is the right-half-plane zero. Unlike buck converters, boost converters exhibit a non-minimum phase response: when duty cycle increases to raise output voltage, the output initially drops before rising. This limits the achievable control bandwidth.
For the 5V→12V example: f_RHPZ = (1-0.625)² × 12 / (2π × 33µH) = 0.14 × 12 / (2π × 33×10⁻⁶) ≈ 8.1kHz. The control loop crossover frequency must stay well below this — typically below f_RHPZ/5 ≈ 1.6kHz — or instability will occur.
FAQ
1. When should I use a boost converter vs a buck converter?
Use a boost converter when Vout > Vin (e.g., 3.7V battery to 5V USB, or 5V to 12V). Use a buck converter when Vout < Vin (e.g., 12V to 5V). If the input can be both higher and lower than the output, use a buck-boost or SEPIC converter instead.
2. Why is my boost converter output voltage collapsing under load?
Three common causes: (1) inductor saturation — the inductor can’t store enough energy per cycle, (2) current limit triggering — input current exceeds the controller’s limit, or (3) insufficient input capacitance causing input voltage droop. Check that Iin = Iout × Vout/(Vin × η) does not exceed your source capability.
3. How do I reduce output ripple in a boost converter?
Increase output capacitance, use low-ESR MLCC capacitors, add a second-stage LC filter on the output, or increase switching frequency. Note that boost converters inherently have higher output ripple than buck converters because output current is discontinuous (only flows during switch-off time).
4. Can I parallel two boost converters for more power?
Yes, but it requires careful design. Use a multiphase boost controller or external synchronization to interleave the switching phases. This cancels input and output ripple and doubles effective power. Never directly parallel two independent boost converters — they will fight each other due to small differences in reference voltages.
5. What happens if the load is suddenly removed from a boost converter?
Output voltage will soar rapidly because stored inductor energy has nowhere to go — this is the boost converter’s most dangerous failure mode. Always include overvoltage protection (OVP). A minimum load (preload resistor) of 1–5mA helps, but active OVP that turns off the switch when Vout exceeds the setpoint by 10% is essential.
6. Why does my boost converter draw huge current at startup?
At startup, the output capacitor is discharged (0V). The inductor sees Vin directly and current ramps with di/dt = Vin/L. Without soft-start, the peak current can reach many amps before the feedback loop takes control. Implement soft-start by gradually increasing the current limit or duty cycle over 1–5ms.
7. What is the efficiency limit of a boost converter?
Practical boost converters achieve 85–93% efficiency. Loss sources: MOSFET conduction (I²R), switching loss, diode forward drop (0.3–0.5V at output current), inductor core and copper losses, and ESR of capacitors. Higher step-up ratios reduce efficiency — a 3.7V to 50V boost will be far less efficient than a 10V to 12V boost.
8. How do I choose between asynchronous (diode) and synchronous (MOSFET) rectification?
For output currents below 0.5A, an asynchronous boost with Schottky diode is simpler and sufficient. Above 1A, synchronous rectification using a MOSFET becomes worthwhile — it eliminates the diode Vf loss (Pdiode = Vf × Iout ≈ 0.4 × 1 = 0.4W). The tradeoff is added complexity: you need a high-side gate driver and dead-time control.