Synchronous Buck Converters
Replace the freewheel diode with a controlled MOSFET to reduce loss while managing dead time and reverse current.
A synchronous buck replaces a one-way check valve with a second controlled valve. It wastes less pressure, but both valves must be timed carefully.
A conventional buck uses a diode to carry inductor current when the high-side switch is off. A synchronous buck replaces that diode with a low-side MOSFET. When timed correctly, current flows through the MOSFET's low channel resistance instead of a relatively large diode forward drop.
The two MOSFET commands are complementary with dead time. During dead time the body diode often conducts briefly. Too little dead time risks shoot-through; too much increases diode loss. Controllers may adapt dead time because the best value changes with devices, current, temperature and layout.
At light load, the inductor current may reach zero and try to reverse. Some controllers allow reverse current for continuous conduction; others enter diode-emulation or pulse-skipping modes to improve efficiency. The operating mode affects ripple, noise and whether energy can flow back toward the input.
Higher efficiency does not remove switching loss, gate-drive loss, inductor loss or layout requirements. Compare total loss across the expected input, output and load range instead of comparing only diode drop with RDS(on).
Replace the freewheel diode with a controlled MOSFET to reduce loss while managing dead time and reverse current.
Core ideas
- The low-side MOSFET performs synchronous rectification.
- Dead time trades shoot-through risk against diode conduction loss.
- Light-load control determines whether negative inductor current is allowed.
Diode loss estimate: P ≈ IavgVFMOSFET conduction estimate: P ≈ Irms²RDS(on)Ideal buck relation: Vout ≈ D × VinWork through a practical example
1Compare a 5 A freewheel path with VF = 0.5 V against a 15 mΩ MOSFET.
2The diode estimate is 5 × 0.5 = 2.5 W while the MOSFET estimate is 5² × 0.015 = 0.375 W during equal conduction time.
3The real comparison must scale each loss by its conduction interval and add switching, body-diode and gate-drive losses.
4Verify controller behavior at light load and during startup before assuming current always remains positive.
Assuming the low-side MOSFET can simply stay on whenever the high side is off, including conditions where reverse current or startup states are unsafe.