Synchronous vs non-synchronous
Why use a synchronous buck instead of a diode? Are there cases where the diode is better?
| Level | ● screening |
| Track | Power |
| Really testing | Loss comparison and light-load behaviour. |
Going deeper
Extra explanation that is not in the book: worked numbers, where people go wrong, and how to say it in the room.
Worked example — 24 V → 3.3 V, 3 A (D ≈ 0.14).
- Schottky, Vf = 0.5 V: 0.5 × 3 × 0.86 ≈ 1.3 W — about 12% of the 9.9 W output.
- Low-side FET, 10 mΩ: 3² × 0.01 × 0.86 ≈ 0.08 W.
But the FET is not free. Gate drive: 10 nC × 5 V × 500 kHz = 25 mW. Body diode during dead time: 0.7 V × 3 A × (2 × 20 ns) × 500 kHz ≈ 42 mW. Still far less than the diode at 3 A.
At light load the picture flips. At 10 mA out (33 mW), that 25 mW of gate drive alone is most of the output power. That is why controllers skip pulses or emulate a diode at light load.
Where people go wrong
- Comparing Vf × I against I²R without the (1 − D) factor — at low output voltage the freewheel path conducts most of the cycle.
- Forgetting reverse recovery: a PN diode is much worse than a Schottky in this position.
Errata
No corrections known for this question. All errata
Ask about this question
Stuck on a step, or think something is wrong? Send it with PW-02 in the subject. Good questions get answered on this page, and corrections go into the next printing.