Direct Answer: For a standard query converting a 50W input through a buck converter operating at 92% efficiency, you will get 46W of usable output power and dissipate 4W as heat.
The formula to convert input power to output power is straightforward: P_out = P_in × (η / 100). Substituting our baseline values: 50W × 0.92 = 46W. The remaining power is lost primarily as heat through MOSFET switching, inductor core losses, and copper I²R resistance. Calculate the heat loss by subtracting the output from the input: 50W - 46W = 4W. While the math is simple, treating that 92% figure as a universal constant is a fast track to melting your PCB traces.
The Core Conversion: Input Power to Output and Heat
When sizing thermal vias or selecting an inductor, you need to know exactly how much heat your buck converter will dump into the ambient environment. The table below maps the neighboring values for a ±20% range around our 50W baseline, assuming a fixed 92% efficiency curve in Continuous Conduction Mode (CCM).
| Input Power (W) | Efficiency (η) | Output Power (W) | Heat Loss (W) |
|---|---|---|---|
| 40W | 92% | 36.8W | 3.2W |
| 45W | 92% | 41.4W | 3.6W |
| 50W | 92% | 46.0W | 4.0W |
| 55W | 92% | 50.6W | 4.4W |
| 60W | 92% | 55.2W | 4.8W |
A 4.8W heat loss in a small QFN or SOIC package without an exposed thermal pad will easily push the silicon junction temperature past 125°C, triggering thermal shutdown. Always pair these calculations with the IC's thermal resistance (θ_JA) spec.
What Assumptions Fix Your Efficiency Number?
The 92% efficiency figure printed on a datasheet is not a law of physics; it is a snapshot taken under highly specific laboratory conditions. Three primary assumptions fix this number:
- Continuous Conduction Mode (CCM): The inductor current never drops to zero. If your load drops and the converter enters Discontinuous Conduction Mode (DCM) or pulse-skipping, the efficiency curve shifts dramatically.
- Nominal Switching Frequency: Switching losses scale linearly with frequency. An IC clocked at 500kHz will yield higher efficiency than the same IC clocked at 2MHz, assuming identical load and voltage conditions.
- Specific V_in / V_out Ratio: Efficiency peaks when the duty cycle is near 50%.
How the Math Shifts Across Voltage Ratios
If you are stepping 12V down to 5V (duty cycle ~42%), a modern synchronous buck like the TI TPS5430 will easily hit 92-95% efficiency because conduction losses dominate and switching transitions are manageable. However, if you shift to stepping 24V down to 3.3V (duty cycle ~14%), the high-side MOSFET is on for a very short time. The gate charge losses and switching transition losses eat up a much larger percentage of the total power budget, often dragging real-world efficiency down to 85-88%, meaning your 50W input now yields 42.5W out and 7.5W of heat.
How the Math Shifts for Rectified Mains (120V/230V) and 3-Phase Designs
Standard buck converters are strictly DC-DC devices. However, when engineers ask about '120V vs 230V' in the context of buck topologies, they are usually referring to off-line rectified DC bus voltages. Rectified 120V AC yields roughly 170V DC, while rectified 230V AC yields roughly 325V DC.
If you attempt to step 325V DC down to 12V DC using a single-stage buck converter, your duty cycle drops to a microscopic 3.6%. The required minimum on-time of the MOSFET will likely exceed the switching period, causing the converter to skip pulses or fail to regulate. Furthermore, the massive voltage spike across the switching node during turn-off creates devastating switching losses, rendering standard efficiency calculations meaningless. For these high-ratio step-downs, you must use a Flyback, LLC resonant, or active-clamp forward topology instead.
The 3-Phase (Multi-Phase) Shift
In DC-DC design, '3-phase' refers to a 3-phase interleaved buck converter, commonly used in CPU Voltage Regulator Modules (VRMs). Instead of one massive inductor handling 60A, three smaller inductors handle 20A each, switching 120° out of phase. Because conduction losses scale with the square of the current (I²R), splitting the current across three phases drastically reduces MOSFET and inductor DCR (Direct Current Resistance) losses. A 3-phase interleaved buck will typically push efficiency up by 2% to 4% at heavy loads compared to a single-phase equivalent, shifting a 50W/92% calculation closer to 50W/95%.
When the Efficiency Conversion is Meaningless
The P_out = P_in × η conversion completely falls apart under light-load or no-load conditions. Every buck converter draws a quiescent current (I_q) just to keep its internal logic, voltage reference, and gate drivers alive.
Imagine your 50W-rated buck converter is powering a microcontroller in sleep mode, drawing just 10mA at 3.3V (33mW of output power). If the IC has a quiescent current of 5mA and is fed from a 12V source, it is consuming 60mW just to stay awake. Your input power is 93mW, your output is 33mW, and your actual efficiency has plummeted to 35%. Applying the datasheet's 92% nominal figure here would lead you to expect 85mW of output, which is physically impossible. For light loads, you must calculate losses based on I_q and switching overhead, not the peak CCM efficiency percentage.
Frequently Asked Questions
How does switching frequency affect buck converter efficiency?
Higher switching frequencies allow you to use physically smaller inductors and capacitors, but they directly increase switching losses (gate charge and transition overlap). Every time the MOSFETs toggle, a small packet of energy is burned. If you double the switching frequency from 500kHz to 1MHz, you roughly double the switching losses, which can drop your overall efficiency by 2% to 5% depending on the input voltage. Always use the lowest switching frequency that your ripple and transient response requirements will allow.
Why is my buck converter efficiency dropping at light loads?
At light loads, the fixed overhead of the converter (quiescent current, gate drive power, and core losses) becomes a massive percentage of the total power budget. To combat this, modern ICs like the advanced synchronous bucks employ 'Eco-mode', 'PFM' (Pulse Frequency Modulation), or 'Burst Mode'. These modes halt switching entirely when the output capacitor is charged, dropping the quiescent current to microamps and rescuing light-load efficiency, though at the cost of increased output voltage ripple.
Does the inductor choice change the buck converter efficiency calculation?
Absolutely. The inductor is often the largest source of loss after the IC itself. Inductor losses are split into two categories: DC losses (caused by the wire's DCR) and AC core losses (caused by hysteresis and eddy currents in the ferrite). Choosing a shielded inductor with a 40mΩ DCR instead of an unshielded one with 80mΩ DCR will cut your inductor conduction losses in half. If you are pushing 5A through that inductor, the 40mΩ part wastes 1W (5² × 0.04), while the 80mΩ part wastes 2W. That 1W difference directly alters your final heat dissipation calculation.
