If you are stepping down a 12V input to a 5V output to drive a 3A load (15W output) using a typical 90% efficient buck converter, your required input current is 1.39 Amps. The foundational formula fixing this answer is based on power conservation adjusted for efficiency: I_in = (V_out × I_out) / (V_in × η). Substituting our exact values: I_in = (5V × 3A) / (12V × 0.90) = 15W / 10.8W = 1.388A. The primary assumption that fixes this answer is the converter's continuous efficiency (η) at that specific load point; if efficiency drops to 80% due to thermal throttling or light-load burst mode, the input current jumps to 1.56A.
Below is a quick-reference table showing how the input current shifts across a ±20% range of output current for this exact 12V-to-5V scenario, assuming a flat 90% efficiency:
| Output Current (A) | Output Power (W) | Input Current @ 12V (A) | Input Power (W) |
|---|---|---|---|
| 2.4A (-20%) | 12.0W | 1.11A | 13.3W |
| 3.0A (Nominal) | 15.0W | 1.39A | 16.7W |
| 3.6A (+20%) | 18.0W | 1.67A | 20.0W |
Real-World DC-DC Converter IC and Module Data
Theoretical 90% efficiency is a useful baseline, but bench reality varies wildly depending on the silicon, switching frequency, and inductor core losses. When sizing your upstream power supply or battery pack, you must use the actual efficiency curves of your chosen regulator. Below is a data-dense specification table of common DC-DC converters used in 2026 hobbyist and commercial designs, reflecting real-world continuous limits rather than marketing peak numbers.
| Part / Module | Topology | Input Range (V) | Max Continuous Out (A) | Typical Efficiency | Switching Freq | Approx. Cost (USD) |
|---|---|---|---|---|---|---|
| LM2596 Module | Buck | 4.5 - 40V | 2.0A (w/ heatsink) | 73% - 82% | 150 kHz | $1.50 - $2.50 |
| XL4015 5A Module | Buck | 5.0 - 32V | 3.5A (w/ heatsink) | 85% - 92% | 180 kHz | $2.50 - $4.00 |
| TI TPS5430DDA | Buck (IC) | 5.5 - 36V | 3.0A | 88% - 95% | 500 kHz | $1.80 - $2.20 |
| MPS MP2315GJ | Buck (IC) | 4.5 - 24V | 3.0A | 92% - 97% | 1.4 MHz | $1.10 - $1.50 |
| TI LM3478 | Boost (IC) | 2.9 - 40V | Depends on FET | 80% - 90% | 50k - 1MHz | $2.00 - $2.80 |
Source: Manufacturer datasheets from Texas Instruments and Monolithic Power Systems.
Why AC Mains Metrics (Power Factor, 3-Phase) Are Meaningless Here
A common point of confusion for makers transitioning from AC wiring to PCB design is asking how DC-DC converter math shifts for 120V vs 230V AC mains, or how 3-phase power affects the calculation. The direct answer: it doesn't.
Power factor (PF), phase angles, and AC RMS voltages are strictly metrics for the upstream AC-DC power supply (like a Mean Well LRS-350 or an ATX computer supply). By the time power reaches the DC-DC converter stage, it has already been rectified and bulk-filtered into DC. The DC-DC stage only "sees" the DC bus voltage. Therefore, a 120V AC input and a 230V AC input will both yield roughly the same 12V DC bus (assuming the AC-DC supply is rated for universal input), meaning the DC-DC converter's input current calculation remains exactly 1.39A in both scenarios.
The
I_in formula breaks down and becomes meaningless under three conditions:1. Dropout: When
V_in approaches V_out plus the dropout voltage, the duty cycle hits 100%, the switch stays on continuously, and efficiency collapses.2. Light-Load Burst Mode: Below ~10% load, modern ICs like the MP2315 enter pulse-skipping mode. The average efficiency curve becomes highly non-linear, making the simple η multiplier inaccurate.
3. Thermal Shutdown: If the silicon junction hits 150°C, the converter will fold back or shut down entirely, rendering steady-state power equations void.
Sizing Input Traces and Capacitors for the Calculated Current
Knowing your input current is 1.39A is only step one. The most frequent bench failure I see with DC-DC converters is not the IC burning up, but the input capacitor exploding or the PCB trace delaminating. This happens because the input current to a buck converter is not a smooth 1.39A DC; it is a harsh, chopped square wave.
While the average input current is 1.39A, the RMS ripple current (I_cin(rms)) that the input capacitor must handle is calculated as:
I_cin(rms) = I_out × √(D × (1 - D))
Where D is the duty cycle (V_out / V_in = 5/12 = 0.416). Substituting our values: 3A × √(0.416 × 0.584) = 3A × 0.492 = 1.47A RMS.
Actionable Sizing Rules:
- Capacitor Selection: You must use low-ESR ceramic (MLCC) or polymer capacitors on the input. A standard electrolytic capacitor rated for 1.39A DC will overheat and vent from the 1.47A high-frequency RMS ripple. Place a 10µF to 22µF X7R ceramic capacitor as physically close to the IC's VIN and PGND pins as possible.
- PCB Trace Width: For a 1.39A continuous DC input on a standard 1oz copper outer layer, the IPC-2221 standard dictates a minimum trace width of roughly 0.5mm (20 mils) for a 10°C temperature rise. However, because you must account for the peak switching current (which can be
I_out + (I_ripple / 2)), sizing the input trace to 1.0mm (40 mils) provides a necessary safety margin and reduces parasitic inductance. - Wire Sizing: If wiring this module from a battery, 22 AWG stranded silicone wire (rated for ~7A in free air) is more than sufficient for the 1.39A average draw, but keep the leads under 6 inches to prevent input voltage ringing.
FAQ: DC-DC Conversion Edge Cases
Does a boost converter draw more input current than a buck converter for the same output?
Yes. In a boost topology (e.g., stepping 3.7V Li-ion up to 12V at 1A), the input current is significantly higher because the input voltage is lower. Using the same formula: I_in = (12V × 1A) / (3.7V × 0.85) = 12W / 3.145W = 3.81A. The input source must be sized for this higher continuous draw, and the inductor must not saturate at the peak input current.
How do I measure the actual input current on my bench?
Do not rely on the display of a cheap variable bench power supply; their sampling rates are too slow to capture the dynamic load changes of a switching regulator. Use a true-RMS multimeter in series for the average DC draw, and use an oscilloscope with a current probe (like a Tektronix TCP0030A) clamped around the input wire to view the peak-to-peak switching ripple.
What happens if my input voltage sags under load?
If your 12V battery sags to 10V under the 1.39A load, the converter will automatically draw more current to maintain the 15W output (15W / (10V × 0.90) = 1.66A). This can trigger a cascading failure where the battery sags further, current increases further, and the converter's internal UVLO (Under Voltage Lock Out) eventually trips, shutting the system down.
