Choosing between a linear regulator vs buck converter is not a question of which is better but of what you are willing to pay. A linear regulator conducts the full load current while dropping the difference between input and output, so every volt of that difference becomes heat. A buck converter moves energy rather than throwing it away, and charges you in complexity and switching noise.
The arithmetic that decides the linear case
In a linear regulator the current entering the input is essentially the current leaving the output plus a small quiescent term. Efficiency therefore reduces to the ratio of output voltage to input voltage, and nothing you do to the layout, the capacitors or the load changes that. Dropping a rail to roughly half its input voltage means about half the power drawn is dissipated in the regulator. Dropping a larger ratio makes it worse quickly. The dissipation is the product of the voltage difference and the load current, and it appears in one small package that must get rid of it through its leads, its thermal pad and the board copper beneath it. This is why linear regulators are excellent at small currents and small ratios and become impractical outside that region.
Why a switching converter escapes the same arithmetic
A buck converter does not conduct continuously from input to output. It switches a device on and off, storing energy in an inductor during one part of the cycle and releasing it during the other, so the average output voltage is a fraction of the input while the average input current is a correspondingly smaller fraction of the output current. Power in approximately equals power out, less the real losses: conduction in the switches and the inductor winding, switching losses in the transistors, core loss in the inductor, and losses in the control circuitry. Those losses do not scale with the voltage ratio in the same punishing way, which is why a converter can step a large ratio down at high efficiency. The efficiency is not constant, though, and typically falls at very light load and at very high current.
What the switcher costs you
Switching creates fast current and voltage edges, and those edges radiate and conduct interference. A converter needs careful layout, particularly of the loops carrying discontinuous current, and it produces output ripple at the switching frequency plus higher-frequency content from the edges themselves. Sensitive analog circuitry, precision references, radio receivers and high-resolution converters can all be degraded by that noise. The converter also needs an inductor, several capacitors and often a feedback network, taking more board area and adding parts that can fail. Its control loop can be unstable if the output capacitor is changed carelessly. A linear regulator by comparison is small, quiet, simple and forgiving, which is exactly why it remains the right choice in many places.
Using both, which is what most designs actually do
The productive way to frame the choice is to stop treating it as exclusive. A typical system uses a buck converter to bring a high input rail down efficiently to an intermediate voltage, then linear regulators with modest headroom to derive the quiet rails for analog and radio sections. The converter handles the bulk energy conversion where efficiency matters, and the linear stage provides the supply rejection where noise matters. Sizing the intermediate rail is the design decision: enough headroom to stay out of dropout under the worst load and temperature, and no more, because every extra volt multiplies by the load current and becomes heat again. Measure the resulting case temperature rather than trusting a calculation.
Heat is a design constraint, not an afterthought
When a linear regulator is the right answer, the dissipation still has to leave the package. Thermal resistance figures in a datasheet are measured on a specified test board, and your board is not that board, so treat them as comparative rather than predictive. Copper area, the number of thermal vias, adjacent heat sources, enclosure ventilation and ambient temperature all change the result. Junction temperature drives long-term reliability, so a part that runs hot but does not shut down is still consuming life. Verify with a thermocouple or thermal camera on a fully assembled unit at maximum load and maximum ambient, and confirm any thermal shutdown feature is a protection of last resort rather than something the design relies on.
What to check before you act
- Multiply the voltage difference by the load current to get linear dissipation.
- Check converter efficiency at your light-load condition, not only at full load.
- Measure the real case temperature on an assembled unit at maximum ambient.
- Consider a converter feeding a linear stage rather than choosing one alone.
Common questions
Is a low-dropout regulator more efficient than an ordinary one?
Only because it allows a smaller input-to-output difference. At the same voltage difference and current, the dissipation is the same. Efficiency follows the ratio either way.
Why is my buck converter inefficient at very light load?
Fixed losses such as switching, gate drive and controller quiescent current become a large fraction of a small output power. Many converters offer a light-load mode that reduces switching activity.
The practical takeaway
Pick a linear regulator when the voltage ratio is small, the current is modest and the noise requirement is strict, and a switching converter when the ratio or the current makes the heat unmanageable. Most good designs use both, with the converter setting up a rail just far enough above the linear stage to keep it out of dropout.
References and further reading
- Department of Energy: how electronics losses turn into wasted energy and heat
- ENERGY STAR: what a product must demonstrate to earn certification
- Department of Energy: putting a running cost on the electronics you already own
Numerical scenarios are illustrative unless identified otherwise. Follow the exact product instructions; component ratings and local installation requirements can differ.
