What Is a Linear Power Supply?
A linear power supply is a type of DC power supply that converts AC mains electricity into a stable, low-noise DC output using a step-down transformer and a linear voltage regulator — a transistor that operates in its active region, behaving like a continuously adjustable resistor to maintain a precise output voltage.
Think of it as a meticulous craftsman: it wastes more energy than it needs to, but the result is exceptionally clean. Linear power supplies typically achieve efficiency between 40% and 60% — meaning nearly half the input power can be converted to heat rather than useful output (Acopian). In exchange, they produce output ripple as low as 0.25 mV RMS — 20 to 50 times quieter than a typical switching supply.
They are heavy, they run hot, and their design has barely changed in decades. Yet in laboratories, recording studios, hospitals, and anywhere that demands absolutely clean DC power, linear supplies remain irreplaceable. To understand why, you need to see how they work.
How Does a Linear Power Supply Work?
Every linear power supply follows the same four-stage signal path: AC mains enters, gets stepped down, gets rectified and filtered into rough DC, and finally gets polished by a linear regulator into clean, stable output. That journey runs through four stages.
Step 1 — Transformer: Stepping Down the Voltage
The first thing the AC mains encounters is a large iron-core transformer. Its job is straightforward: take the 110V or 220V from the wall and step it down to a much lower AC voltage — typically somewhere between 12V and 24V AC on the secondary side. It also provides galvanic isolation, separating your circuit from the mains for safety.
This transformer is the single biggest reason linear power supplies are so large and heavy. It operates at the mains frequency — 50 or 60 Hz — and at these low frequencies, an efficient transformer needs a substantial iron core and many turns of copper wire. A 100-watt linear supply’s transformer alone can weigh 1.5 to 2.5 kilograms.
Compare that to a switching power supply, where the transformer runs at 50 kHz or higher: the core shrinks to a fraction of the size because the required magnetic cross-section decreases as frequency increases. But in a linear design, there is no getting around the bulk — it is the price of simplicity.
Step 2 & 3 — Rectification and Filtering: From AC to Rough DC
After the transformer delivers low-voltage AC, a bridge rectifier — four diodes arranged in a diamond configuration — flips the negative half of each AC cycle upward, turning the alternating sine wave into a series of positive-going humps. Engineers call this “pulsating DC.”
It is a start, but not a usable one. The voltage still swings from zero to peak 100 or 120 times per second (twice the mains frequency, since both halves of the cycle now point the same direction). To smooth this out, a large electrolytic capacitor — typically several thousand microfarads, say 4,700 µF or 10,000 µF — sits across the rectifier output. This reservoir capacitor charges to the peak voltage during each hump and discharges into the load during the gaps between them, flattening the waveform into something that resembles DC, but still carries ripple.
A useful mental model: the capacitor acts like a water reservoir. Upstream, the rectifier delivers water in pulses. Downstream, the load draws continuously. The reservoir absorbs the surges and releases a steady flow — but the surface still ripples.
Step 4 — Voltage Regulation: The Heart of Linear Design
This is where linear power supplies earn their name. The roughly-filtered DC now enters a series pass transistor — the regulating element — which operates in its linear or active region. Unlike a switch that is either fully on or fully off, this transistor is always partially on, behaving like a variable resistor.
A feedback circuit continuously monitors the output voltage. It compares the actual output against a precision reference voltage (typically accurate to within ±2% to ±5%) and dynamically adjusts the transistor’s resistance: if the output drifts too high, the transistor increases its resistance to drop more voltage across itself; if too low, it eases off. The result is a rock-steady output regardless of changes in input voltage or load current.
The trade-off is captured in a single equation:
P_dissipated = (V_in − V_out) × I_load
Take a real example: a linear regulator receiving 18V at its input, delivering 12V at 2A to the load. The voltage drop across the transistor is 6V, and with 2A flowing through it, that is 12 watts of heat — pure waste, conducted away by a heatsink. This is why linear supplies run hot and why their efficiency is fundamentally limited: the larger the gap between input and output voltage, the more energy becomes heat.
Classic integrated linear regulators like the LM317 (adjustable) and the 78xx fixed-voltage series (7805 for 5V, 7812 for 12V) package this entire feedback-and-regulation circuit into a single three-pin chip. They have been workhorses of electronics design for over four decades — simple, reliable, and nearly impossible to kill if you give them adequate cooling.
Steps down AC voltage
AC rough DC
Precision voltage control
<1 mV ripple
Linear vs. Switching Power Supplies: Key Differences
If you have spent any time around power electronics, you have heard the debate: linear or switching? The comparison table below gives you the numbers at a glance. But numbers without context can mislead — so each dimension gets unpacked below.
| Boyut | Doğrusal Güç Kaynağı | Anahtarlamalı Güç Kaynağı (SMPS) |
|---|---|---|
| Verimlilik | 40–60% | 70–90%+ |
| Output Ripple (RMS) | 0.25–1.5 mV | 10–25 mV |
| Boyut ve Ağırlık | Large, heavy | Compact, lightweight |
| Circuit Complexity | Basit | Complex |
| EMI / RFI | Very low | High — requires shielding |
| Transient Recovery | 50–100 µs | ~300 µs |
| Yük Regülasyonu | 0.005–0.2% | 0.05–0.5% |
Source: Acopian Power Supply Comparison
Efficiency and Heat — The Fundamental Trade-off
The efficiency gap between linear and switching supplies is not a matter of engineering refinement — it is baked into their operating principles.
In a linear supply, the series pass transistor is always conducting, always dropping voltage, always dissipating power. Every watt not delivered to the load becomes heat. When the input-output voltage difference is large, efficiency plummets: a 5V output from an 18V input wastes over 70% of the input power as heat.
A switching supply avoids this entirely. Its transistor — typically a MOSFET — operates in only two states: fully on (where voltage across it is near zero) or fully off (where current through it is near zero). In both states, the product of voltage and current — power dissipation — is negligible. Losses occur only during the brief switching transitions, and at switching frequencies of 50 kHz to 1 MHz, modern MOSFETs with on-resistances in the single-digit milliohms keep those losses remarkably small. The result: a 100W switching supply might waste 10 to 15 watts as heat, while a linear equivalent wastes 50 to 60 watts.
The analogy that helps: a linear supply is like a driver who controls speed by riding the brake pedal — constant friction, constant waste. A switching supply is like a driver who rapidly taps the brake 100,000 times per second — the average braking force is minimal, and almost all the energy goes to forward motion.
Every watt of voltage drop becomes heat. This single equation explains why linear supplies run hot — and why heatsinks are not optional.
Noise and Ripple — Why Clean Power Matters
If efficiency is switching’s advantage, noise performance is linear’s fortress.
Linear supplies produce ripple almost entirely at the rectified mains frequency — 100 Hz or 120 Hz — and its low-order harmonics. This is low-frequency ripple, which is relatively easy to filter further downstream and does not radiate effectively. The result: output noise typically below 1 mV RMS.
Switching supplies, by contrast, generate broadband noise from the sharp edges of their switching waveforms. A MOSFET turning on and off in tens of nanoseconds produces frequency components stretching into the tens of megahertz. This noise couples through conduction (along the output wires) and radiation (through the air), and can corrupt sensitive analog signals in ways that are hard to diagnose and harder to fix after the fact.
Why does this matter in practice? Three concrete examples:
- Audio equipment: 50 mV of switching noise on a preamplifier’s power rail, after 40 dB of gain, becomes 5V of audible hiss and buzz at the speaker. Recording engineers can hear the difference — and so can their clients.
- Precision measurement: A 16-bit ADC operating with a 2.5V reference resolves signals as small as 38 microvolts per bit. Power supply ripple above this level directly translates to jitter in the least significant bits — your readings drift for no apparent reason.
- RF circuits: A receiver’s low-noise amplifier (LNA) amplifies everything on its input — including power supply noise that falls within its bandwidth. Switching harmonics can create phantom signals or raise the noise floor enough to bury weak stations.
Size, Weight, and Cost — Practical Realities
Walk into any electronics lab and the difference is immediately visible. A 100W linear bench supply sits on the desk like a small anvil — most of that weight is the transformer’s iron core and copper windings, plus the aluminum heatsink fins needed to dissipate tens of watts of waste heat. A 100W switching adapter for a laptop fits in the palm of your hand.
There is a cost crossover, too. Below roughly 50 watts, linear supplies remain price-competitive because their simplicity offsets the cost of the bulky transformer. Above 100 watts, switching supplies pull decisively ahead — the cost of copper and iron in a linear design grows faster than the cost of silicon and ferrite in a switching design. For most modern industrial and consumer applications above a few dozen watts, switching has won on economics alone.
Trade-off at a Glance
| Linear | Switching | |
|---|---|---|
Verimlilik higher is better |
40–60% Düşük |
80–95% Best |
Ripple lower is better |
<1 mV Best |
10–25 mV Daha yüksek |
Ağırlık lower is better |
Heavy 1.5–2.5 kg (100W) |
Light Compact, portable |
EMI Noise lower is better |
Çok Düşük Best |
Yüksek Needs shielding |
When Should You Choose a Linear Power Supply?
This is the question that most “linear vs. switching” articles never answer — and it is the one that actually matters once you understand the basics.
Three questions cut through the noise:
- Is your application noise-sensitive? If you are powering audio amplifiers, precision ADCs, sensitive RF receivers, or medical instrumentation, linear’s ultra-low ripple is not a luxury — it is a requirement. The cost and complexity of filtering a switching supply down to linear levels often exceeds the cost of just using a linear supply in the first place.
- Do you have strict size or weight constraints? If the supply must fit in a compact enclosure or be carried — portable test equipment, drone payloads, embedded field devices — switching is the practical choice. A linear supply’s transformer alone may exceed your entire weight budget.
- What is your power level? Below 50W, either topology works; let noise sensitivity decide. Above 500W, linear becomes impractical — the transformer and heatsink would be enormous, and the wasted energy would require active cooling. Between 50W and 500W, it is genuinely an engineering trade-off.
Run these questions against real scenarios:
- Recording studio / audiophile setup: Linear. No contest. The entire signal chain — microphone preamps, DACs, headphone amplifiers — lives or dies by power cleanliness.
- Laboratory bench testing: Linear dominates. When you are characterizing a circuit and need to know that the 2 mV deviation you see is real and not power supply artifact, clean power is non-negotiable.
- Industrial control cabinet (DIN rail): Switching dominates — efficiency and compact size matter more — but analog sensor front-ends within the same cabinet may still benefit from a local linear regulator stage.
- Medical diagnostic equipment: Mixed. The sensitive analog front-end (ECG, EEG, pulse oximetry) typically uses linear regulation; the digital processing and display sections run on switching supplies.
- Embedded / IoT prototyping: Switching is almost always sufficient. A modern low-dropout (LDO) linear regulator downstream of a switching pre-regulator can clean up the last few millivolts of noise if a specific analog sensor requires it.
There is a professional trick worth knowing: the hybrid approach. Use a switching supply as a pre-regulator to drop the voltage efficiently to just above the target, then follow it with a linear regulator to strip away the remaining ripple. This is how Tektronix built some of their legendary oscilloscopes, and it remains the architecture of choice for high-end bench power supplies today. As Analog Devices has demonstrated, modern switching regulators with proper post-filtering can now power 16-bit high-speed ADCs with no measurable degradation versus a pure linear supply — while cutting power consumption by roughly 35% (Analog Devices). The “linear vs. switching” debate is increasingly becoming “linear and switching, working together.”
What to Look for in a Linear Power Supply
If your application calls for a linear supply, knowing what separates a well-built unit from a corner-cutting one saves you from chasing problems later. Five dimensions matter.
1. Output specifications. Check the line regulation (how much the output changes when the input voltage varies) and load regulation (how much it changes when the load current changes). Quality linear supplies achieve line regulation below 0.01% per volt and load regulation below 0.1%. The ripple-and-noise figure on the datasheet — look for it in mV RMS, not just mV peak-to-peak — should be under 1 mV RMS for a serious instrument-grade supply. But read the test conditions: a ripple spec measured at 10% load can be 5 to 10 times worse at full load.
2. Protection features. Over-current protection (OCP), over-voltage protection (OVP), and thermal shutdown are not optional in industrial environments. A supply without them can destroy both itself and your circuit when something goes wrong — and something always goes wrong eventually.
3. Build quality. Open the case if you can. Look at the transformer — silicon steel laminations indicate a proper core; a suspiciously light unit may use inferior materials that vibrate and hum. Look at the filter capacitors — Japanese brands (Nichicon, Rubycon, Panasonic) consistently outlast generic alternatives. Look at the heatsink — is it generously sized, or does it look like an afterthought? A linear supply’s reliability is directly proportional to its cooling margin.
4. Certifications. CE marking, UL listing, and RoHS compliance are baselines, not differentiators — but their absence is a red flag. Safety certifications like IEC 62368-1 (replacing the older IEC 60950) and EMC compliance to EN 55032 Class B tell you the manufacturer has invested in testing and compliance, not just assembly.
5. Manufacturer reputation. A power supply is a long-term component. You want a manufacturer with a track record — someone who has been building power supplies long enough to understand the engineering trade-offs between different topologies, who stands behind their products with a real warranty, and who can provide technical support when you need it. A supplier who understands both linear and switching design can give you an honest recommendation rather than simply selling whatever they happen to make.
Making the Right Choice for Your Power Needs
If you take one idea away from this article, let it be this: a linear power supply is a trade — efficiency and compactness for purity and simplicity. The transistor that regulates its output is the same transistor that wastes its energy. That is not a design flaw; it is the physics of the topology.
For a quick decision heuristic: if your circuit cares more about clean power than about efficiency or size, linear is likely the right answer. If efficiency and power density dominate, switching is the default. And if you are not sure — or if your system has both analog and digital sections with competing requirements — the hybrid approach of a switching pre-regulator followed by linear post-regulation is not a compromise. It is the state of the art.
Choosing a power supply is not a religious debate. There is no universally correct topology — only the one that fits your specific application. Good engineers understand both, and then put the right tool in the right place.
Need Help Choosing the Right Power Solution?
Whether linear or switching, we’ll help you find the right fit for your application.
Talk to an EngineerReferences
- Acopian. “Power Supply Comparisons: Linear Power Supply vs. Switching Power Supply vs. Unregulated Power Supply.” https://www.acopian.com/linear-power-supply-vs-switching-power-supply-vs-unregulated-power-supply.html
- Analog Devices. “Powering High Speed Analog-to-Digital Converters with Switching Power Supplies.” https://www.analog.com/jp/resources/technical-articles/powering-high-speed-analog-to-digital-converters-with-switching-power-supplies.html
- Siglent NA. “What Is the Difference Between a Switching and Linear Power Supply?” https://siglentna.com/operating-tip/difference-switching-linear-power-supply/
- OMCH. “Switch Mode Power Supply.” https://www.omch.com/switch-mode-power-supply/
- OMCH. “About OMCH.” https://www.omch.com/about-omch/
- OMCH. Homepage. https://www.omch.com/



