Amplifier Classes: A, AB and D Explained

Read any amplifier review and you will meet a letter: Class A, Class AB, Class D. The letter describes one specific thing — how the output devices behave across the audio waveform — and it has real consequences for efficiency, heat, size and cost. What it does not do is settle how an amplifier sounds. Excellent and mediocre amplifiers exist in every class, and the gap between two implementations of the same class is routinely larger than the gap between classes. This guide explains what each letter actually means, then draws the line between what the class predicts and what it does not.

What "Class" Is Describing

An amplifier's output stage has to reproduce a signal that swings above and below zero. The classification describes how much of that waveform each output device conducts for — the conduction angle — and, in the case of Class D, whether the devices are behaving as linear amplifiers at all.

Think of it as a spectrum with a trade-off at each end. Keeping output devices conducting all the time is linear and free of a particular kind of distortion, but wastes enormous energy as heat. Switching them off when they are not needed is far more efficient, but introduces the problem of handing over cleanly between devices. Every class is a different answer to that trade.

Class G and Class H: Rail Switching

Two further classes turn up mainly in high-power amplifiers, and both are variations on the same idea: instead of running the output stage from one fixed high-voltage supply all the time, vary the supply to suit the signal.

Class G uses two or more fixed supply rails and switches to the higher one only when a loud passage demands it. Class H modulates the rail voltage continuously so it tracks just above the signal. Both keep a conventional linear (usually Class AB) output stage, so their behaviour resembles Class AB, but they waste far less power as heat during quiet passages. The result is high peak power without a huge heatsink — useful in professional amplifiers and some high-output home designs.

What the Class Actually Predicts

The class reliably tells you about the physical and practical characteristics of an amplifier: how hot it runs, how much it weighs, how much power it can package into a given size, and roughly how much it costs per watt. Those are genuinely useful things to know when you are choosing a component for a real room and a real shelf.

What the class does not reliably predict is sound quality. A well-executed Class D amplifier can measure superbly and drive difficult loads with authority. A poorly-executed Class A amplifier with an inadequate power supply can sound worse than a competent Class AB at a third of the price. Modern Class D in particular has moved on enormously from its early reputation — output filter design, feedback around the filter, and switching-frequency choices have improved to the point where the best implementations are among the most transparent amplifiers available at any price.

Where Each Class Makes Practical Sense

Because the classes differ mostly in physical consequences, matching them to circumstances is straightforward.

• Class A suits low-power systems with sensitive speakers, in rooms where the heat is tolerable and running costs are not a concern. It is the least practical and the most romantic option.
• Class AB suits almost everything. It scales from modest integrateds to serious power amplifiers, delivers plenty of current for difficult loads, and manages heat without extreme measures.
• Class D suits high power in a small space, active speakers, subwoofers, multi-channel systems, and anywhere efficiency or heat is a constraint. It is also the sensible choice if the amplifier will be left switched on continuously.
• Class G and H suit high-output applications where peak capability matters more than average power — large rooms, insensitive speakers, and professional use.

One consequence worth planning for: a Class A amplifier needs genuine ventilation. Placing one in a closed cabinet is not a cosmetic decision but a reliability one, and the same caution applies at a lower level to high-bias Class AB designs.

Bias, Idle Current and "First Watt" Class A

A detail that causes confusion: many Class AB amplifiers are described as operating "in Class A for the first few watts". This is literally true and follows from how bias works. Below the current level set by the bias, both output devices are still conducting, so the amplifier is genuinely in Class A operation; above it, one device hands over to the other and it becomes Class AB.

Whether that matters depends on your listening level. With sensitive speakers at moderate volume, average power demand can genuinely sit within the first watt for much of the time — which is the argument behind high-bias designs. With insensitive speakers in a large room, the amplifier spends most of its time well outside that region and the distinction is academic. As so often, the answer depends on the speakers and the room rather than on the amplifier alone.

Efficiency at the Levels You Actually Use

The efficiency figures quoted for each class are measured at or near full output, and very little listening happens there. At normal levels with speakers of average sensitivity, the average power demand is commonly around a watt or less, with brief peaks far higher. At that operating point, what the amplifier wastes is dominated by its idle draw rather than its full-power efficiency.

The theory makes the point clearly. A pure Class B output stage, the most efficient linear arrangement, has a theoretical maximum efficiency of about 78.5% at full output, but that maximum falls in proportion to output voltage. At one-hundredth of full power — 1 W from a 100 W amplifier — the theoretical ceiling drops below 8%, before bias current and the rest of the circuit are counted. Class A is worse still, since it draws its full current regardless. Class D is the one class whose efficiency holds up reasonably well at low output, which is a large part of why it suits equipment that is left on.

Heat and Running Cost in Practice

Idle consumption turns these differences into numbers on an electricity bill. A stereo Class A amplifier rated at 25 W per channel might draw around 225 W continuously whether playing or not. Used four hours a day, that comes to about 330 kWh a year, multiplied by your local tariff. A Class AB integrated of 100 W per channel might idle at 40–80 W, and many Class D designs idle at 10–30 W, or less in standby. These are illustrative figures rather than specifications, so check the power consumption given in a particular model's manual or measured in reviews.

The heat side follows directly: almost all of a Class A amplifier's draw ends up warming the room. In a small room in summer, a couple of hundred watts of continuous heat is comparable to having an extra person or two in the room, which is a practical consideration as well as an environmental one.

Class D: What Changed

Class D's early reputation for a hard or variable treble came largely from one source: the output filter. The filter that removes the switching frequency is tuned assuming a particular load, and in early designs without feedback after the filter, the frequency response in the top octave changed with the speaker's impedance — rising into some loads, falling into others. Paired with a speaker whose impedance varied at high frequencies, the treble balance became unpredictable.

Modern designs address this directly, typically by taking feedback from after the output filter or by using self-oscillating topologies that include the filter in the control loop. Switching frequencies of several hundred kilohertz to over a megahertz put the residual far above the audio band. The best current Class D modules measure with distortion and noise comparable to the finest linear amplifiers, and their response is largely independent of the load. One residual quirk is that a small amount of switching-frequency energy remains at the output; it is inaudible, but it is why measuring a Class D amplifier properly requires a specific filter on the test equipment.

The Same Letters in Valve Amplifiers

The classes are not specific to transistors. A single-ended valve amplifier, with one output valve per channel conducting continuously, is always Class A, which is why such designs are typically low-powered and run hot. Push-pull valve amplifiers use pairs of output valves and are commonly biased in Class AB, gaining power and efficiency in exactly the same way as their solid-state counterparts. When a valve amplifier is described as Class A, it tells you the same thing it does for a transistor amplifier: low power for its size and heat, and no crossover region.

Quick Reference: Classes Compared

Class	Operation	Efficiency	Typical Use
Class A	Devices conduct for the full waveform; no crossover region	20–30%	Low-power designs with sensitive speakers; runs very hot
Class AB	Slight overlap around zero crossing, set by bias current	50–65%	Mainstream integrateds and power amplifiers
Class D	Devices switch fully on/off at high frequency; output filtered	85–95%	Subwoofers, active speakers, compact and high-power designs
Class G	Linear output stage with stepped supply rails	Better than AB	High-power amplifiers needing peak headroom
Class H	Linear output stage with continuously tracked rail voltage	Better than AB	Professional and high-output home amplifiers

More Guides