Class A Class D Amplifier Family Map
There are three output-stage families in a power amplifier: Class A, Class AB, and Class D. One physical axis separates them: whether the output transistor works in its smooth linear region or is switched hard on and hard off. Once you know which side a design sits on, the rest of its behaviour follows.
Key Takeaways
- The axis is drive state, not a label. Whether the transistor runs in its linear region or gets pulsed fully on and off is the one fact that separates the three families and explains their heat, headroom, and listening character.
- Class A trades heat for smoothness. A constant bias current keeps the device in its linear zone across the full listening range, giving near-zero crossover distortion but turning roughly 75 percent of input power into waste heat.
- Class D trades a little purity for raw output. Pulse-width modulation switches the transistor hard on or hard off, cutting thermal load to about 20 percent and freeing up wattage headroom for large subwoofers or small cabinets.
- Class AB straddles the line. At low volume it mimics Class A; at higher levels the devices hand off and it begins to switch. Where it lands depends on the bias point the designer chose.
The Big Picture First
Every power amplifier you buy answers the same physical question with a different mechanism: how does the output transistor deliver current to the speaker? Listeners keep meeting this fork because the marketing labels (“solid-state,” “tube-like warmth,” “pure audio”) describe construction choices or adjectives, not the actual drive method. Think of a gas pedal on a car. In one design, you ease the pedal smoothly and the engine revs in tiny, linear steps. In another design, you stomp the pedal fully down or lift it fully off, and a filter smooths the on-off pulses into a usable average speed. The amplifier families are built around that same choice: smooth, continuous conduction versus hard on-off switching. Everything else, such as heat, headroom, the way a bass note decays, and the way a drum hit attacks, flows from where a design lands on that single axis.
The Whole Family At A Glance
| Family | One-Line Working Principle | Signature Sound Trait | Where You Typically Meet It |
|---|---|---|---|
| Class A | A constant bias current keeps the output device in its linear region across the full volume range | Smooth, low crossover distortion; a continuous bass that does not snap off | Small-room desktop amps; high-impedance planar speakers; boutique solid-state circuits |
| Class AB | The device runs in its linear zone at low level, then the two push-pull halves hand off and begin to switch at higher drive | Clean at low volume with an efficiency gain as volume rises; less heat than Class A | Most mid-priced stereo receivers; integrated amplifiers for general home listening |
| Class D | Pulse-width modulation switches the output stage hard on and hard off at a very high rate; a low-pass filter reconstructs the average | Tight, controlled bass with high headroom in a small enclosure; very low heat | Subwoofer amplifiers; compact multi-channel AV receivers; portable powered speakers |
The Families, One By One
Class A: The Continuous Conductor
In a Class A stage, the output transistor sits with a steady bias current flowing through it even when no signal is present. The device does not fully cut off. It conducts a little, a lot, or somewhere in between, but it stays in the smooth, proportional part of its transfer curve throughout the listening range. Because the transistor does not cross the off-to-on threshold, there is no crossover region where one device hands the signal to its partner. The result is very low distortion right at the zero-crossing point. The cost is heat. Because the device carries current even in silence, roughly 75 percent of the electrical power drawn from the wall ends up as waste heat in the heatsink. That is why Class A stages are usually rated at modest wattage and are found in small-room or high-impedance speaker applications where the thermal load stays manageable. Listening impression: bass notes start and stop with a smooth, almost liquid continuity. There is no audible click when a bass tone fades toward silence. The whole performance feels slightly rounded and continuous, which is why many listeners associate Class A with a warm character even when the circuit is entirely solid-state.
Class AB: The Crossover Mover
Class AB borrows the push-pull topology of Class B but adds a small quiescent bias so both output devices conduct for a brief window around the zero-crossing point. At low volume, both devices are on, and the stage behaves much like Class A: smooth and low in distortion. As the signal grows louder, the two devices take turns carrying more of the load, and the stage starts to behave more like a switching arrangement. The efficiency climbs because you are no longer paying the full Class A idle current at high output. The amount of quiescent bias the designer sets is the key variable. More bias pushes the hand-off point further out, keeping the stage in its linear zone for a wider volume range. Less bias lets it switch sooner and saves heat. A hybrid amplifier that leans toward one family or the other is really a matter of where that bias point was set during the design. Listening impression: at the volume you would use for background music in a living room, the difference from Class A is subtle. Push the volume to party levels and you may notice a slight shift in how bass transients attack. The bass gets a touch tighter and more edged, because the stage has moved into the switching side of the axis.
Class D: The Switching Workhorse
A Class D stage does not ask the transistor to sit in its linear region at all. Instead, a pulse-width modulation (PWM) circuit chops the output signal into very short on-off pulses. The duty cycle, which is the fraction of time the pulse is on versus off, encodes the amplitude of the desired signal. A low-pass filter after the switching stage smooths those pulses back into a usable waveform for the speaker. Because the transistor is either fully on, carrying current with near-zero voltage across it, or fully off, carrying no current, it dissipates very little power as heat. Thermal load drops to about 20 percent of what a Class A stage would waste. The freed-up thermal budget means the amplifier can deliver high wattage into small enclosures without overheating. Listening impression: bass is tight, controlled, and fast. The amplifier can drive a large subwoofer to high excursion without the voice-coil overheating, because the amp itself stays cool. In a well-designed Class D stage, the reconstruction filter keeps any switching-frequency noise well below normal listening thresholds, so the high end stays clean.
The Difference That Counts
The physical fact that separates the families is simple. Does the output transistor conduct in its proportional, linear region, or does it sit at one of two extremes, fully on or fully off? Everything else, heat, headroom, the feel of a bass note’s decay, flows from that one choice. Think of a water pipe with a valve. In a Class A stage, you turn the valve to any angle between wide open and fully shut, and the flow rate tracks the angle smoothly. In a Class D stage, you only have two positions: wide open or fully shut, and you click between them many times per second. A flow meter downstream averages those clicks into a steady reading. The “Class AB” label sits between. The valve spends most of its time at a gentle partial opening, but clicks fully open or shut when the flow demand gets high. Surface labels like “pure audio” or “high-resolution” tell you about marketing, not about where the valve actually sits. Two amplifiers can carry the same adjectives yet land on opposite sides of the axis, which is why the drive-state question is the one that actually settles your choice at the listening level you care about.
Where The Lines Blur
A few real-world designs mix characteristics from more than one family, and the blur makes the labels confusing. Here is how to sort them using the axis.
High-bias Class AB pushed into near-Class-A territory. Some boutique amplifiers set the quiescent current so high that both output devices stay in their linear region across nearly the entire usable output range. The stage conducts like Class A for almost every note you would play at normal listening levels. It still has the hand-off structure of Class AB, but the bias point pushes that hand-off so far out that it rarely happens in practice. Ask yourself: does the device actually cross into its switching region at the volume I will use? If the answer is essentially no, treat it as a Class A in listening behaviour even though the circuit diagram says AB.
Class D with a small linear output buffer. Some high-end subwoofer amplifiers use a Class D front end for the bulk of the power but add a small Class A output buffer on top. The Class D section handles the heavy lifting at high excursion. The Class A buffer smooths the last few millimetres of cone travel. The family you hear depends on how loud you play. At low and mid levels the linear buffer dominates the sound. At full throttle the switching section takes over.
Gate-driven ultra-linear stages. A handful of designs use a gate-driven MOSFET arrangement that keeps the transistor in its linear region while running at higher efficiency through a clever bias network. These sit closer to Class A on the axis than to Class D, even though the topology looks unusual on a schematic. The test is the same in every case: is the device conducting proportionally to the input signal, or is it being pulsed between two extremes?
In each of these hybrids, the question that resolves the blur is the axis from the previous section. Where does the output device actually sit in its transfer curve at the listening level you care about? That answer tells you which family it truly leans toward.
Questions About The Types
How Do I Tell Which Family An Amplifier Uses Just By Looking At It?
You usually cannot tell from the outside. Two cabinets can look identical, carry the same wattage number, and share the same marketing adjectives. The drive-state axis is internal. Look for clues: a very large heatsink with a modest wattage rating points toward Class A or high-bias AB. A small, light enclosure with a high wattage rating and low idle power draw points toward Class D. The most reliable method is to read the circuit description in the user manual or the manufacturer’s technical notes, which will state the output-stage topology in plain terms.
Does A Higher Wattage Number Mean A Better Amplifier?
No. Wattage on the box measures how much power the amplifier can deliver under lab conditions, not how clean or smooth it sounds at the volume you actually listen at. A 200-watt Class D subwoofer amplifier and a 20-watt Class A power stage can both sound excellent in their intended roles. The wattage number tells you headroom and thermal design, not the drive-state axis. A small Class A stage driving a high-impedance speaker in a quiet room can deliver more musical detail at listening level than a large Class D stage that is far more efficient but not optimised for that impedance load.
Why Does Class A Sound “Warmer” Even With All-Solid-State Parts?
The warmth comes from the constant bias current. Because the transistor is conducting across the full signal swing, the current path is continuous and the output waveform does not pass through a dead zone where both devices are off. That absence of a crossover gap means the low-order harmonics of the signal are generated by a single, steady conduction path rather than by two devices taking turns. The result is a smooth, slightly rounded character in the mid-bass and bass that many listeners describe as warm or tube-like, even though no vacuum tube is involved.
Can I Mix Families In One System, Like A Class A Front And A Class D Back?
Yes, and many multi-channel AV receivers do exactly this. The front stereo channels may use a Class AB output stage for general programme material, while the dedicated subwoofer channel uses a Class D stage for high-exursion bass. Each channel is an independent amplifier with its own drive-state topology. The listening result is a blend: the stereo image and mid-range follow the AB behaviour, while the sub-bass follows the D behaviour. Each channel can be matched to the speaker and role it is assigned, so one family does not need to win across the whole system.
(This article first appeared on hifireport.com. Reproduction without permission is not allowed.)