LESSON 09 · AMPLIFIERS · CIRCUIT TECHNOLOGY
Does Class A, Class AB or Class D decide how an amplifier sounds?
Learn what Class A, AB, D, G and H actually do, why people might reasonably prefer one over another, and which sound-quality claims need real evidence.
Start with the exact rating. Then compare the alternatives.
WHAT THE EXAMPLE STAT ACTUALLY MEANS
Class AB · MOSFET output stage
Illustrative, not a verified product measurement.
Read every part of the label
Class AB describes the operating method: complementary output devices share the audio signal with some overlap. MOSFET describes the kind of transistor used. Both terms can describe one amplifier, because operating class and transistor type are independent design choices.
What it doesn't establish: The label doesn't reveal the amplifier's wattage, distortion, output impedance or noise. It isn't enough to predict whether the amplifier will sound warm, clinical or transparent.
SEE THE DIFFERENCE
Three ways to make a speaker move
Inside an amplifier, transistors are tiny electronic controls that manage the electricity sent to your speakers. Think of them as very fast, precise switches or adjustable taps. Class A and AB control how much current passes; Class D switches fully on and off extremely quickly. Choose a class below to see how that works. The same example audio wave appears underneath each picture so you can compare what changes inside the amplifier. These simplified diagrams are not measurements.
Class A
The transistors stay on, with current changing as the music changes
In this simplified push-pull example, the two output transistors both carry some current throughout the entire music wave. The blue and green lines move up and down as the music changes, but neither line reaches the off position. Other Class A circuit layouts exist. That avoids a handoff between them, but electricity is used and heat is produced even when the music is quiet.
Why people choose it: continuous conduction avoids the particular output-device handoff associated with crossover distortion. Some listeners value that design philosophy, along with the experience of owning a traditional high-bias amplifier.
Why others do not: substantial idle heat, electricity use, large heat sinks and often less practical power per dollar. A Class A label alone does not prove warmer or more accurate sound.
Some amplifiers marketed with a Class A operating mode leave that mode at higher power or under particular loads. Check the actual specification and bias conditions.
Class AB
Two transistors take turns, with a small overlap
Unlike Class A, each transistor here handles mostly one half of the musical signal. The blue and green lines fall to the off position during their resting portion. Near each handoff, both briefly carry current to avoid a gap in the sound. This wastes less energy than many Class A designs while still moving the speaker the same way.
Why people choose it: a familiar linear topology balancing crossover control with better efficiency and more usable output than many pure Class A designs. A well-biased Class AB amp can be extremely clean.
Why others do not: it still wastes heat, and poor bias control or thermal design can leave crossover-related artifacts. Properly designed AB does not need to sound harsher than Class A.
Bias and measured distortion matter more than the AB badge. Some AB stages remain Class A-like for small signals, but that region depends on the actual circuit, load and bias.
Class D
Transistors switch on and off extremely quickly
Instead of acting like adjustable taps, the output transistors switch rapidly between on and off. In the common pulse-width method shown here, the time spent on versus off changes with the music. A filter (a circuit that removes the unwanted high-frequency switching) helps leave the intended audio signal for the speaker. Not every Class D amplifier uses exactly this switching method. It can be very efficient and accurate. The D does not stand for digital audio.
Why people choose it: high efficiency, small size, less heat, and potentially strong output from a compact or multichannel amplifier. Good modern examples can be audibly transparent.
Why others do not: the switching design introduces filter, electromagnetic-interference and feedback challenges that must be handled properly. Some designs interact with the load; some listeners simply prefer the construction or serviceability of a linear amplifier.
Class D does not mean digital sound. Never assume it sounds cold, sharp, or inferior without controlled evidence about the specific unit.
THE EVERYDAY EXPLANATION
Operating class and device type are separate choices
An amplifier's class describes how its output circuitry handles electrical power as the music changes. In Class A, the output devices stay active through the whole signal cycle. Class B divides the work between devices, and Class AB gives them a little overlap to smooth the handoff. Class D switches its output devices rapidly and filters the result. Those methods affect heat and efficiency; the class name alone does not tell you what you will hear.
The parts inside are a separate question. Vacuum tubes and transistors control the electrical current. BJT and MOSFET are two families of transistor, not amplifier classes. For example, an amplifier can be Class AB and use MOSFETs. A tube amplifier can also operate in Class A or AB.
Classes G and H manage the supply-rail voltages to reduce waste. They may be combined with linear output stages. Practical concerns such as cooling, reliability, output limits and power consumption often differ by design.
THE REASONS BEHIND THE PREFERENCE
Why would somebody actually choose Class A, AB, D, G or H?
There are legitimate trade-offs here. What changes most reliably is how the amplifier uses electricity, generates heat and delivers power. What you hear depends on the complete circuit, its measured behavior into your speakers, and the listening conditions. Your preference is real as an experience; the class letter alone does not explain its cause.
CLASS A · CONTINUOUS CONDUCTION
Choose it for its operating philosophy
The appeal: output devices remain conducting for the entire signal cycle. Avoiding the usual complementary handoff removes one possible crossover-distortion mechanism. Enthusiasts may like the engineering approach, a particular amplifier's measured performance or simply the equipment itself.
The compromise: inefficient at ordinary listening levels, substantial heat even at idle, more ventilation and generally a less convenient cost-to-power ratio. It does not follow that every Class A amplifier sounds warm, rich or better.
CLASS AB · A PRACTICAL MIDDLE GROUND
Choose it for clean linear power without Class A's constant losses
The appeal: complementary output devices overlap around the crossover region, reducing the dead zone of Class B while dissipating much less idle power than many Class A implementations. It can offer excellent fidelity and sensible size, heat and cost.
The compromise: bias and thermal tracking must be right, and efficiency still falls well short of many Class D designs. AB is not inherently coarse or gritty; bad crossover behavior is a design problem, not the required 'AB sound.'
CLASS D · SWITCHING EFFICIENCY
Choose it for compact, cool-running high output
The appeal: high efficiency means less waste heat and more practical output in compact amplifiers, powered speakers, subwoofers and multichannel systems. Well-executed modern Class D designs can measure extremely well.
The compromise: switching noise, output filters, load-dependent response in some designs, EMI and control-loop behavior must be engineered carefully. Those are reasons to examine the individual model, not proof that the entire class has a cold or digital tone.
CLASS G / H · SMARTER POWER-RAIL USE
Choose it when output demand varies widely
The appeal: multiple or tracking supply rails can reduce wasted energy in an amplifier that would otherwise operate with unnecessarily high supply voltage. That can help with heat and power delivery in suitable designs.
The compromise: rail-control transitions, power-supply behavior, protection and implementation complexity deserve scrutiny. G and H primarily describe rail management, and can coexist with other output-stage classes. They do not promise a sonic personality.
What about 'I prefer the way this one sounds'?
That's a valid listening preference. A particular tube or solid-state amplifier may have audible differences from another because of its distortion spectrum, output impedance, noise, frequency response, clipping behavior or interaction with speakers. Some listeners even prefer a specific, mildly colored presentation. Preference does not require the least distortion, but the class label does not prove that coloration exists or caused the preference.
For a useful comparison, match playback levels, keep the same source, speakers and placement, operate both amplifiers safely within their limits, and compare actual measurements where available. If a difference survives a controlled listening test, investigate its source rather than attributing it automatically to 'Class A warmth' or 'Class D harshness.'
An important hybrid case: Some nominally Class AB amplifiers bias their output devices so low-power operation remains in a Class A region, then transition as current demand rises. The size of that region is model- and load-dependent. A front-panel Class A switch or marketing label is not enough to establish its actual power envelope or the audibility of the transition.
WHAT MIGHT YOU HEAR?
Connect the measurements to listening
A carefully engineered Class D amplifier can reproduce music very accurately, just as a well-made Class A or AB amplifier can. The class label alone does not prove that one will have smoother guitars or punchier drums. Why might a listener prefer one? They might hear a real model-specific difference, like a particular distortion pattern or speaker interaction, or value its physical design, power delivery, efficiency and ownership experience. Those explanations should be investigated separately.
Some tube designs or other deliberately colored circuits can add audible harmonics. That coloration depends on the particular circuitry and load, not merely the presence of tubes, MOSFETs or a Class A badge.
WHAT AUDIOPHILES SAY
“I love the warmth of my Class A amp,” “Class AB sounds more natural,” “MOSFETs sound like tubes” and “Class D sounds cold” describe listener impressions worth investigating, not universal laws. A preference is meaningful; its cause remains a separate question.
WHAT THE EVIDENCE TELLS US
Amplifier operating class is directly related to conduction or switching behavior, efficiency and thermal design. It does not, by itself, establish output-stage linearity under real loads or a unique audible tonal character. Compare distortion-versus-power, noise, response into the speaker load, and level-matched listening results.
USE THIS WHEN BUYING
What matters in practice?
- Compare clean power, THD+N at relevant loads and levels, and compatibility with the speakers you actually own.
- Account for idle heat, ventilation and service needs; tube replacement and Class A power consumption can add ongoing costs.
- Read actual product measurements rather than relying on the output-device label as a substitute for sound-quality evidence. Compare THD+N over power, distortion spectra when available, frequency response with your load and clipping behavior.
Other important variables
Bias and thermal stability, output-stage feedback, supply-rail behavior, Class D output-filter behavior, noise, output power, speaker load and the measured operating point at your typical listening level. A different circuit can measure better at 2 W yet worse at 80 W.
Engineering deep dive Optional
In a conventional linear output stage, a Class A device conducts for 360° of the cycle; Class B approximately 180°; Class AB between 180° and 360°. Class D uses switching rather than the same conduction-angle classification. Class G/H implement different supply-rail management strategies.
Sources & further reading
All diagrams are original schematic teaching illustrations.