An amplifier class defines the fraction of the input signal cycle during which the active output devices conduct current, dictating the fundamental trade-off between power efficiency and signal fidelity. Choosing a specific class changes your heat sink mass, power supply current requirements, printed circuit board (PCB) layout constraints, and the total harmonic distortion (THD) you will measure at the output.
The Core Trade-Off: Fidelity vs. Heat
Every active amplifier relies on transistors (BJTs, MOSFETs, or GaN FETs) acting as variable resistors or switches to modulate power from a supply rail to a load. The "class" simply describes how long those transistors stay turned on during a single 360-degree cycle of the input waveform.
Think of it like a toll booth operator. A Class A operator never leaves the booth, ready to process cars instantly (zero crossover distortion, but they get exhausted and overheat even when no cars are there). A Class B setup uses two operators taking exact 50/50 shifts, but there is a slight delay when they hand off the shift to each other (crossover distortion).
The Major Amplifier Classes at a Glance
Before you select output devices or design a power supply, you need to know the theoretical limits of your chosen class. Here is how the standard classes stack up on the bench.
| Class | Conduction Angle | Theoretical Max Efficiency | Typical Real-World Efficiency | Primary Use Case |
|---|---|---|---|---|
| A | 360° (Full cycle) | 50% (25% typical) | 15% - 30% | Boutique hi-fi audio, low-noise RF preamps |
| B | 180° (Half cycle) | 78.5% | Rarely used alone due to high crossover distortion | RF push-pull stages |
| AB | >180° to <360° | 78.5% | 50% - 65% | AV receivers, guitar amps, studio monitors |
| D | Switching (PWM) | 100% | 85% - 95% | Car audio, portable speakers, active subwoofers |
| G / H | Variable rail AB | ~80% | 65% - 80% | Pro audio PA amplifiers, high-power home theater |
Bench Math: Sizing Heat Sinks for Class A vs. Class AB
Let us move past theory and look at what this means when you are bolting TO-3P transistors to a heat sink. Suppose we are designing a 50W RMS audio amplifier driving an 8-ohm load, and we need to size the thermal management.
The Class AB Scenario:
Assuming a realistic 60% efficiency, the total power drawn from the supply is 50W / 0.60 = 83.3W. The heat dissipated by the output transistors is 83.3W - 50W = 33.3W. A standard extruded aluminum heat sink (like a Wakefield Engineering 441 series) with a thermal resistance of roughly 1.5°C/W will keep the transistors well within their safe operating area (SOA) in free air.
The Class A Scenario:
Class A practical efficiency hovers around 25%. To deliver 50W to the load, the amplifier must draw 50W / 0.25 = 200W from the supply. The heat dissipated is 200W - 50W = 150W.
To dissipate 150W while keeping the transistor junction temperature under 100°C in a 25°C room, you need a heat sink with a thermal resistance of roughly 0.5°C/W or lower. In practice, this means a massive, multi-pound block of aluminum, often requiring forced air cooling. Furthermore, a Class A amp dissipates that 150W of heat even when no music is playing, because the bias current is constant.
Where You Meet Amplifier Classes in Practice
You will encounter different classes depending on the physical constraints and fidelity requirements of your project:
- Class A: Found in high-end boutique audio (like Pass Labs) where heat and power consumption are secondary to eliminating crossover distortion. Also standard in low-noise RF front-ends where linearity is critical to prevent intermodulation distortion.
- Class AB: The workhorse of the industry. If you open a Denon or Marantz AV receiver, or a solid-state guitar amplifier, you will find Class AB push-pull output stages. It offers a great compromise between the heat of Class A and the distortion of Class B.
- Class D: Dominates applications where size, weight, and battery life matter. Active studio monitors (like Kali Audio or Genelec), car audio subwoofers, and portable Bluetooth speakers all use Class D. As of 2026, the adoption of Gallium Nitride (GaN) transistors in Class D designs has pushed switching frequencies into the MHz range, drastically shrinking the required LC output filters.
- Class G / H: Used in professional PA amplifiers (QSC, Crown). These use a Class AB output stage but switch between multiple power supply rails (Class G) or modulate the rail voltage dynamically (Class H) to handle dynamic musical peaks without wasting power as heat during quiet passages.
War Story: When a Class D Subwoofer Amp Turned into an EMI Generator
Class D amplifiers are incredibly efficient, but their high-frequency switching nodes make them notorious for electromagnetic interference (EMI) if the PCB layout is flawed. I learned this the hard way during a custom car audio build.
The Setup: I was building a custom 500W Class D subwoofer amplifier for a car trunk using an IRS2092 evaluation board paired with a 12V-to-±40V isolated DC-DC converter.
The Numbers: The PWM switching frequency was set to 300 kHz. The output filter inductor was rated for 15A continuous, and the 12V power supply was capable of delivering 45A at 14.4V. On the bench, it pushed clean bass into a dummy load.
The Outcome: Once installed in the vehicle, the amp produced excellent bass, but a massive 1kHz alternator whine bled into the head unit, and the AM radio was completely blanked out within a 50-foot radius of the car.
What Went Wrong & The Fix:
- Ground Loop Error: The ground return for the high-current 12V input shared a thin wire with the low-level audio ground. I fixed this by implementing a strict star-ground topology, separating the high-current power ground from the sensitive audio signal ground.
- Capacitive Coupling: The high-frequency PWM traces on the PCB were routed parallel to the unshielded RCA input cables. I re-routed the RCAs away from the switching node and wrapped them in a grounded copper braid shield.
- Conducted Emissions: The 300 kHz switching harmonics were beating against the alternator's ripple on the 12V line. Adding a high-current common-mode choke and a bank of low-ESR ceramic capacitors directly at the amp's 12V input terminals choked off the high-frequency noise before it could reach the car's wiring harness.
Frequently Asked Questions
Is Class D always better than Class AB for audio?
Not necessarily. While modern Class D chips (like those from Texas Instruments or Infineon) achieve THD+N figures below 0.01%, rivaling Class AB, they require careful output filter design to maintain a flat frequency response across varying speaker impedances. If a speaker's impedance drops significantly at high frequencies, a poorly designed Class D filter will cause a treble peak. Class AB is largely immune to this load-dependent frequency shift.
Why do Class A amplifiers run so hot even with no music playing?
In a Class A design, the output transistors are biased heavily into their linear region so they never turn off. This constant DC bias current flows through the transistors continuously, burning power as heat regardless of whether an AC audio signal is present. The amplifier is essentially a 150W space heater that happens to produce music.
What is the difference between Class G and Class H?
Both aim to improve the efficiency of a Class AB output stage. Class G uses discrete, stepped power supply rails (e.g., ±35V for quiet passages, switching to ±70V for loud peaks). Class H uses a continuously variable tracking power supply that modulates the rail voltage in real-time to stay just slightly above the audio signal envelope, offering smoother transitions and slightly better efficiency at the cost of a more complex power supply design.
