Class-D amplifier
A Class-D amplifier is a switching audio power amplifier that utilizes pulse-width modulation (PWM) to drive output transistors between fully on and off states, thereby achieving high efficiency by minimizing power dissipation in the switching devices.[1] Unlike linear amplifier classes such as A, B, or AB, which operate transistors in their linear region and convert excess power to heat, Class-D designs theoretically approach 100% efficiency through binary switching, with practical efficiencies often exceeding 90% at rated output power.[2] The core operation involves a modulator that compares the input audio signal to a high-frequency carrier (typically 200 kHz or higher) to generate a PWM waveform, followed by a power stage of MOSFETs or other switches and a low-pass output filter to reconstruct the amplified audio while attenuating switching harmonics.[1] The concept of Class-D amplification, invented by Alec Reeves in the early 1950s using vacuum tube technology, with the "D" designation simply following the alphabetical progression after Class-C amplifiers, rather than denoting "digital" operation.[1] Commercial availability began in the 1960s with early kits from Sinclair Radionics, though initial designs suffered from performance limitations; significant advancements occurred in the 1990s driven by affordable, high-speed MOSFETs and integrated circuits from companies like Tripath Technology and Philips.[3] A pivotal development was the Universal Class-D (UcD) topology invented by Bruno Putzeys at Philips in the late 1990s and early 2000s, which achieved low distortion (around 0.02%) alongside high efficiency, enabling widespread adoption in consumer audio.[4] Class-D amplifiers excel in applications requiring compact size and extended battery life, such as portable devices, automotive sound systems, and home theater receivers, where their reduced heat dissipation allows for smaller heat sinks and power supplies compared to traditional linear amplifiers (typically 65-70% efficient).[5] They now dominate modern audio equipment, including smartphones and smart speakers, due to their ability to deliver high-fidelity sound with minimal energy loss, though challenges like electromagnetic interference from switching frequencies necessitate careful filtering and shielding.[6] Ongoing innovations, such as gallium nitride (GaN) transistors, continue to enhance performance by further reducing distortion and enabling higher power outputs.[7]Fundamentals
Definition and Principles
A Class-D amplifier is a type of switching audio power amplifier that utilizes power transistors, such as MOSFETs, operated strictly as on/off switches rather than linear devices to amplify an input signal. The process involves converting the analog audio input into a train of high-frequency pulses whose width or density encodes the signal's amplitude, followed by low-pass filtering to recover the amplified analog output. This switching paradigm contrasts with linear amplifiers by avoiding continuous power dissipation in the output stage.[8][9] The fundamental principle relies on pulse-width modulation (PWM) or analogous techniques, where the duty cycle of the pulses proportionally represents the input signal's amplitude. In ideal operation, the transistors experience either full current with minimal voltage drop (when on) or full voltage with no current (when off), resulting in negligible power dissipation across the switches. This enables a theoretical efficiency approaching 100%, as the output power nearly equals the input power from the supply.[8][9] Efficiency is defined as \eta = \frac{P_\text{out}}{P_\text{in}} \approx 100\% under ideal conditions, stemming from minimized dissipation P_\text{diss} in the switches, where conduction losses are limited to I^2 R_\text{on} during the on-state and switching losses are reduced by high-speed transitions. Practical efficiencies often exceed 90%, far surpassing linear classes due to these principles.[8][10][9] The designation "Class-D" originated in the 1950s to signify this switching-based operation, following the alphabetical progression after Class-C amplifiers.[1]Comparison with Other Amplifier Classes
Class-A amplifiers operate with constant conduction throughout the entire input cycle, achieving a theoretical maximum efficiency of 50% while providing very low distortion levels, typically below 0.01% total harmonic distortion (THD).[11][8] Class-B and Class-AB amplifiers employ a push-pull configuration, where Class-B conducts for exactly half the cycle to reach a theoretical maximum efficiency of 78.5%, though practical implementations often yield 50-70% due to crossover distortion in Class-B, which is mitigated in Class-AB at the cost of slightly higher quiescent power.[12][13] Class-C amplifiers, biased beyond cutoff for less than half-cycle conduction, achieve efficiencies exceeding 80%—up to 90% theoretically with tuned loads—but introduce significant nonlinearity and high distortion, making them unsuitable for audio applications and primarily used in radio frequency (RF) systems.[14][15] In contrast, Class-D amplifiers utilize pulse-width modulation (PWM) or similar switching techniques to achieve practical efficiencies of 85-95%, far surpassing linear classes by minimizing power dissipation as heat during conduction.[16] This high efficiency enables compact, battery-friendly designs for portable audio, with output filtering required to reconstruct the analog signal and suppress switching artifacts.[17] While early Class-D designs suffered from higher THD compared to linear classes, modern implementations with feedback loops achieve THD below 0.1%, approaching the linearity of Class-AB amplifiers (typically <0.01% THD).[13][8] The following table summarizes key performance metrics for comparison:| Amplifier Class | Theoretical Max Efficiency (%) | Practical Efficiency Range (%) | Typical THD (%) | Primary Application Notes |
|---|---|---|---|---|
| Class-A | 50 | 20-50 | <0.01 | Low distortion, high heat for audio |
| Class-B/AB | 78.5 | 50-70 | <0.01 (AB) | Balanced for audio, moderate heat |
| Class-C | 90 | >80 | High (>10) | Nonlinear, RF-tuned loads |
| Class-D | ~100 | 85-95 | <0.1 (with feedback) | High efficiency, audio with filtering |