I. Core Classification: By Amplification Method

This is the most fundamental difference between audio amplifier chips, directly determining efficiency, sound quality, and heat generation.

1. Class A

Principle: The amplifying element is always conducting throughout the entire signal cycle, resulting in a very large quiescent current.

Characteristics: Excellent linearity, extremely low distortion, and the best sound quality; however, efficiency is extremely low (theoretically up to 25%, practically <20%), and the chip gets extremely hot year-round.

Current Status: Rarely used in integrated chips; primarily found in high-end discrete component Hi-Fi amplifiers.

2. Class B

Principle: Uses two sets of amplifying elements to process the positive and negative half-waves respectively, cutting off when there is no signal.

Characteristics: Improved efficiency (theoretically 78.5%), but suffers from severe crossover distortion.

Current Status: Pure Class B audio amplifiers are virtually nonexistent, having been replaced by Class AB.

3. Class AB – Classic and Balanced

Principle: Applying a small bias current to the Class AB amplifying components allows them to conduct slightly near the zero signal, eliminating crossover distortion.

Characteristics:

Efficiency is significantly higher than Class A (often reaching 50%-70% in practice)

Distortion is significantly lower than Class B, with sound quality approaching Class A

Simple design and peripherals, extremely low EMI

Representatives: Countless classic amplifier chips belong to this category, such as the LM386 (low-voltage portable), TDA2030/LM1875 (early active speakers), LM3886/TDA7293 (high-fidelity amplifiers), and NE5532 (op-amp, often used in conjunction with power amplifiers).

Applications: Suitable for applications requiring high sound quality, low power consumption, and ample space for heat dissipation, such as Hi-Fi desktop amplifiers, monitor speakers, and original equipment car head units.

4. Class D – High Efficiency and Mainstream

Principle: The analog audio signal is compared with a high-frequency triangular wave, modulated into a PWM square wave (switching signal), driving the power transistor to fully conduct or cut off. The signal is then restored to the audio signal through an LC low-pass filter.

Core Advantages: Extremely high efficiency (generally >80%, even >90%), minimal heat generation, eliminating the need for a large heatsink, suitable for battery-powered and high-power applications.

Technical Keywords:

Filterless: Improved modulation methods (such as spread spectrum modulation) directly use the inductance of the speaker coil for filtering, eliminating the need for an LC filter. Examples: TPA3116D2, MAX98357, etc.

Spread Spectrum and Low EMI: Reducing radiation by dithering the switching frequency is a key feature of Class D design.

Representative examples: TPA3116/3118 (classic high-power), TPA3255 (high-fidelity high-power), TAS5825M (TI‘s DSP digital input Class D), MAX98357 (low-power I2S input), and various smart amplifiers from domestic manufacturers such as Silergy, Southchip, and Shanghai Awinic.

Applications: Mobile phones, tablets, TVs, Bluetooth speakers, TWS earphones, car amplifiers, subwoofers, portable devices… almost the absolute mainstay of modern audio equipment.

5. Class G and Class H (variants with improved efficiency)

Class G: Automatically switches between high and low power rails based on signal amplitude, using low voltage power supply at low volumes to improve efficiency. Commonly found in low-power headphone amplifiers.

Class H: Power supply voltage is continuously adjustable with the signal envelope (similar to envelope tracking), smoother and more efficient than Class G, but more complex to control. Often used in high-power professional amplifiers and some mobile phone smart PAs (adjusting the power supply voltage through boost).

II. How to interpret key parameters? To understand the datasheet, focus on these 8 key metrics:

**Output Power:** Always consider load impedance, supply voltage, and THD+N conditions. For example, "15W @ 8Ω, 24V, THD+N < 1%". Power figures without these conditions are meaningless.

**THD+N (Total Harmonic Distortion + Noise):** Lower is better. High-fidelity chips often achieve below 0.01% or even 0.001%; ​​<0.1% is good for ordinary TVs/speakers; Class D chips have slightly higher high-frequency distortion than Class AB chips.

**Signal-to-Noise Ratio (SNR):** The difference between the signal and the noise floor. Higher SNR results in a darker background. Ideally, it should be >90dB, with high-end models exceeding 110dB.

**Power Supply Rejection Ratio (PSRR):** The ability to suppress power supply ripple. Class AB chips typically have high PSRR (>60dB), while Class D chips are inherently lower, requiring high power supply decoupling. Some models improve this with built-in LDOs.

**Efficiency:** Directly related to heat generation and battery life. Class D > Class AB; Class D‘s advantages are more pronounced at low output power.

Quiet current and shutdown current: Extremely critical for battery-powered devices. Class D amplifiers in TWS earphones have quiescent currents below 1mA and shutdown currents <1μA.

Gain: Fixed gain is simple, while adjustable gain is flexible. Digital input amplifiers typically use I2C digital adjustment.

Protection functions: Over-temperature, over-current, short-circuit, under-voltage, and DC offset protection. Modern chips also include clipping detection and thermal foldback to protect the speaker and prevent harsh noise from hard clipping.

III. Typical Chips by Application Scenarios

Application Scenarios | Main Types | Typical Characteristics | Representative Chips/Series

Mobile Phones / Tablets | Class D Smart Amplifier (Smart PA) | Built-in boost, IV feedback, speaker protection, DSP, I2S/TDM input | Awinic AW87/AW88 series, TI TAS2563, Cirrus Logic CS35L41

TWS Earphones | Ultra-low power consumption, Class D | Quiescent current <1mA, I2S input, no filtering required, extremely small package | Integrated main control from companies like Actions, BES, and Airoha; independent amplifiers such as ESMT AD52068

Bluetooth Speakers | Class D (5-30W) | Single power supply, no filtering required, low EMI, with direct Bluetooth main control connection | TI TPA3136D2, Shanghai Silergy, Nanjing Topmicro, etc.

TVs / Soundbars | Class D Digital Amplifier | Multi-channel (2.0/2.1), high power, built-in DSP/sound effects | TI TAS5825M, ST STA350BW, NXP TDF8530

Desktop Hi-Fi / Active Monitor

Class AB or High-Fidelity Class D

Extremely low THD+N, high SNR, may require dual power supply

LM3886, TDA7293 (AB); TPA3255, Merus™ MA12070 (Class D)

Car Amplifier

Primarily Class D

High voltage (12V/24V), high power, with load diagnostics, high reliability, low EMI

ST HFA80A, TI TAS5421-Q1, ADI/MAXIM MAX13331

Minimalist DIY / Toy

Class AB

Minimalist peripherals, wide voltage range, easy to hand-solder

LM386 (timeless classic, mono low voltage)

IV. Three Steps in Selection

Determine Load and Power

Determine the speaker impedance (2Ω/4Ω/8Ω) and the required maximum undistorted power.

Based on the power and impedance, work backwards from the "Power-Voltage" curve in the chip datasheet to deduce the minimum supply voltage.

Select Amplification Type and Input Interface

Battery powered/limited heat dissipation → Class D is a must.

Demanding sound quality/analog preference → Class AB, but heat dissipation must be carefully calculated.

Input Signal: Is it analog single-ended/differential, or I2S/TDM digital audio from the CPU? Digital input eliminates the need for a DAC, but requires software development; analog input is the most universal.

Assess System Requirements

Is a DSP (equalizer, dynamic bass, 3D surround) required?

Is a boost converter needed to achieve high power at low battery voltages (Smart PA)?

Are EMI requirements stringent (e.g., automotive, portable electronics)? Consider the chip‘s spread spectrum and edge control capabilities.

Is protection comprehensive? In high-power scenarios, short-circuit, over-temperature, and DC detection are essential, especially clipping control to prevent damage to the tweeter.

V. Several Trends in 2026

Maturation of the All-Digital Link: Amplifier chips directly receive I2S/TDM signals from the SoC, internally performing decoding, DSP audio effects, and amplification, reducing conversion steps.

"Smart Amplifiers" Continue to Deploy: Speaker amplitude protection, temperature prediction, and dynamic compensation algorithms, previously only used in flagship phones, are now increasingly appearing in mid-range phones, tablets, and laptops.

Higher Voltage, Smaller Package: Thanks to advancements in manufacturing processes, the on-resistance of Class D chips is decreasing, enabling higher power output with the same heat dissipation.

Deep Efficiency Optimization: In real-world usage scenarios (below medium volume), greater emphasis is placed on light-load efficiency and static power consumption, which significantly impacts battery life.

High-Fidelity Class D Approaching Class AB: Through higher-frequency modulation and advanced feedback loops, excellent Class D amplifiers can achieve a stable THD+N level of 0.001%, challenging the dominance of traditional Hi-Fi amplifiers.