Understanding the AB Class Amplifier: Theory, Biasing, and Real-World Math

A Class AB amplifier is a push-pull audio circuit where both output transistors conduct slightly more than half a cycle, eliminating the dead-zone crossover distortion of Class B while maintaining much higher efficiency than Class A. In a real PCB layout or schematic, adopting an AB class amplifier topology changes your thermal management requirements and forces the addition of a bias spreader network to keep the output devices slightly turned on at idle. Hobbyists and junior engineers frequently confuse this topology with Class D (which uses high-frequency PWM switching and LC filters) or Class H (which dynamically switches rail voltages to improve AB efficiency), but Class AB remains a purely linear, analog output stage.

The Core Mechanics of a Class AB Amplifier

To understand why the AB class amplifier exists, you have to look at the flaws of its predecessors. A pure Class A amplifier keeps its output transistor(s) conducting 100% of the time. It sounds incredible because there is zero switching delay, but it burns massive amounts of power as heat even when no music is playing. A Class B amplifier uses two transistors in a push-pull configuration: one handles the positive half of the audio waveform, and the other handles the negative half. The problem? Transistors require about 0.6V to 0.7V at the base-emitter junction to start conducting.

This creates a "dead zone" at the zero-crossing point where neither transistor is fully on, resulting in harsh crossover distortion. Think of it like a relay race baton handoff where both runners stop completely before passing the baton; the momentum is ruined. The AB class amplifier solves this by applying a small DC bias voltage to the bases of both transistors, keeping them just barely on the edge of conduction. When the signal crosses zero, the baton is passed seamlessly because both runners are already moving.

Worked Example: Biasing and Thermal Math for a 50W Output Stage

Let's design the bias and calculate the heatsink requirements for a discrete AB class amplifier output stage using complementary ON Semiconductor MJL21193 (NPN) and MJL21194 (PNP) power transistors.

Assumptions & Circuit Parameters:

• Power Supply: Dual rails at ±35V DC
• Load: 8Ω nominal loudspeaker
• Transistor Saturation Voltage (V_CE(sat)): ~3V
• Target Quiescent Current (I_q): 50mA (standard for eliminating the crossover notch without excessive idle heat)

1. Maximum Output Power Calculation

The peak voltage before clipping is the rail voltage minus the saturation drop: 35V - 3V = 32V peak.
Converting to RMS: 32V / √2 = 22.6V RMS.
Maximum continuous power into 8Ω: P = (22.6²) / 8 = 63.8 Watts.

2. Quiescent (Idle) Power Dissipation

With no audio signal, the 50mA bias current flows through both transistors from the positive rail to the negative rail. The voltage across each transistor is 35V.
Idle dissipation per transistor: P_q = 35V × 0.05A = 1.75 Watts.
This is remarkably low compared to Class A, meaning the transistors only need a tiny heatsink just to survive at idle.

3. Heatsink Sizing Under Full Load

The real thermal challenge happens under load. In a Class AB topology, maximum average power dissipation in the output devices does not occur at maximum volume; it occurs when the output voltage is roughly 63% of the rail voltage.
Max average dissipation per transistor: P_D(max) ≈ V_CC² / (π² × R_L) = 35² / (9.87 × 8) = 15.5 Watts.

Now we calculate the required heatsink thermal resistance (θ_SA):

• Max Junction Temp (T_j): 150°C
• Ambient Temp (T_a): 25°C
• Junction-to-Case (θ_JC): 1.5°C/W (TO-3P package)
• Case-to-Sink (θ_CS): 0.5°C/W (with thermal pad)

θ_SA = (T_j - T_a) / P_D(max) - (θ_JC + θ_CS)
θ_SA = (150 - 25) / 15.5 - (1.5 + 0.5) = 8.06 - 2.0 = 6.06°C/W.

You must source an extruded aluminum heatsink rated for 6.0°C/W or lower per channel. If you omit the thermal pad or use a smaller sink, the transistors will hit thermal shutdown or destructively avalanche during loud bass transients.

Where You Meet Class AB Topologies in Practice

While Class D switching amplifiers have largely taken over portable Bluetooth speakers, subwoofers, and budget AV receivers due to their 90%+ efficiency, the AB class amplifier remains the gold standard in specific high-fidelity and low-noise environments in 2026.

• Powered Studio Monitors: High-end nearfield monitors (like classic Yamaha or Genelec models) often use Class AB for their tweeter amplifiers. The high-frequency open-loop gain of a linear AB stage is superior, and it avoids the high-frequency EMI noise that Class D LC filters can sometimes inject into sensitive studio environments.
• Chip Amplifiers: The legendary Texas Instruments LM3886 and STMicroelectronics TDA7294 are internally Class AB. They are heavily favored by DIY audio builders because they require minimal external components, offer excellent thermal protection, and deliver 50W-70W of pristine audio without the complexity of discrete bias networks.
• Headphone Amplifiers: Desktop headphone amps (such as those using the TI OPA1612 or discrete Burr-Brown designs) almost exclusively use Class AB. Headphones are high-impedance, low-power loads, meaning the efficiency penalty of Class AB is irrelevant, while its ultra-low noise floor is critical.

For a deeper look into linear audio topologies, the Texas Instruments audio portfolio documentation provides excellent comparative data on when to select linear AB versus switching D architectures. Additionally, electronics-tutorials.ws offers foundational schematic breakdowns of the V_BE multiplier bias networks discussed above.