How a Class B Amplifier Works: Theory, Efficiency, and Crossover Distortion

A Class B amplifier is an electronic circuit where the output transistors conduct for exactly 180 degrees of the input signal cycle, relying on a complementary push-pull pair to reconstruct the full AC waveform. In a real circuit, shifting from a Class A topology to an amplifier B configuration slashes idle power dissipation and boosts theoretical efficiency from a maximum of 25% up to 78.5%, but it trades that thermal headroom for zero-crossing artifacts. Beginners frequently confuse pure Class B with Class AB; the latter adds a small quiescent bias voltage to keep both transistors slightly on, eliminating the dead zone inherent to pure Class B designs.

The Core Mechanics of a Class B Amplifier

Unlike Class A amplifiers, which bias the transistor in the middle of its linear region so it conducts for the full 360 degrees of the waveform, a pure Class B stage biases the transistors exactly at cutoff. This means the quiescent current (idle current) is essentially zero. When no audio or RF signal is present, the amplifier draws almost no power and generates virtually no heat.

To handle a full AC waveform, an amplifier B circuit uses two transistors in a "push-pull" arrangement: an NPN transistor handles the positive half-cycle, while a PNP transistor handles the negative half-cycle. Think of it like a two-lane bridge where northbound traffic uses one lane and southbound traffic uses the other, but the toll booth requires a 0.6V "toll" to open the gate, causing a traffic jam at the exact moment the light changes direction.

Because silicon bipolar junction transistors (BJTs) require approximately 0.6V across the base-emitter junction to begin conducting, neither transistor turns on when the input signal is between -0.6V and +0.6V. This creates a flat spot in the output waveform known as crossover distortion. While highly efficient, this distortion makes pure Class B largely unsuitable for high-fidelity audio without modification.

Numeric Breakdown: Efficiency and Power Dissipation

To understand why engineers accept the hassle of crossover distortion in certain applications, we need to look at the math. Let us calculate the maximum theoretical performance of a complementary push-pull amplifier B stage driving a standard speaker load.

1. Maximum AC Power Output ($P_{out}$)
The maximum power delivered to the load occurs at peak voltage swing.
$P_{out} = \frac{V_{peak}^2}{2 \times R_L} = \frac{12^2}{2 \times 8} = \frac{144}{16} = \mathbf{9 \text{ Watts}}$

2. DC Power Drawn from the Supply ($P_{DC}$)
The power supply provides current in half-sine pulses. The average DC current is $I_{peak} / \pi$.
$I_{peak} = \frac{12V}{8\Omega} = 1.5A$
$P_{DC} = \frac{2 \times V_{CC} \times I_{peak}}{\pi} = \frac{2 \times 12 \times 1.5}{3.14159} = \mathbf{11.46 \text{ Watts}}$

3. Maximum Efficiency ($\eta$)
$\eta = \frac{P_{out}}{P_{DC}} = \frac{9}{11.46} = 0.785 \text{ or } \mathbf{78.5\%}$

4. Transistor Power Dissipation ($P_{diss}$)
The remaining power is burned off as heat in the transistors.
$P_{diss} = P_{DC} - P_{out} = 11.46W - 9W = \mathbf{2.46 \text{ Watts}}$

At 9W of audio output, the transistors only need to dissipate 2.46W combined (about 1.23W each). In a Class A amplifier delivering the same 9W, the transistors would dissipate at least 18W continuously, requiring massive heatsinks. For a deeper look at the derivations behind these formulas, the All About Circuits semiconductor textbook provides excellent foundational math.

Where You Meet This in Practice

In 2026, you will almost never find a pure Class B amplifier in commercial audio equipment. The crossover distortion is simply too harsh for the human ear, generating high-order harmonics that fatigue the listener. Instead, audio designers use Class AB, which injects a tiny bias current to push the transistors just past the 0.6V threshold, smoothing out the zero-crossing deadband at the cost of a slight efficiency drop.

However, you will still encounter pure Class B topologies in specific niches:

• RF Power Amplifiers: In radio frequency transmitters, the signal is often a continuous wave or heavily filtered, and the human ear is not listening directly to the raw zero-crossing. Push-pull Class B (or Class C) is common in RF final amplifier stages where efficiency dictates battery life or thermal limits.
• Transformer-Coupled Legacy Designs: Older tube amplifiers and vintage transistor radios sometimes used center-tapped transformers to sum the push-pull halves, naturally masking some of the sharp crossover edges through inductive smoothing.
• Motor Drive H-Bridges: While not amplifying audio, the concept of alternating conduction paths for positive and negative cycles is the exact foundational theory behind DC motor H-bridge drivers and basic PWM switching stages.

Bench Walkthrough: Building, Testing, and Failing with Pure Class B

Theory is clean; the workbench is messy. Here is a real-world scenario demonstrating exactly why pure Class B fails for audio, complete with the measurements that prove it.

The Setup
We breadboard a complementary push-pull amplifier B stage using a 2N3904 (NPN) and a 2N3906 (PNP). The bases are tied together to receive the input signal. The emitters are tied together to drive an 8$\Omega$ 5W power resistor acting as a dummy load. We connect a 12V DC bench supply and feed a 1kHz sine wave from a function generator into the bases via a coupling capacitor.

The Numbers
We set the function generator to output a 2V peak-to-peak (Vpp) sine wave. We connect Channel 1 of the oscilloscope to the input and Channel 2 to the 8$\Omega$ load.

The Outcome
Looking at the oscilloscope, the output waveform is not a clean sine wave. Instead, it looks like a square wave with rounded tops. Every time the input signal crosses the 0V center line, the output voltage flatlines at exactly 0V for a duration of roughly 1.2 microseconds. When we connect a small speaker to the output, the 1kHz tone sounds incredibly "buzzy" and harsh, completely lacking the smooth character of the original signal.

What Went Wrong
The flatline is the crossover distortion in action. The 2N3904 requires $V_{BE}$ to reach +0.6V to turn on, and the 2N3906 requires -0.6V. Because the bases are tied directly together with no biasing network, there is a 1.2V "dead zone" where neither transistor is conducting. The load sees zero current during this transition. According to Electronics Tutorials, this dead zone generates massive odd-order harmonic distortion. To fix this on the bench, we would insert two 1N4148 diodes between the bases, forward-biasing them with a resistor to the supply rail. This drops exactly 1.2V across the diodes, keeping both transistors on the very edge of conduction and instantly transforming our harsh Class B circuit into a smooth Class AB amplifier.

Frequently Asked Questions