Introduction to Power Amplifier Circuits

Power amplifiers (PAs) are foundational in a vast range of electronic systems—from the tiny output stage in a smartphone’s RF front-end to the massive tubes driving a concert sound system. Their primary job is to take a weak input signal and deliver a powerful, faithful copy at the output. Two metrics dominate the performance evaluation of any PA: power gain and linearity. Gain tells us how much the signal is amplified; linearity tells us how cleanly that amplification happens. This article explores both in depth, their interplay, the engineering trade-offs involved, and the practical techniques used to balance them in real-world designs.

What Is Power Gain?

Power gain is the most direct measure of an amplifier’s ability to increase signal strength. It is defined as the ratio of output power to input power, usually expressed in decibels (dB) because dB scales compress wide dynamic ranges into manageable numbers.

Mathematically, power gain G in dB is:

G (dB) = 10 log₁₀ (Pout / Pin)

For example, if an amplifier delivers 1 watt from a 1 milliwatt input, the gain is 10 log₁₀(1/0.001) = 30 dB. A gain of 20 dB corresponds to a factor of 100 in power, while 10 dB is a factor of 10. In practice, PAs in base stations may have gains of 30–50 dB, while handset PAs often range from 20–35 dB.

Measuring Power Gain

Gain is rarely constant across frequency, temperature, or output power level. Engineers specify small-signal gain (measured with very low input power, where the device is fully linear) and large-signal gain (at a given saturated output power). Three common gain definitions are used:

  • Transducer power gain (GT) – assumes source and load are matched to the amplifier’s ports. This is the most practical definition for system design.
  • Available gain (GA) – optimizes the source impedance for maximum output, ignoring load mismatch. Often used in noise analysis.
  • Operating power gain (GP) – defined as power delivered to the load divided by power input to the amplifier. Useful for stability analysis.

Modern vector network analyzers (VNAs) and spectrum analyzers make these measurements routine. For a deeper dive into gain measurement methods, see this Analog Devices application note on RF power gain measurements.

Understanding Linearity in Power Amplifiers

Linearity describes how faithfully an amplifier reproduces the input signal’s amplitude and phase. A perfectly linear amplifier would output an exact scaled replica of the input; any deviation constitutes distortion. In real PAs, linearity degrades as output power approaches the device’s maximum capability. This non-linear behavior generates harmonics, intermodulation products, and compression – all of which degrade signal quality in communication systems.

Key Metrics of Linearity

Several figures of merit quantify PA linearity:

  • 1-dB Compression Point (P1dB) – The output power level at which the actual gain falls 1 dB below the ideal linear gain. It is the simplest metric for gain compression. Below P1dB the amplifier is essentially linear; above it, distortion increases rapidly.
  • Third-Order Intercept Point (IP3) – A theoretical point where the power of the third-order intermodulation products equals the fundamental signal power. Higher IP3 indicates better linearity. It is usually measured as Output IP3 (OIP3) or Input IP3 (IIP3).
  • Adjacent Channel Power Ratio (ACPR) – The ratio of in-channel power to power spilled into adjacent channels due to spectral regrowth. Crucial for cellular and Wi-Fi transmitters to meet regulatory limits.
  • Error Vector Magnitude (EVM) – A measure of modulation accuracy. High linearity yields low EVM; practical limits for LTE/5G are typically 3–8% EVM.

A detailed discussion of these metrics can be found in this Microwaves101 linearity encyclopedia entry.

Types of Nonlinear Distortion

The nonlinear transfer function of a PA creates several distortion products:

  • Harmonic Distortion – Output signals at integer multiples of the input frequency. Harmonics are usually filtered out but can degrade efficiency if reflected back into the device.
  • Intermodulation Distortion (IMD) – When two or more tones pass through a nonlinear PA, sum and difference frequencies appear. Third-order (IM3) and fifth-order (IM5) are most troublesome because they fall close to the fundamental and cannot be easily filtered.
  • AM-AM & AM-PM Conversion – Amplitude-to-amplitude (AM-AM) and amplitude-to-phase (AM-PM) conversions describe how the output amplitude and phase vary with input power. These effects directly cause distortion in modulated signals.

The Gain–Linearity Trade-Off

There is a fundamental tension between achieving high power gain and maintaining excellent linearity. Every transistor has a finite transconductance and output swing; driving it harder to get more gain also pushes it into saturation, where gain compresses and distortion rises. The relationship is governed by the device’s nonlinear characteristics and the circuit topology chosen.

Why the Conflict Exists

In a simplified transistor model, the output current is a polynomial function of the input voltage. The linear term provides gain; higher-order terms (square, cube, etc.) create distortion. As you increase input drive to extract more linear gain, the higher-order terms become significant. Furthermore, the transistor’s parasitic capacitances vary with voltage (Ccb, Cbe), producing AM-PM distortion that worsens near saturation. High gain often requires higher drain or collector currents, which can degrade efficiency and thermal performance, indirectly harming linearity.

Design Techniques to Balance Gain and Linearity

Engineers employ a range of methods to improve linearity without sacrificing too much gain:

  • Negative Feedback – Feeding a portion of the output back to the input in opposite polarity reduces overall gain but linearizes the transfer characteristic. Common in audio op-amps and some RF PAs.
  • Biasing Optimization – Selecting the quiescent point (Class A, AB, B, or C) directly affects linearity and gain. Class A gives best linearity but poorest efficiency; Class AB balances both; Class C offers high efficiency at the cost of severe nonlinearity.
  • Emitter or Source Degeneration – Adding a small resistor (or inductive degeneration at RF) introduces local feedback, linearizing the amplifier at a mild gain penalty.
  • Digital Predistortion (DPD) – An advanced technique where the input signal is pre-distorted in the digital domain to exactly cancel the amplifier’s nonlinearities. DPD can improve ACPR by 10–20 dB and is standard in modern basestation PAs.
  • Doherty Architecture – Uses a main amplifier (biased in Class AB) and a peaking amplifier (Class C) that turns on at high power. The combination yields high back-off efficiency and good linearity across a wide power range.

For a comprehensive overview of modern linearity enhancement, refer to this EDN article on RF power amplifier linearity enhancement.

Practical Applications and Performance Demands

Different applications impose distinct requirements on gain and linearity. Below we explore three major domains.

Radio Transmitters (Cellular Base Stations, Broadcast)

Modern cellular standards (4G LTE, 5G NR) use complex modulation schemes like 256-QAM OFDM, which have high peak-to-average power ratios (PAPR) of 8–12 dB. The PA must remain linear for large envelope swings to avoid spectral splatter into adjacent channels. Basestation PAs typically achieve 50–60 dB gain using GaN or LDMOS transistors, with OIP3 above 50 dBm. DPD and Doherty architectures are essential to meet stringent ACPR and EVM masks while keeping efficiency above 40%.

Audio Amplifiers

Audio PAs prioritize linearity (low total harmonic distortion, THD) over raw gain. The human ear is sensitive to even 0.1% THD, especially in midrange frequencies. Many high-fidelity audio amplifiers operate in Class A or Class AB with heavy global feedback to reduce distortion to below 0.01%. Gain is typically moderate (20–30 dB) because the preceding preamplifier already provides substantial voltage gain.

RF Communication Handsets and IoT

Small, battery-powered devices need PAs that are simultaneously efficient, compact, and linear enough for the modulation scheme. For example, a Wi‑Fi 6 PA in a smartphone must deliver +17 dBm output with EVM under 3% (64-QAM MCS7) while drawing less than 100 mA. CMOS PAs with on-chip linearization (e.g., adaptive biasing, capacitive compensation) are common. The gain is lower (20–25 dB) to reduce risk of instability in the compact layout.

Instrumentation and Test Equipment

Wideband laboratory amplifiers that drive signal generators or RF power meters often stress gain flatness and low distortion over many octaves. They use Class A biasing with very high quiescent current to maintain linearity over a >10 dB back-off. Gain is typically 30–40 dB, with OIP3 specifications given at 1 dB compression point headroom.

Emerging Technologies and Future Directions

The relentless push for higher data rates and lower power consumption drives innovation in PA design.

  • GaN on SiC – Gallium Nitride transistors offer very high breakdown voltage and current density, enabling PAs with gain >20 dB up to millimeter-wave frequencies while maintaining excellent linearity.
  • Envelope Tracking (ET) – The supply voltage to the PA is modulated dynamically to track the signal envelope. ET improves efficiency at back-off by reducing dissipation, and careful design preserves linearity.
  • Load Modulation and Outphasing – Architectures like Chireix and LINC (Linear amplification using Nonlinear Components) use two or more PAs combined with varying phase to reconstruct a linear signal from nonlinear building blocks.
  • Machine Learning for PA Linearization – Neural networks and reinforcement learning are being applied to optimize DPD coefficients in real time, achieving better correction with lower computational overhead than traditional polynomial models.

For further reading on GaN PA linearity, see this Qorvo blog on GaN power amplifier linearity.

Conclusion

Power gain and linearity are two pillars of power amplifier design, locked in a constant trade-off that engineers must carefully manage. Gain provides the brute strength to boost weak signals; linearity ensures that the amplified signal remains clean and usable. Through clever circuit techniques—feedback, biasing, Doherty architectures, and digital predistortion—designers push the envelope of both performance metrics simultaneously. As next-generation wireless systems demand ever-higher throughput and lower power consumption, the art of balancing gain and linearity will remain at the heart of innovation in power amplifier circuits.