What are the trade-offs between linearity and efficiency in RF power amplifiers?

1 Linearity vs. Efficiency

Linearity is the ability of a PA to amplify a signal without distorting its shape or frequency. A linear PA preserves the information content and quality of the signal, and reduces the interference with other signals in the same frequency band. However, linearity comes at a cost of efficiency, which is the ratio of the output power to the input power of a PA. An efficient PA consumes less power and generates less heat, which are desirable for battery-operated and compact devices. However, efficiency often requires operating the PA in a nonlinear region, where the output signal is not proportional to the input signal, and hence introduces distortion and harmonics.

2 Sources of Nonlinearity

Nonlinearity in PAs can arise from several sources, such as the intrinsic characteristics of the active devices (e.g., transistors), the parasitic elements of the passive components (e.g., resistors, capacitors, inductors), and the mismatch between the input and output impedances of the PA and the signal source and load. These sources can cause nonlinear phenomena such as amplitude modulation (AM), frequency modulation (FM), intermodulation distortion (IMD), and memory effects. These phenomena can degrade the signal quality and spectral efficiency, and increase the bit error rate (BER) and adjacent channel power ratio (ACPR) of the communication system.

3 Measures of Linearity

When considering the linearity of a PA, several metrics can be used depending on the application and signal characteristics. Common metrics include the 1 dB compression point (P1dB), which is the output power level at which the gain of the PA drops by 1 dB from its linear value, the third-order intercept point (IP3), which is the hypothetical output power level at which third-order intermodulation products (IM3) have the same power as the fundamental output signals, error vector magnitude (EVM), which is the ratio of root mean square (RMS) error between ideal and actual output signal vectors to RMS magnitude of ideal signal vector, and spurious-free dynamic range (SFDR), which is the difference between maximum output signal power and minimum spurious signal power within a given frequency band.

5 Architectures for Efficiency

To improve the efficiency of a power amplifier (PA), several architectures can be used, such as Class A, B, C, D, E, F, Doherty, and Outphasing. Class A involves biasing the active device in the linear region and conducting for the entire input signal cycle, resulting in high linearity but low efficiency (25-50%). Class B biases the active device at the cutoff point and conducts for half of the input signal cycle, providing moderate linearity and efficiency (50-70%). Class C biases the active device below the cutoff point and conducts for less than half of the input signal cycle, resulting in low linearity but high efficiency (70-90%). Class D switches on and off using a pulse-width modulated (PWM) signal that yields very high efficiency (90-100%) but high distortion and switching noise. Class E switches on and off with a sinusoidal signal that produces high efficiency (80-90%) with low distortion and switching noise due to a resonant circuit. Class F utilizes a harmonic-rich signal to achieve high efficiency (80-90%) with low distortion and switching noise thanks to a harmonic filter. Doherty combines two PAs - one operating in class B or C as the main amplifier and one operating in class C as the peaking amplifier - that are activated according to the input power level, yielding high efficiency over a wide dynamic range. Lastly, Outphasing combines two PAs operating in class D or E that are driven by phase-modulated signals, leading to high efficiency and linearity due to a combiner circuit.

6 Here’s what else to consider

This is a space to share examples, stories, or insights that don’t fit into any of the previous sections. What else would you like to add?