Linear amplifiers are the core of most electronic signal processing, and RF transmitters are no exception. Linear circuitry always operates with the transistor as a controlled current source (CCS). We can state both that a linear amplifier operates as a CCS, and the converse that with CCS operation we have a linear amplifier. With a century of history, the design of linear power amplifiers is covered extensively in the literature [4-1] and will not be repeated here.

In this chapter, the points about linear RF power amplifiers that are most important to their application in DPS architectures are described.

Overview

The entire objective of any linear amplifier is to provide an output signal y(t) that is proportionally scaled from the input signal x(t). The constant of proportionality is called the gain of the amplifier. Mathematically we write this as

y(t) = ax(t) (4 : 1)

where the proportionality constant α is the amplifier gain.

In the real world, the relationship in (4.1) is an ideal goal that is never precisely reached. How hard we must work to make our amplifier approximate this ideal performance more closely is dependent on the signal type we are to amplify, and the output performance specifications we need to meet. In general, the greater the signal order is [4-2], which is the number of possible information values that can be transmitted in any signal symbol, the more precise the amplifier linearity performance must be.

It is very important to be clear that the concept of a linear amplifier is a port-based specification, as shown in Figure 4-1. What the actual circuitry is within the amplifier, and how it precisely operates, does not matter. It is very common to implement a linear amplifier function using linear circuitry. But this is not necessary at all.

What is a requirement, though, is that the input and output signals be essentially sinusoidal in wave shape.