Introduction

The power amplification of amplitude modulated RF signals has two inherent problems. The first is that the envelope, and hence the modulating signal, will be distorted to some degree, if the power amplifying device is used at its full rated RF power level. The second is that conventional power amplifier designs only give maximum efficiency at a single power level, which is dependent on the circuit design but usually near the maximum rated power for the device. As the drive power is backed off from this point, the efficiency drops sharply and the heat dissipation can increase, even though the RF output power is also decreasing. The overall effect is therefore to measure a mean efficiency which is much lower than the efficiency at the maximum power, or PEP, level.

Solutions for the second of these problems have been available for many years and can be quite effective; they will be termed efficiency enhancement techniques. The first problem, which essentially requires the transfer characteristic of the amplifier to be made more linear over a given power range, represents a bigger challenge, and the solutions always have some limitations. In particular, higher linearity will always come at a price, usually in the form of lower efficiency but also in baseband frequency limitations. Such techniques are nevertheless widely used, and are quite indispensable in many transmitter applications. Logically, the term linearization should only be applied to this latter category. Inevitably, some implementations use both efficiency enhancement and linearization techniques together, and...