Harmonic load modulation in Doherty amplifiers

Abstract

In a Doherty amplifier employing a class-B main amplifier and a class-C peaking amplifier, harmonics are normally suppressed at each amplifier. If odd harmonic currents generated by the peaking amplifier are permitted to flow into the main amplifier, the loadline of the main amplifier is shaped in such a manner as to allow a lower supply voltage to be used.Introduction: As the RF power amplifier draws a significant amount of current from the power supply, it is important to maximise its DC to RF conversion efficiency over a wide dynamic range. However, a conventional RF power amplifier only achieves maximum efficiency at maximum drive and this means low average efficiency when the envelope is modulated. The Doherty amplifier [1] alleviates this problem by using two transistors with one, the main amplifier, operating over the entire range of input powers, and the other, the peaking amplifier, only operating over the upper end (usually the top 6 dB) of the dynamic range. Over the upper end of the dynamic range, the load resistance seen by the main amplifier decreases with input level, causing it to operate at maximum efficiency in this range; this is called load modulation. Usually the main amplifier operates in class-B, and the peaking amplifier in class-C. It is often assumed that harmonic suppression occurs within these two amplifiers, and only the fundamental plays a role in the Doherty amplifier operation.In this Letter we show that odd harmonics generated in the peaking amplifier can be used by the main amplifier to further enhance efficiency. We call this harmonic modulation, and the main amplifier operates in a similar manner to class-F operation [2]; but this approach is distinct from class-F Doherty amplifiers that use separate class-F main and peaking amplifiers [3,4].

High Efficiency Power Amplifiers for Modern Mobile Communications: The Load-Modulation Approach

“…In [31,32], two extremely compact output combining networks integrating harmonic load terminations are proposed, based on a lumped-element and a harmonic-trap network, respectively. As a final example, in [33], the odd harmonics generated in the Auxiliary amplifier are exploited to shape the drain voltage waveform of the Main device allowing it to operate at a reduced supply voltage, thereby increasing efficiency.…”

A compact class‐F/class‐C Doherty amplifier

“…Figure 1 shows the basic schematic of the Doherty amplifier that employs harmonic load modulation [9]. There are three key transmission lines (TL m , TL D , and TL p ), each quarter‐wave at the fundamental frequency.…”

“…As odd harmonics from the peaking FET flow through TL D and TL m , odd‐harmonic standing waves are established on TL D with antinodes at the main FET. These odd‐harmonics “square” the main FET drain voltage waveform and we denote this harmonic load modulation [9]. The third harmonic is the larger of all odd harmonics [2], so for argument sake, we consider the third harmonic and the fundamental in the analysis that follows.…”

“…If V knee is the drain voltage delineating the triode region, the minimum value of V DDmain is approximately: Whereas for class‐B operation, the minimum supply voltage is V knee + a 1main R Lmain . Hence with harmonic load modulation, we can reduce the supply voltage thereby increasing efficiency, without compromising the output power [9].…”

“…The authors have recently proposed a Doherty amplifier where odd harmonics generated in the peaking amplifier flow to the main amplifier where they perform useful waveform shaping [9]. Aside from improvements in efficiency, this approach leads to a simplified circuit without complicated harmonic terminating networks.…”