Comparative efficiency analysis of dynamically supplied power amplifiers (PA)

Abstract

Data throughput, reliability, bit-error rates (BER), and backward compatibility complicate modern protocols so much that power amplifiers (PA) drive and demand most of the power mobile devices consume. As a result, PA power efficiency often limits battery life, which is why dynamically adaptive supplies are popular today. Justifying one supply scheme over another in terms of efficiency, however, is not straightforward. This paper analyzes the fundamental loss mechanisms in PAs to compare how power-and envelope-tracking (PT and ET) supplies improve efficiency. Results show that, even when ET supply efficiencies are 20% worse than PT's, ET schemes are 4% -29% more efficient than PT systems for CDMA 1895 (and UMT8), CDMA 2000, and 802.11 a/g protocols across input power, and better in newer protocols where peak-average ratios (PAR) are higher. In fact, envelope elimination and restoration (EER) with a dynamic supply and feedback is optimal with 29% -35% more efficiency than Class-A ET because a nonlinear PA dissipates little conduction power, the smart supply loses less power than the nonlinear PA saves, and feedback corrects nonlinear errors.Index Terms-Envelope Elimination and Restoration (EER), Envelope Tracking, Power Tracking, Power Amplifier (PA). I. POWER-AMPLIFIER EFFICIENCYS MALL and large-scale systems save energy, time, cost, and lives by sharing information across a networked space. In this context, mobile and portable wireless devices play a pivotal role in gathering and disseminating intelligence. Unfortunately, replacing or recharging batteries across a wide area network demands personnel costs and disrupts the network's connectivity, ultimately curbing the benefits of communication. Battery life, as a result, is a critical parameter.Prolonging the single-charge life of a small (and easily exhaustible) state-of-the-art battery amounts to decreasing losses in the system. Regrettably, radiating electromagnetic waves is fundamentally lossy because distant receivers only absorb a small fraction of what a transmitter delivers. As a result, the radio-frequency (RF) power amplifier (PA) demands most of the power a system dissipates [1]-[2], which means that extending life hinges on increasing PA efficiency. II. SUPPLy-MODULATED SCHEMESPAs fall under two categories: linear and nonlinear. While linear PAs amplify their inputs linearly to produce outputs Manuscript ).978-1-4244-8157-6/1 0/$26.00 ©2010 IEEE 607 with proportionately higher power, nonlinear PAs generate constant output power, irrespective of the input. Power in nonlinear outputs, however, increases with supply voltage so adjusting the supply modulates the nonlinear PA's power gain. A. Linear Power AmplifiersIn linear PAs, maximum output power dictates the physical size and conduction time of the lossy power switch for a particular supply voltage V 00. As such, conventional designs favor peak-load efficiency T]PK over its backed-off counterpart T]80 because the voltage (power) dropped across the switch is high during back-off conditions, as the los...