Dual-Path Power Amplifier Switching for Low-Power RF Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing power amplifier systems face efficiency challenges, particularly at low power levels, due to inefficiencies in DC-DC converters, which become significant limitations in modern RF communication systems, especially with the increased time spent in low-power modes by 5G handsets, necessitating a new architecture for improved efficiency across various power ranges.
Innovation Solution
A power amplifier system with a reconfigurable switch network and a dedicated CMOS low-power path that reuses the shunt switch of the main high-power path as the series switch for the low-power path, combined with a shared reconfigurable input impedance matching network, allowing for efficient operation across high and low power levels without significant additional cost or die area.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Power
If a standard power amplifier configuration with DC-DC converter is used, then high power output (30 dBm) can be achieved, but efficiency drops significantly at low power levels (0-5 dBm)
Solution Approach 1:
The power amplifier system is divided into two separate amplifier paths: a high-power path for 20-30 dBm output and a low-power path for 0-15 dBm output. Each path is optimized for its specific power range, with the low-power path eliminating the inefficient DC-DC converter for low power operation. This segmentation allows the system to achieve high efficiency across the entire power range by selecting the appropriate path based on required output power.
Solution Approach 2:
The system dynamically switches between high-power and low-power amplifier paths based on the required output power level. A switch network controlled by power level detection enables seamless transition between paths, ensuring optimal efficiency for the current operating condition. This dynamic reconfiguration resolves the contradiction by adapting the system architecture to match the instantaneous power requirements.
2Loss of energy
If separate power amplifiers for different power ranges are used, then efficiency across all power levels improves, but die area and cost increase
Solution Approach 1:
The input impedance matching network is merged and shared between the high-power and low-power amplifier paths. This shared network eliminates the need for separate matching networks for each path, significantly reducing die area while maintaining the efficiency benefits of having separate amplifier paths. The shared resource approach resolves the contradiction between performance and area by consolidating common functions.
Solution Approach 2:
The shared input impedance matching network serves multiple functions: it provides impedance matching for both high-power and low-power paths, and it reduces overall component count. This multi-functionality allows the system to achieve wide power range efficiency without proportionally increasing die area, as the same hardware structure serves multiple operational modes.
3Use of energy by moving object
If supply voltage is scaled down for low power mode, then power consumption decreases, but DC-DC converter efficiency drops below 50%
Solution Approach 1:
The DC-DC converter is extracted from the low-power path and retained only in the high-power path. For low-power operation (0-15 dBm), the system bypasses the DC-DC converter entirely and uses direct power supply, eliminating the source of efficiency loss. This extraction resolves the contradiction by removing the inefficient component from the operational path where it causes harm, while preserving it where it remains beneficial.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A power amplifier system is disclosed having a first amplifier with a high-power input and a high-power output. A second amplifier has a low-power input and a low-power output. A reconfigurable mode switch network has a first series switch branch coupled between the high-power output and an RF output, a first shunt branch is coupled between the RF output and a fixed voltage node, and a second series switch branch is coupled between the low-power output and a shared node of the first shunt branch. The shared node separates the first shunt branch into a first shared section that is between the RF output and the shared node and a second shared section that is between the shared node and the fixed voltage node.