Exploring and modeling of non-linearity in BJT to optimize it for Wi-Fi 7 power amplifiers

Abstract

Power Amplifiers are an integral part of modern-day wireless communication systems. Nonlinearity in power amplifiers is an undesirable problem but is a common phenomenon in its applications. For advanced wireless applications of modern days, reducing nonlinearities has become a challenging task. Studies have been done to understand the origin mechanisms of these nonlinearities and their mitigation techniques. HBTs provide higher gain and speed compared to CMOS, but the advanced HBTs have complicated architecture, and the Equivalent Circuits are often not comprehensible for designers to gain insight into the origin of nonlinearity. Thus, an efficient method to investigate these nonlinearities and find an easier way to understand them has always been of interest to designers. A systematic way of investigating the nonlinearity in HBT of PA was reported in this study. A custom model was built to help circuit designers understand the physical effects involved in the nonlinearity, without having to dive deep into transistor architecture or the equivalent model. The key effects and regions playing a role in this nonlinearity were identified, such as exponential IV characteristics between IC and VBE, especially at high power, the BC saturation effect, the early effect especially the reverse early effect, the high-injection effect, and the transit time effect. To get more insight into the origin of these nonlinearities, the components of the HICUM/L0 model were also investigated. From over a hundred components/parameters that make up the equivalent model HICUM/L0, the key parameters that impact the linearity of the power amplifier’s performance were identified. With the help of Monte Carlo simulation, a comprehensive relation between these parameters and the transistor’s physical architecture is also presented in the report. The correlation between these parameters/regions with the transistor and the performance metrics was also identified. The proposed framework provides a systematic way of analyzing nonlinearity in the PA with key insights into the nonlinear effects and their origins.