混合增益控制機制之CMOS可程式化增益放大器

Abstract

本論文針對可程式化增益放大器（Programmable Gain Amplifier, PGA）提出了幾種增益控制機制。在超音波影像接收器中，需要一個精確的分貝線性PGA，將反射訊號的動態範圍，轉換至類比轉數位轉換器可接受的範圍。為了在很小的增益誤差下，得到較寬的增益範圍與精確的解析度，新提出的PGA採用了指數與近似指數函數兩種變化模式。本論文中討論且分析了幾種可變增益放大器的架構，以及近似指數函數。為了比較兩個不同架構，PGA以0.18-μm互補式金屬氧化物半導體製程製作，包含開路與閉路兩種架構。開路的PGA可以得到較佳的雜訊特性，而閉路的PGA則可以提供較佳的線性度效能。

Several gain control mechanisms for the design of programmable gain amplifier (PGA) are presented in this thesis. In ultrasound imaging receivers, an accurate decibel (dB)-linear PGA is required to convert the dynamic range of reflected signal into an acceptable range for the analog-to-digital converter (ADC). In order to obtain a wide dB-linear gain range and fine step resolution with small gain error, both exponential and exponential approximation functions are used in the proposed PGAs. Several variable gain amplifier architectures and exponential approximation functions are discussed and analyzed in this thesis. Two different PGA architectures are fabricated in 0.18-μm CMOS process for comparison, including an open-loop based and a closed-loop based architectures. The open-loop based PGA can achieve a low noise characteristic, while the closed-loop based PGA can provide a better linearity performance.