Mixed-signal processing-the integration of digital and analog circuitry within computer system-enables systems to take signals from the analog world and process them within a digital system. In fact, recent advances in VLSI technology performance now allow for the integration of digital and analog circuits on a single chip and for the process that requires the use of analog pre- and post-processing system such as data converters, filters, sensors, drivers, buffers, and actuators. This thesis investigates mixed-signal CMOS inte-grated circuit designs for medical X-ray image sensor and liquid crystal display (LCD) column driver chip.
In Chapter 1, a direct photon-counting X-ray image detector with a HgI2 photoconductor is presented for high-quality medical imaging applications. The proposed sampling-based charge preamplifier with asyn-chronous self-reset enables a pixel to detect single X-ray photon energy with higher sensitivity and faster pro-cessing rate. The use of the correlated double sampling enabled by the sampling-based architecture also re-duces flicker noise and contributes to the achievement of high pixel-to-pixel uniformity. Discrimination of the energy level of the detected X-rays is performed by the proposed compact in-pixel ADC with low power con-sumption. Three 15-bit counters in each pixel count up energy-discriminated photons for the reconstruction of multispectral X-ray images. A 128 × 128 X-ray image detector with a pixel size of 60 × 60 μm2 is implemented and measured using a 0.13-μm/0.35-μm standard CMOS process. It discriminates 3 energy levels of photon energy with a gain of 107 mV/ke- and a static power consumption of 4.6 μW/pixel. The measured equivalent noise charge (ENC) and minimum detectable energy level of the detector pixel are 68 e- rms and 290 e , respectively. The measured maximum threshold dispersion in the pixel array is 164 e- rms without any calibration. The functionality of our chip is also successfully demonstrated usin...