(A) low-power high-resolution CMOS sensor system for IoT applications

Abstract

With the emergence of the Internet-of-Things (IoT), the number of IoT devices exceeded 50 billion, which rapidly increases the demand for sensors. Moreover, most sensor systems are driven using batteries with limited power capacity. Therefore, low-cost, low-power sensor systems have inevitably gained attention. In this thesis, sensing transducers for air pressure, relative humidity, and acceleration using a cheap standard CMOS process, low-power high-resolution sensor interfaces, and a clock generator were proposed. First, a 0.66pJ/c.step energy-efficient dual quantization-based capacitance-to-digital converter (CDC) based on dual quantization was proposed. The proposed CDC achieves high linearity of 11.8bits due to the single-bit quantization, and simply reduces the quantization noise without a high-order DS modulator. Next, a 3.68aFrms ultra-high resolution continuous-time (CT) bandpass (BP) delta-sigma (DS) CDC was proposed in order for full-CMOS sensors whose transducer have low sensitivity compared to MEMS counterpart. A CT DS operation is employed to overcome thermal noise limitation and BP DSM architecture is employed to save power consumption. The proposed CTBP DS CDC achieves the highest FoMS of 183dB which is more than 2x improvement over the recent stat-of-the-art CDCs. For a clock generator, a 43nW ultra-low power temperature compensated crystal oscillator (TCXO) with high frequency accuracy of $\pm$5ppm from -20$^o$C to 85$^o$C was proposed.