The rapid advancement of wearable electronics in smart textiles has driven the need for highly efficient and precise timing solutions. Traditional crystal oscillators, while accurate, are too bulky for seamless integration into fabric-based computing environments. My research addresses this challenge by designing a tunable, ultra-low power on-chip oscillator, specifically optimized for real-time synchronization in multi-chip fabric-integrated systems.

This work introduces a current-starved ring oscillator (CSRO) running in sub-threshold weak inversion, achieving an exceptionally low power consumption of 170 femtojoules per cycle. Operating at 1.1V, the oscillator provides a frequency range of 88.35 kHz to 108.7 kHz with an ultra-fine tuning step size of 80 Hz (0.08%), ensuring precise clock adaptability for dynamic sensing and processing tasks.

Designed for integration within ARM Cortex M0+ based System-on-Chip (SoC) architectures, the oscillator enables efficient multi-chip synchronization, ensuring timing accuracy within ±100 microseconds across a network of embedded sensors. This solution is a significant step forward for next-generation textile computing, offering a compact, energy-efficient, and scalable clocking solution for real-time applications.

Thesis Defence Slide: Ms_presentation_Final.pptx 

Overview:

A Flash Analog-to-Digital Converter (ADC) is the fastest ADC architecture, utilizing a parallel conversion technique to achieve extremely high-speed data acquisition.

Working Principle:

Uses 2ⁿ - 1 comparators for an n-bit ADC. A resistor ladder generates reference voltages. Each comparator compares the input voltage (V_in) with its reference voltage. The outputs of the comparators form a thermometer code, which is then converted into a binary output using an encoder.

Key Components: Resistor Ladder, Comparators (2 stage opamp has been used here as comparator), Thermometer-to-Binary Encoder