This thesis presents innovations for a special class of circuits called Time Difference (TD) circuits. We introduce a signal processing methodology with TD signals that alters the target signal from a magnitude perspective to time interval between two time events and systematically organizes the primary TD functions abstracted from existing TD circuits and systems. The TD circuits draw attention from a broad range of application fields. In addition, highly evolved complementary metal-oxide-semiconductor (CMOS) technology suffers from various problems related to voltage and current amplitude signal processing methods. Compared to traditional analog and digital circuits, TD circuits bring several compelling features: high-resolution, high-throughput, and low-design complexity with digital integration capability. Further, the fabrication technology is advancing into the nanometer regime; the reduction in voltage headroom limits the performance of traditional analog/mixed-signal designs. All-digital design of time-difference circuit needs to be stressed to adapt to the low-cost, low-power, and high-portability applications. We focus on Time-to-Digital Converters (TDC), one of the crucial building blocks in TD circuits. A novel algorithmic architecture is proposed based on a binary search algorithm and validated with both simulation and fabricated silicon. An all-digital structure Time-difference Amplifier (TDA) is designed and implemented to make FPGA and other all-digital implementations for TDC and related TD circuits feasible. Besides, we propose an all-digital timing measurement circuit based on the process variation from CMOS fabrication: PVTMC, which achieves a high measurement resolution: $