Fellow, affiliation:
02/2023-01/2025: Analogue Circuits and Image Sensors, Uni Siegen
Design and analyses of a scalable power delivery network for human sensing application with novel on-chip harmonic mitigation circuit.
The past decade has witnessed tremendous progress in sensors and sensing technologies. Sensing technologies have ubiquitous applications in the field of the design of human-machine interface development. These designs are compact, low-voltage and require extensive computational efforts; so that a two-way communication/interaction between humans and machines is possible in an efficient manner. The sensing and interface devices require a signal processing circuitry, which consists of analog, digital and mixed-mode circuits to transfer information in a readable form. The performance of the signal processing circuits largely depends on the quality of the power supply voltages. In today’s deep-micron technologies, maintaining the quality of voltage in the presence of internal and external factors such as nonlinear distortions, crosstalk, high switching activities, etc., is a very challenging task for designers. This work aims to design a robust, efficient and scalable power delivery network (PDN) to increase the on-chip battery life for sensing applications. An efficient and scalable (supply voltage) PDN will be designed in this work, including a novel harmonic calibration circuit. The work also includes the analysis of the nonlinear behaviour of the PDN. Based on the harmonics analysis, a harmonic calibration circuit and variable negative impedance circuit will be designed that mitigates the harmonic values at the output of a PDN. Therefore, the AC fluctuations present at the output of a PDN can be reduced, which eventually enhance the performance of the sensing circuits. The work will also include the design of a differential bootstrapped driver circuit that will help to negate the effect of harmonics due to supply fluctuations in a PDN. Further, the work includes studying the effects at a system level and will determine the parameters degrading the desired output in a closed-form equation.