Executive Summary : | Modern cellular/wireless receivers operate in multiple standards and bands, making analog-to-digital converters (ADC) essential. Conventional ADCs can cause "aliasing," where frequency components around the multiples of the sampling frequency get down-converted to the desired signal bandwidth, potentially corrupting or drowning the desired signal. An anti-alias filter (AAF) is needed to tackle this issue, but it has strict specifications due to strong blocker signals and multi-band operation. A potential solution to aliasing is to sample the signal at non-uniform time intervals. By sampling and digitizing the signal only when it crosses specific threshold voltages, a signal-dependent or level-crossing sampling can be done, avoiding aliasing. These non-uniform digital samples can be fed to a digital filter to filter out-of-band components before converting them to uniform samples. This digital AAF can be readily programmed and reconfigured, relaxing or even eliminating the need for an analog AAF. Non-uniform sampling (NUS) ADCs have gained traction among ADC designers, but they consume more than 50% of the total power. Reducing the number of comparators in NUS ADCs will greatly reduce power consumption of the digital section and the ADC overall. This work aims to investigate two NUS ADC architectures that reduce the number of comparators needed and improve the power efficiency of the NUS ADC, making it an excellent choice for multi-band wireless receivers. |