Low-noise Image Sensors with Shifted Pseudo-correlated Multiple Sampling Method

Abstract

This paper presents a low-noise image sensor with a shifted pseudo-correlated multiple sampling (S-PMS) method based on a conventional single-slope analog-to-digital converter (SS-ADC) structure. By shifting the reference voltage of the ramp signal, a single ramp generator acts as a dual ramp generator, leading to reduced area and power consumption. In addition, by dividing the two illumination regions and performing PMS, the analog-to-digital conversion time can be reduced by half, resulting in an increase in frame rates. A prototype SS-ADC with S-PMS for a low-noise image sensor was fabricated using a 28 nm process. The results showed that the proposed S-PMS increased by 49% of frames per second and reduced random noise by 53%. © 2024 IEEE.