This paper focuses on the assessment of using programmable gain amplifiers on fast image sensor devices targeting low-light noise performance, comparing the responses of three on-chip column gain amplifier structures. Subsequent conclusions are derived from the usage of column amplifiers for imagers, retrieved from the comparison work. Based on the conducted study and the fabricated silicon device outputs, conclusive directions are pointed to determine the most optimal actions to adopt on future low-noise CMOS imagers. This work has revealed that the usage of amplification stages are essential for image sensors targeting low-noise, employing fast high-order incremental sigma-delta converters. Furthermore, the newest characterization results are also presented, done at a higher pixel supply voltage level. The measured performance improved from previous and early characterization, exhibiting 0.77% overall system non-linearity and featuring 2.91e-input-referred noise for the total readout chain noise at the same region of interest in the earlier work. Increasing the pixel supply has led to superior overall sensor performance.

本文着重评估在针对低光噪声性能的快速图像传感器设备上使用可编程增益放大器的性能，并比较了三种片上列增益放大器结构的响应。从比较工作中获得的结论是从成像器的列放大器的使用中得出的。根据进行的研究和制造的硅器件输出，指出最终的方向，以确定在未来的低噪声CMOS成像器上要采用的最佳方案。这项工作表明，使用快速的高阶增量sigma-delta转换器，放大级的使用对于以低噪声为目标的图像传感器至关重要。此外，还提供了最新的表征结果，这些结果是在较高的像素电源电压电平下完成的。测量的性能较之前和早期的特性有所提高，表现出0.77％的整体系统非线性度，在较早的研究中，相同关注区域的总读出链噪声具有2.91e输入参考噪声。像素供应的增加带来了卓越的整体传感器性能。