This paper presents a background calibration method assisted with noise-shaping, for a 10-b bridged SAR ADC. It proposes calibrating the mismatches from the MSB capacitors using the LSBs. First-order noise-shaping has been employed to facilitate the calibration as well as the analog-to-digital conversion. Since noise-shaping is able to shape both the comparator input and quantization noise, it is expected to improve the calibration effects as well as the SNR, for the SAR ADC. Measurement results show, upon the proposed calibration, the SAR’s INL and SNR are improved by 3$$LSB and 9$$dB, respectively, compared to when the noise shaping is off. Fabricated using a 55$$nm CMOS MOMCAP technology and tested with a −2.5$$dB sinewave input, the bridged SAR ADC achieves an SNR of 62.2$$dB and an INL better than 1$$LSB, when consuming a power of $72\mu \mathrm{\text{W}}$, at a supply voltage of 1.2$$V and sampling frequency of 50$$kHz. The ADC’s active area is $300\times 380\mu {\mathrm{\text{m}}}^2$.

本文提出了一种用于 10-b 桥接 SAR ADC 的辅助噪声整形的背景校准方法。它建议使用 LSB 来校准 MSB 电容器的失配。采用一阶噪声整形来促进校准以及模数转换。由于噪声整形能够整形比较器输入和量化噪声，因此有望改善 SAR ADC 的校准效果以及 SNR。测量结果表明，经过所提出的校准后，SAR的INL和SNR提高了3$$LSB 和 9$$分别与噪声整形关闭时相比。使用 55 制造$$nm CMOS MOMCAP 技术并使用 −2.5 进行测试$$dB 正弦波输入，桥接 SAR ADC 的 SNR 达到 62.2$$dB 和 INL 优于 1$$LSB，当消耗功率时$72\mu \mathrm{\text{瓦}}$，电源电压为 1.2$$V 和采样频率为 50$$千赫。ADC 的有效区域为$300\times 380\mu {\mathrm{\text{米}}}^2$。