针对人体内神经电信号非常微弱、噪声大、环境干扰大等特点，研究与设计了一款应用于神经信号采集的高电源抑制比（ PSRR ）和共模抑制比（ CMRR ）的低噪声植入式模拟前端 . 该模拟前端采用全差分结构来实现模拟前端中的前置放大器、开关电容滤波器及可变增益放大器，使得电路具有较好的电源抑制比和共模抑制比；采用斩波调制技术来抑制电路的低频噪声，并通过带电流数模转换器（ DAC ）的纹波抑制环路来抑制前置放大器的输出纹波，从而使该模拟前端在具有高 PSRR 和 CMRR 的同时能保持低噪声性能 . 文中采用 0.18μm CMOS 工艺设计该模拟前端芯片，版图后仿真结果表明，该模拟前端在
0.1Hz~10kHz 内的等效输入噪声为 2.59μV ， 实现了 46.35 、 52.18 、 60.02 、 65.95dB 可调增益， CMRR 和 PSRR 分别可达146 及 108dB ，很好地满足了植入式神经信号采集的要求

As the in vivo electroneurographic signals are weak with high noise and strong environment interference ， an implantable low-noise analog front-end with high power supply rejection ratio （ PSRR ） and com-mon mode rejection ratio （ CMRR ） for neural signal acquisition is proposed. Fully differential structure is used to design the preamplifier ， switch capacitor filter and variable gain amplifier of the analog front-end. Chopper stabilization technology is used to reduce the low-frequency noise and a ripple reduction loop with current DAC is used to reduce the output ripple. In this way ， the proposed analog front-end can achieve high PSRR and CMRR without increasing the original low noise. Moreover ， a front-end chip is designed via 0.18μm CMOS technology and is used for simulation. The results of layout show that （ 1 ） input-referred noise of the analog front-end is 2.6μV at the interval of 0.1Hz~10kHz ；（ 2 ） the gain of the analog front-end can switch among 46.35 ， 52.18 ， 60.02 and 66.95dB ； and （ 3 ） the CMRR and PSRR of the analog front-end are respectively 146 and 108 dB. Thus ， it draws the conclusion that the proposed analog front-end is suitable for neural signal acquisition.