传感器常用于将外界的各种物理信号、化学信号和生物信号等转化为电流或电压等模拟电参量。即使在集成电路高度集成化和数字化的发展趋势下，传感器作为连接自然环境和机器的桥梁，依旧发挥着不可替代的作用。为了能更好地识别传感器产生的电信号，高精度的模拟前端电路常被用于对信号进行放大与滤波，使之通过模数转换器后被精确地转换为数字信号并被微控制单元所识别。在大数据和人工智能等信息技术的潮流下，工业监测、自动驾驶、医疗电子和航空航天等领域对外界环境的信息获取需求愈加扩大，然而这些应用场景为模拟前端电路带来了巨大的挑战：（1）芯片要求具有更强壮的鲁棒性来应对各种恶劣的工作环境；（2）应用领域的扩展对噪声的要求越来越高，并要求电路在不同工作条件下依旧保持高精确度；（3）传感器的发展要求电路具有更高的输入动态范围，从而识别更宽范围的信号。基于这个背景，本论文针对模拟前端电路做了以下研究：第一，本论文针对输出电流信号的传感器设计了一款跨阻放大器芯片。该跨阻放大器芯片由跨阻放大器、直流失调抑制电路和可编程增益放大器阻成。跨阻放大器采用积分器-微分器级联的结构实现；直流失调抑制电路采用积分器实现，防止第一级电路的输出端因直流输入成分而饱和，影响了整个电路的正常工作；同时本论文提出了一中新型具有工艺-电压-稳度稳定性的高线性GΩ等效电阻电路，避免了片外电阻的使用。PGA由三级电阻反馈放大器和五阶巴特沃斯有源RC滤波器实现。第二，本论文针对输出电压信号的传感器设计了一款斩波电容耦合仪表放大器芯片。斩波技术可以消除电路的1/f噪声和输入失调电压，电路的精度完全由放大器的热噪声决定。针对放大器的输入失调电压因斩波开关而在输出端产生幅度很大的波纹，本论文提出一种新型的双路波纹衰减环路结构，极大程度地抑制了波纹的幅度；与此同时，对于波纹衰减环路给系统带来的额外噪声，本论文提出了一种噪声-非线性抑制环路结构消除这部分噪声。该仪表放大器放大器具有更低的功耗和噪声效率因子。整个电路能在各种工作条件下保持鲁棒性的同时保持高精度的特点。
Sensors are usually used to convert external physical signals, chemical signals and biological signals into analog electrical parameters, such as current or voltage. Even in the integrated circuit highly integrated and digital development trend, the sensor as a bridge between the natural environment and the machine, still plays an irreplaceable role. In order to better detect the electrical signal generated by sensors, high-precision analog front-end circuits are often used to amplify and filter the signals, so that it can be accurately converted into digital signal through the analog-to-digital converter and recognized by microcontroller units. With the trend of information technology such as big data and artificial intelligence, industrial monitoring, automatic driving, medical electronics, aerospace and so on, have an increasingly expanding demand for information acquisition of the external environment. However, these application scenarios have brought huge challenges to the simulation of analog front end circuits: (1) The chips are required to have more stronger robustness to apply in more harsh working environment; (2) With the expansion of application field, the requirement of noise becomes higher, and the analog front end circuits are required to maintain high accuracy under different working conditions; (3) The development of sensors requires circuits with a higher input dynamic range to recognize a wider range of signals. Based on this background, this paper has done the following research on the AFE circuits:First, a transimpedance amplifier chip is proposed for the sensors of output current signal. It is composed of a transimpedance amplifier, a DC offset cancellation circuit and a programmable gain amplifier. The transimpedance amplifier is based on an integrator followed by a differentiator configuration; DC offset cancellation circuit is realized by integral circuit, which prevents the output end of the first stage circuit from being saturated due to DC current and affects the normal operation of the whole circuit; At the same time, one new GΩ equivalent resistance circuit with high process-voltage-temperature and linearity is put forward to avoid the use of off-chip resistor. The programmable gain amplifier is consisted of three stages of resistive feedback amplifier and a fifth-order Butter worth filter with active RC topology.Second, a novel chopper capacitively-coupled instrumentation amplifier chip is designed for the sensors of output voltage signal. Chopper technique can eliminate the 1/f noise and input offset voltage of amplifier, while the accuracy of the circuit is completely determined by the thermal noise of the amplifier. Due to input offset voltage of amplifier transferring a large ripple at the output with the function of chopping switch, a novel dual ripple reduction loop structure is proposed to strongly inhibit the amplitude of the ripple; In addition, a noise-nonlinearity cancellation loop topology is proposed to eliminate the additional noise caused by the dual ripple reduction loop structure. The instrumentation amplifier has lower power consumption and noise efficiency factor. The whole chip can maintain the characteristics of high precision and robustness under various working conditions.
