A High CMRR Instrumentation Amplifier Employing Pseudo-Differential Inverter for Neural Signal Sensing

The next generation of the wearable and implantable high-density neural signal sensing systems must be multi-channel and power efficient. The integrated instrumentation amplifier (IA) within such sensing system forms the critical element for signal detection and pre-processing. Inverter-based IA provides a low power consumption and a small die area. However, the conventional inverter yields a single-ended (SE) topology, which suffers from a poor common-mode rejection ratio (CMRR). In this paper, a pseudo-differential inverter is proposed. It provides an alternative solution to increase CMRR without the need for external control circuit. The pseudo input terminal is simply employed by equally splitting the input transistor of SE inverter into two and connected in parallel (reuses the active load), to provide a feedback path to dynamically reduce the common-mode interference. With respect to the SE topology, CMRR is much improved (approximately 75% higher) without any significant increasing of die area and power consumption. The proposed IA provides circa 117 dB CMRR, 1.5
$\text{G}\pmb \Omega $
input impedance at 100 Hz, and consumes
$8.3~\pmb \mu \text{W}$
.