Low-noise low-power front-end logarithmic amplifier for neural recording system

Abstract

In this work, a low-power, low-noise logarithmic preamplifier for biopotential and neural recording application is presented. The amplifier is based on a linear limit logarithmic amplifier technique, and an active filter as a DC cancellation filter has been included to its input in order to eliminate DC offsets, which are produced at the electrode-tissue interface. This system has been simulated in a UMC standard 90-nm 1P9M CMOS process. Five dual gain stages are used to produce the required linear limit logarithmic amplifier. The dynamic range of the amplifier is measured to be 48 dB which covers the signals with amplitude from 20 μV to 5 mV. The amplifier consumes 23.5 μW from a 1.2-V power supply and has a maximum gain of 69.8 dB. The simulated input referred noise is 5.3 μV over 0.1 Hz to 20 kHz. Copyright © 2012 John Wiley & Sons, Ltd. Due to problems caused using current systems, a novel method has presented. High-pass filter has been used to eliminate electrode offsets. Logarithmic amplifier has been used to amplify the neural signals. © 2012 John Wiley & Sons, Ltd.