A Sub-μW Tuneable Switched-Capacitor Amplifier-Filter for Neural Recording using a Class-C Inverter

A two stage sub-μW Inverter-based switched-capacitor amplifier-filter is presented which is capable of amplifying both spikes and Local Field Potentials (LFP) signals. Here we employ a switched capacitor technique for frequency tuning and reducing of 1/f noise of two stages. The reduction of power consumption is very necessary for neural recording devices; however, in Switched Capacitor (SC) circuits OTA is a major building block that consumes most of the power. Therefore an OTA-less technique utilizing a classC inverter is employed that significantly reduces the power consumption. A detailed analysis of noise performance for the inverter-based SC circuits is presented. A mathematical model useful for analysis of such SC integrators is derived and a good agreement has been obtained between simulation and analytical results. With a supply voltage of 0.7V and using 0.18 μm CMOS technology, this design can achieves a power consumption of about 538 nW. The designed amplifier-filter has the gains 18.6 dB and 28.2 dB for low pass only and cascaded filter, respectively. By applying different sampling frequencies, the filter attains a reconfigurable bandwidth.