Design and Analysis of a Fifth Order Low Pass G m -C Filter for Seizure Detection

Electroencephalogram (EEG) signals have weak amplitude and have susceptibility to noise from supplies, electrodes and added sources. The preamplifier in analog front end of a neural acquisition system will magnify the signal at low power and low noise, and a low pass filter will restrict the bandwidth. The low power methods are necessary to reduce heat dissipation; otherwise, they will damage the surrounding tissues. These techniques can reduce the use of bulk sized batteries, which will increase the battery lifetime and can avoid the distress of frequent battery replacements. In this article, the focus is given on the design of an OTA-C filter, with challenges in reducing power of the circuit without compromising the noise and bandwidth parameters. These challenges are addressed in the proposed work with comprehensive analysis. The low frequency circuits necessitate huge capacitors and resistors, which occupy large chip-area, but the requirement is for reduced area. The OTA-C filters are widespread because they are fundamentally tunable, resistor-less and can be engineered on small chip areas. The novelty of this work is to reduce power and to create a higher order G m -C topology filter that can record FRs or high focal oscillations in order to record epileptic signals. The proposed design implements current division and source degeneration techniques to bring down the transconductances in nano scale. The design achieves a mid-band gain of 54.2 dB and a − 3 dB bandwidth in the range 40–650 Hz, with simulations performed in 180 nm SCL Cadence environment.