@inproceedings{Luan:2017:10.1109/BioCAS.2016.7833817,
author = {Luan, S and Liu, Y and Williams, I and Constandinou, TG},
doi = {10.1109/BioCAS.2016.7833817},
pages = {404--407},
publisher = {IEEE},
title = {An Event-Driven SoC for Neural Recording},
url = {http://dx.doi.org/10.1109/BioCAS.2016.7833817},
year = {2017}
}

Download

TY  - CPAPER
AB  - This paper presents a novel 64-channel ultra-low power/low noise neural recording System-on-Chip (SoC) featuring a highly reconfigurable Analogue Front-End (AFE) and block-selectable data-driven output. This allows a tunable bandwidth/sampling rate for extracting Local Field Potentials (LFPs)and/or Extracellular Action Potentials (EAPs). Realtime spike detection utilises a dual polarity simple threshold to enable an event driven output for neural spikes (16-sample window). The 64-channels are organised into 16 sets of 4-channel recording blocks, with each block having a dedicated 10-bit SAR ADC that is time division multiplexed among the 4 channels. Eachchannel can be individually powered down and configured for bandwidth, gain and detection threshold. The output can thus combine continuous-streaming and event-driven data packets with the system configured as SPI slave. The SoC is implemented in a commercially-available 0.35u m CMOS technology occupying a silicon area of 19.1mm^2 (0.3mm^2 gross per channel) and requiring 32uW/channel power consumption (AFE only).
AU  - Luan,S
AU  - Liu,Y
AU  - Williams,I
AU  - Constandinou,TG
DO  - 10.1109/BioCAS.2016.7833817
EP  - 407
PB  - IEEE
PY  - 2017///
SP  - 404
TI  - An Event-Driven SoC for Neural Recording
UR  - http://dx.doi.org/10.1109/BioCAS.2016.7833817
UR  - http://hdl.handle.net/10044/1/39163
ER  - 

Download