A Hybrid 1st/2nd-Order VCO-Based CTDSM With Rail-to-Rail Artifact Tolerance for Bidirectional Neural Interface

Abstract

Bi-directional brain machine interfaces enable simultaneous brain activity monitoring and neural modulation. However stimulation artifact can saturate instrumentation front-end, while concurrent on-site recording is needed. This brief presents a voltage controlled-oscillator (VCO) based continuous-time <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\rm \Delta \Sigma $ </tex-math></inline-formula> modulator (CTDSM) with rail-to-rail input range and fast artifact tracking. A hybrid <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$1^{st}/2^{nd}$ </tex-math></inline-formula> -order loop is designed to achieve high dynamic range (DR) and large input range. Stimulation artifact is detected by a phase counter and compensated by the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$1^{st}$ </tex-math></inline-formula> -order loop. The residue signal is digitized by the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$2^{nd}$ </tex-math></inline-formula> -order loop for high precision. Redundancy between the two loops is implemented as feedback capacitors elements with non-binary ratio to guarantee feedback stability and linearity. Fabricated in a 55-nm CMOS process, the prototype achieves 65.7dB SNDR across 10 kHz bandwidth with a full scale of 200 mV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">pp</sub> . And a ±1.2 V input range is achieved to suppress artifacts. Saline based experiment with simultaneous stimulation and recording demonstrates that the implemented system can track and tolerate rail-to-rail stimulation artifact within 30 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu \text{s}$ </tex-math></inline-formula> while small neural signal can be continuously monitored.