Research On 24-channel High-Precision Brain-Computer Interface Analog Front-end Chip

With the rapid development of integrated circuit technology,wearable medical electronic devices play an increasingly important role in people's daily lives.The emergence of new medical treatment methods such as telemedicine and smart hospitals can greatly facilitate human life and improve human health.For wearable medical electronic devices,it mainly diagnoses human health by acquiring bioelectrical signals from the human body,and EEG signals can often provide doctors with more information about human health,so it is critical.However,since EEG signals are weaker than other bioelectrical signals and are easily susceptible to environmental factors,the integrated circuit technology for acquiring EEG signals has become a research hotspot.EEG signals are low in frequency and weak in amplitude.It places higher requirements on the noise performance,chip area,and power consumption of the analog front-end circuit.During the EEG signal acquisition process,it is extremely susceptible to power line interference and cable motion artifact interference in the environment.The impedance deviation caused by the multi-channel bio-electrode's contact with the human body will cause signal distortion.With the increase of the number of channels,the power consumption and heat generation of a single channel of the system are strictly limited.These factors bring great challenges to the realization of high-performance EEG signal acquisition front-end..This article first describes the characteristics of EEG signals and the selection of front-end electrode types.The use of chopper and source resistance degradation technology improves the noise performance of the system without increasing power consumption.Improve the traditional input impedance improvement technology,and use the substrate modulation technology to increase the amplifier's common mode input range under low voltage conditions.Improve the connection method of traditional pseudo-resistors,improved amplifier stability in different processes.In addition,the traditional method of using a single right-leg drive loop to suppress power lines interference is improved and combined with the current common-mode feedforward technology to further improve the common-mode rejection ratio.This thesis implements a low power and low noise EEG signal acquisition front-end.The low-power biological front-end circuit architecture proposed in this thesis includes a DCcoupled low-noise chopper amplifier,an ultra-low cut-off frequency low-pass filter,a variable gain amplifier with high-pass filtering,and a low-power CT Delta-Sigma modulator.And use the improved right-leg drive loop to suppress power line interference in the environment.The use of a four-input fully differential amplifier,substrate modulation technology and a new class AB output stage greatly improves the performance of the low noise chopper amplifier.Based on the SMIC 180nm CMOS process,the layout and circuit design were carried out,and the tape was completed.The front-end low-noise amplifier has an integrated noise of 0.96 ?Vrms at 1-100 Hz,an input impedance of more than 10 G? at 100 Hz,and can tolerate ± 450 mV electrode offset.The effective bandwidth of the CT DeltaSigma modulator is 500 Hz.When the sampling frequency is 640 kHz,a 183 Hz sinusoidal signal is input to achieve an ENOB of 13.05 bit.The overall power consumption is 56 ?W,which Delta-Sigma modulator consumes 40 ?W.