Research And Design On High CMRR Delta-sigma Modulator For Neural Interface

With the development of medical technology and neuroscience,bioelectrical signals have become an indispensable means of diagnosing and treating patients.As an important category of human bioelectric signals,neural signals contain a wealth of information about brain neuroelectric activity.The recording of neural signals is of great significance to brain research,physiological research and clinical brain disease diagnosis.However,most hardware solutions available for neural recordings are primarily wired or tethered setups with bulky rack-mounted electronics.The patient has wires coming out of their head or chest cavity through percutaneous connections.These wires carry the neural signals from the implanted electrodes to the external electronics that process these signals.Due to percutaneous connections,such a setup presents a significant risk of infection,limited activities of patients and limited time for diagnosis and treatment.In order to overcome these limitations,miniaturization,fully implantable,and wireless devices are important development directions for neural signal recording systems in the future.Although neural recordings are invaluable,neural stimulation is necessary to administer therapy in patients suffering from neurological ailments that do not respond to pharmacological treatment.Neural stimulation is performed by sending controlled current pulses into specific regions of the brain to modulate brain activity.In the closed-loop neural recording systems formed by the neural recording and neural stimulation,the neural stimulation will generate large stimulation artifacts at the recording sites.Therefore,there are many performance requirements for the analog front end,such as high input impedance,high common-mode rejection ratio,and low power consumption.This paper presents a continuous-time ?? modulator combined with the instrumentation amplifier,which is suitable for the front-end of neural signal recordings.Firstly introduce the characteristics of neural signals,including amplitude characteristics and frequency domain characteristics,introduce the model of the neural recording system,analyze the artifacts that exists during recording,and determine the performance of analog front-end.Secondly,a neural recording system composed of an instrumentation amplifier and a continuous-time ?? modulator is proposed.The key technology used in the instrumentation amplifier is elaborated,including chopping modulation technology,ripple cancellation technology,and positive feedback loop to reduce the front-end DC offset and 1/f noise of the amplifier.Common-mode replication technology to improve the input impedance and common-mode rejection ratio of the amplifier.The basic principle of the continuous-time ?? modulator is introduced,and a variety of discrete time equivalent conversion methods to the continuous time are introduced.The non-ideal factors of the continuous-time ?? modulator,such as excess loop delay and non-linearity of integrator and clock jitter are discussed.Use the SIMULINK tool in MATLAB to perform behavior of modulator,and calculate the values of various parameters in the modulator according to the impulse-invariant transformation.The first-order noise coupling in the digital domain is used to improve the order of noise shaping,and the DWA algorithm is used to improve the linearityof the feedback DAC.The proposed circuit was implentmented on standard CMOS 0.18?m process.Under the supply voltage of 1.8V,the minimum CMRR of instrumention amplifier at 100 Hz achieves 141 d B.For the 361 Hz input signal,the CT ??modulator exhibits a SNRD of 93.1dB and an ENOB of 15.1bit.