Novel Low Noise Amplifier for Neural Signals Based on STT-MTJ Spintronic Device

Low noise amplifier is generally adopted in brain-machine interface (BMI) for the purpose of the extraction of action potentials of neural signals from the Local Field Potentials (LFPs). However, the frequency ranges of these signals are very near to those of the LFPs and other existing noises. So the matter brings us difficulties in order to designing of the amplifier. The purpose of this paper is concentrated on designing of a low noise amplifier specific for the neural signals. Based on the STT-MTJ spintronic devices, the amplifier selectively amplifies the high amplitude peaks, while suppresses the low amplitude peaks. In this way, the novel amplifier can be used to amplify the useful neural signals not only based on the frequency but also on the amplitude. At the same time, the noise signals are suppressed by the amplifier. The low cut-off frequency of the amplifier is adjusted to be 1.24 Hz. The gain of the amplifier is adjusted to be 39.8 dB from 1.24 Hz to 9.1 KHz. In this bandwidth, the input-referred noise is $33.7~{\mathrm {nV}} /\sqrt {\mathrm {Hz}} $ and the RMS noise is $1.9~\mu $ Vrms. This novel amplifier can have potential application in the field of neural signal recording and other related fields.