| Title |
A Neural Recording Analog Front-end IC with Low-noise Folded-cascode OTA |
| Authors |
전대영(Daeyoung Jeon) ; 유효준(Hyojun Yoo) ; 홍순성(Soonseong Hong) ; 박보성(Boseong Park) ; 차혁규(Hyouk-Kyu Cha) |
| DOI |
https://doi.org/10.5573/ieie.2025.62.5.18 |
| Keywords |
Analog front-end IC; Folded cascode OTA; Common-mode cancellation loop; Chopper stabilization |
| Abstract |
This paper presents a low-power, low-noise neural recording system designed to record local field potential (LFP) and action potential (AP) neural signals. The proposed system consists of a low-noise amplifier (LNA) and a programmable gain amplifier (PGA). A chopper stabilization circuit is employed in the LNA to suppress flicker noise in the low-frequency band and a common-mode cancellation loop (CMCL) is added to eliminate common-mode artifacts generated by the stimulator in bidirectional neural interfaces. The proposed LNA utilizes a folded-cascode operational transconductance amplifier with boosted input transconductance to improve the input-referred noise (IRN) performance. The PGA offers gain adjustments of 1, 2, 3, 4 times in addition to the LNA’s 40 dB gain. To record LFP and AP signals simultaneously or separately, the bandwidth of the AFE is designed to be tunable. The simulated IRN performance in the LFP band (1-300 Hz) is 0.4 μVrms, 1.29 μVrms in the AP band (300-10 kHz), and 1.35 μVrms across the combined LFP and AP bands. Within 1% of total harmonic distortion, the CMCL circuit is able to cancel up to 1-VP-P of artifact. The proposed AFE is designed using 180 nm CMOS process and consumes 2.81 μW at 1-V voltage. |