A self-calibration circuit for a neural spike recording channel

This paper presents a self-calibration circuit for a neural spike recording channel. The proposed design tunes the bandwidth of the signal acquisition Band-Pass Filter (BPF), which suffers from process variations corners. It also performs the adjustment of the Programmable Gain Amplifier (PGA) gain to maximize the input voltage range of the analog-to-digital conversion. The circuit, which consists on a frequency-controlled signal generator and a digital processor, operates in foreground, is completely autonomous and integrable in an estimated area of 0.026mm², with a power consumption around 450nW. The calibration procedure takes less than 250ms to select the configuration whose performance is closest to the required one.