An Amplitude and Slope Based Nonlinearity Calibration Algorithm for LC-ADC

Level-crossing analog-to-digital converter (LC-ADC) has attracted widespread attention because of its eventdriven characteristic which boosts the energy efficiency by shrinking the amount of the data to be handled by the digital backend. However, the resolution of the LC-ADC is often less than 10 bits due to nonlinearities such as comparator offset and DAC mismatch. This paper proposes a calibration algorithm to compensate those various nonlinearities. It utilizes the signal amplitude and slope information to avoid too much storage of calibration parameters. The algorithm is verified by the post-layout simulation of a LC-ADC showing over 10dB signal-to-noise distortion ratio (SNDR) improvement. Finally, a LC-ADC with over 70 dB SNDR and 4 ksps equivalent Nyquist sampling rate is achieved drawing $8.3 \mu \mathrm{~A}$ current from 1.8V supply.