Efficiency-Centric Automatic Circuit Design for Feed-Forward Operational Amplifier

This study introduces a novel efficiency-centric automatic circuit sizer, integrating advanced strategies to optimize evolutionary algorithm (EA) based sizing. By combining Latin Hypercube Sampling, optimization focused on the nominal operating corner, and periodic worst-corner selection, this framework significantly increases the quality of the resulting solutions. The method is adaptable, extending beyond traditional EAs to incorporate ML-enhanced approaches. Applied to a feed-forward operational amplifier for automotive use, our integrated method delivers circuit configurations with up to 46% performance improvement over its individual components, all within the same simulation budget.