What problem does it solve?
Numerical cores for discrete-time circuit simulations require robust nonlinear solving and stable time integration. This skill provides guidance on using implicit methods (trapezoidal, backward Euler, GEAR/BDF), Newton-Raphson convergence handling, and sparse linear algebra using Eigen or UMFPACK to debug slow or oscillatory behavior in stiff circuits.
Core Features & Use Cases
- Implicit integration options (trapezoidal, backward Euler, GEAR/BDF) for stiff, switching, and nonlinear networks.
- Newton-Raphson iteration with convergence checks and damping strategies to ensure reliable stepping.
- Sparse linear solver guidance (Eigen, UMFPACK) and topology-change factorization reuse to optimize performance.
- Diagnostic routines for convergence, step-size selection, and accuracy validation with history states.
Quick Start
Run a time-domain simulation with implicit integration to observe nonlinear MNA circuit behavior under switching events.