How do I extract a SPICE model from an EM simulation for use in a circuit simulator?
SPICE Model Extraction from EM Data
SPICE model extraction bridges the gap between electromagnetic simulation (which provides accurate frequency-domain data) and circuit simulation (which requires time-domain compatible models for transient analysis and mixed-signal simulation).
- Performance verification: confirm specifications against the application requirements before finalizing the design
- Environmental factors: temperature range, humidity, and vibration affect long-term reliability and parameter drift
- Cost vs. performance: evaluate whether the application demands premium components or standard commercial grades
Frequently Asked Questions
How do I ensure the SPICE model is passive?
A passive model never generates energy (the real part of the impedance is non-negative at all frequencies). Vector Fitting generates a model that may violate passivity at some frequencies due to numerical fitting errors. Passivity enforcement algorithms (included in commercial tools like Keysight ADS, Cadence Sigrity, and ANSYS Q3D) iteratively adjust the model parameters to ensure passivity while minimizing the fitting error. A non-passive model causes: SPICE convergence failures, artificial energy generation in transient simulation, and physically incorrect results.
What frequency range should I simulate?
Simulate from DC (or the lowest frequency of interest) to at least 3-5× the highest signal frequency. For digital signals: the highest significant frequency is approximately 0.5/t_rise (where t_rise is the signal's 10-90% rise time). For a 1 ns rise time: simulate to at least 1.5-2.5 GHz. For RF signals: simulate to at least 3× the carrier frequency to capture the harmonics generated by nonlinear devices. The S-parameter data must cover the entire frequency range for the rational fit to produce an accurate broadband model.
Can I extract a SPICE model for a multi-port structure?
Yes. Multi-port structures (transmission line crossings, power distribution networks, multi-pin connectors) generate N-port S-parameter data (N = number of ports). The rational fitting algorithm handles multi-port data by fitting all S_ij parameters simultaneously, maintaining the coupling between ports. The resulting SPICE model has N terminals and reproduces the full S-parameter matrix over the fitted frequency range. For large port counts (N > 20): the model size can be very large (thousands of elements), which may slow the SPICE simulation. Model order reduction techniques can compress the model while maintaining accuracy.