What is the recommended mesh size for a Sonnet planar EM simulation of a microstrip filter?
Sonnet Mesh Size for Microstrip Filter Simulation
Sonnet's fixed-grid approach provides guaranteed convergence (finer grid always gives a more accurate result) but requires more cells than adaptive-mesh solvers for the same accuracy. The mesh size is the single most important parameter in a Sonnet 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 does cell size affect simulation time?
Sonnet's simulation time scales approximately as N^2 to N^3, where N is the total number of cells. Halving the cell size quadruples the number of cells (in 2D), which increases the simulation time by approximately 8-16×. For a 10 GHz filter on a 20×20 mm substrate: at lambda/20 (cell=0.9mm): approximately 500 cells, simulation time approximately 10 seconds. At lambda/40 (cell=0.46mm): approximately 2000 cells, approximately 2-5 minutes. At lambda/60 (cell=0.30mm): approximately 4500 cells, approximately 10-30 minutes. At lambda/100 (cell=0.18mm): approximately 12000 cells, approximately 1-3 hours.
What about Sonnet's conformal meshing?
Sonnet Suites version 14+ includes conformal meshing (also called subsectioning) that allows cells near conductor edges to be subdivided without refining the entire grid. This significantly reduces the total cell count (typically 4-10× fewer cells for the same edge resolution) and speeds up the simulation. With conformal meshing: you can use a coarser base grid (lambda/20-30) and the conformal mesh automatically refines near edges. This is the recommended approach for most filter simulations.
How do I know if the mesh is fine enough?
Convergence test: run the simulation at two different cell sizes (e.g., lambda/30 and lambda/60). Compare the S-parameters: if the resonant frequencies shift by less than 0.5% and the insertion loss changes by less than 0.1 dB: the coarser mesh is adequate. If the shift is larger: use the finer mesh or refine further. For coupled-line filters with narrow gaps (less than 0.1 mm): the coupling coefficient is very sensitive to the mesh resolution in the gap. Always verify that the gap is spanned by at least 3 cells.