L
Admin
[Technical article, rfforum Editorial Team]
Adaptive passes that no longer change S11 can still be lying about fields, loss, or pattern. Refining "one more time" is expensive and often not the right next move.
1. S11 convergence ≠ field convergence
Port impedance, loss, and far-field can still walk after |S11| looks flat. Watch the quantity you will ship on (efficiency, gain ripple, isolation), not only the cheapest S-parameter.
2. Three checks before another pass
- Seed the metal edges / slots that carry current; uniform volume refine rarely fixes an under-seeded edge
- Confirm the port and de-embed length are stable (wild port Z → stop refining the DUT)
- Compare two independent meshes or two solvers on a thru; if they disagree, the model or boundary is wrong
3. Habit
Log which observable you converged on. A report that only shows S11 pass history is incomplete for antenna and filter sign-off.
Discussion: what observable do you freeze as the adaptive stop criteria for antennas vs filters, and have you ever passed S11 and failed efficiency after a mesh redo?
— rfforum 编辑部 / Editorial Team
Adaptive passes that no longer change S11 can still be lying about fields, loss, or pattern. Refining "one more time" is expensive and often not the right next move.
1. S11 convergence ≠ field convergence
Port impedance, loss, and far-field can still walk after |S11| looks flat. Watch the quantity you will ship on (efficiency, gain ripple, isolation), not only the cheapest S-parameter.
2. Three checks before another pass
- Seed the metal edges / slots that carry current; uniform volume refine rarely fixes an under-seeded edge
- Confirm the port and de-embed length are stable (wild port Z → stop refining the DUT)
- Compare two independent meshes or two solvers on a thru; if they disagree, the model or boundary is wrong
3. Habit
Log which observable you converged on. A report that only shows S11 pass history is incomplete for antenna and filter sign-off.
Discussion: what observable do you freeze as the adaptive stop criteria for antennas vs filters, and have you ever passed S11 and failed efficiency after a mesh redo?
— rfforum 编辑部 / Editorial Team
