L
Admin
[Technical article / FAQ, rfforum Editorial Team]
Element patterns and array factor are often frozen as if mutual coupling were a small correction. On scanned arrays, surface-wave / Floquet-mode effects can put a null (scan blindness) in a look direction that still looks "fine" on a broadside-only plot.
1. What coupling actually does
- Active impedance moves with scan; a match tuned at broadside can reflect hard at the scan angle that matters for the product
- Scan blindness shows up as a deep active-reflection spike and a pattern null; a passive embedded-element plot alone can miss it
- Ground-plane size, dielectric thickness, and lattice spacing set where the blindness sits — changing one without re-checking scan is a method change
2. Freeze list before you claim "scan OK"
- Active |S11| or active impedance vs scan angle (E/H or uv), not only broadside VSWR
- Lattice spacing, stackup, and whether blindness was searched inside the product scan cone
- Whether cal / pattern claims used the same feed network and edge termination as the product
- Temperature / enclosure note if the dielectric or cavity moves the blindness into band
3. Habit
Never gate an array release on broadside-only match and pattern. Archive the worst-case active reflection inside the scan cone next to every "element matched" claim.
Discussion: on your last scanned array, did blindness show up first in active |S11|, in a pattern null, or only after a field complaint — and do you require an active-impedance sweep before freezing scan performance?
— rfforum 编辑部 / Editorial Team
Element patterns and array factor are often frozen as if mutual coupling were a small correction. On scanned arrays, surface-wave / Floquet-mode effects can put a null (scan blindness) in a look direction that still looks "fine" on a broadside-only plot.
1. What coupling actually does
- Active impedance moves with scan; a match tuned at broadside can reflect hard at the scan angle that matters for the product
- Scan blindness shows up as a deep active-reflection spike and a pattern null; a passive embedded-element plot alone can miss it
- Ground-plane size, dielectric thickness, and lattice spacing set where the blindness sits — changing one without re-checking scan is a method change
2. Freeze list before you claim "scan OK"
- Active |S11| or active impedance vs scan angle (E/H or uv), not only broadside VSWR
- Lattice spacing, stackup, and whether blindness was searched inside the product scan cone
- Whether cal / pattern claims used the same feed network and edge termination as the product
- Temperature / enclosure note if the dielectric or cavity moves the blindness into band
3. Habit
Never gate an array release on broadside-only match and pattern. Archive the worst-case active reflection inside the scan cone next to every "element matched" claim.
Discussion: on your last scanned array, did blindness show up first in active |S11|, in a pattern null, or only after a field complaint — and do you require an active-impedance sweep before freezing scan performance?
— rfforum 编辑部 / Editorial Team
