L
Admin
[Technical article, rfforum Editorial Team]
A single-layer rectangular patch on a thin substrate is narrowband: fractional bandwidth of a few percent is typical once you ask for |S11| < −10 dB. Three families of tricks show up again and again when you need more.
1. Thicker substrate / lower εr
Bandwidth scales roughly with substrate height over wavelength and improves when εr drops. The cost is surface-wave loss, feed inductance (probe or via) and a larger lateral size for the same frequency. Crossing about h ≈ 0.05 λ0 often needs a differently designed feed or you will not match at all.
2. U-slot (and related slots)
A U-shaped slot in the patch introduces a second resonance that can be tuned next to the main TM01 resonance. Done well you get a dual-resonant match that looks like one wide band. Done poorly you get two narrow dips with a high VSWR hill between them. The slot also changes the current path, so cross-pol and pattern tilt need a re-check, not only S11.
3. Stacked / parasitic patches
A second patch above (or beside) the driven one adds another resonance and can broaden both match and gain bandwidth. Alignment, spacer dielectric and the gap are as critical as the copper dimensions. Manufacturing tolerance on the gap is a common reason a stacked design that simulated beautifully fails on the first build.
4. What does not buy radiation bandwidth
A matching network at the connector can make |S11| look broadband while the antenna itself still radiates only in a narrow band. Always separate match bandwidth from radiation / efficiency bandwidth (see also electrically-small discussions on this board).
5. A practical verification order
1. Simulate S11 and total efficiency across the band.
2. Build one board; measure S11 with the final plastic/ground if you have it.
3. If efficiency matters, use a Wheeler-style check for triage and a chamber for the number you will publish.
Discussion: for a single-feed GPS/GNSS or 2.4 GHz IoT patch that must stay under a few millimetres of height, which trick has given you the best measured |S11| bandwidth without destroying efficiency — U-slot, stacked, or simply a thicker foam spacer?
— rfforum 编辑部 / Editorial Team
A single-layer rectangular patch on a thin substrate is narrowband: fractional bandwidth of a few percent is typical once you ask for |S11| < −10 dB. Three families of tricks show up again and again when you need more.
1. Thicker substrate / lower εr
Bandwidth scales roughly with substrate height over wavelength and improves when εr drops. The cost is surface-wave loss, feed inductance (probe or via) and a larger lateral size for the same frequency. Crossing about h ≈ 0.05 λ0 often needs a differently designed feed or you will not match at all.
2. U-slot (and related slots)
A U-shaped slot in the patch introduces a second resonance that can be tuned next to the main TM01 resonance. Done well you get a dual-resonant match that looks like one wide band. Done poorly you get two narrow dips with a high VSWR hill between them. The slot also changes the current path, so cross-pol and pattern tilt need a re-check, not only S11.
3. Stacked / parasitic patches
A second patch above (or beside) the driven one adds another resonance and can broaden both match and gain bandwidth. Alignment, spacer dielectric and the gap are as critical as the copper dimensions. Manufacturing tolerance on the gap is a common reason a stacked design that simulated beautifully fails on the first build.
4. What does not buy radiation bandwidth
A matching network at the connector can make |S11| look broadband while the antenna itself still radiates only in a narrow band. Always separate match bandwidth from radiation / efficiency bandwidth (see also electrically-small discussions on this board).
5. A practical verification order
1. Simulate S11 and total efficiency across the band.
2. Build one board; measure S11 with the final plastic/ground if you have it.
3. If efficiency matters, use a Wheeler-style check for triage and a chamber for the number you will publish.
Discussion: for a single-feed GPS/GNSS or 2.4 GHz IoT patch that must stay under a few millimetres of height, which trick has given you the best measured |S11| bandwidth without destroying efficiency — U-slot, stacked, or simply a thicker foam spacer?
— rfforum 编辑部 / Editorial Team
