R
Member
A plain rectangular patch is easy to match but its 2:1 VSWR bandwidth is often only 2-3%, which is rarely enough once the enclosure and nearby metal are added. Two techniques have actually worked for me on the bench: slotting the patch and stacking a parasitic patch above it.
Slotting first. A U-slot cut into the patch creates a second resonance close to the main one. When the two resonances are spaced right, the return-loss curve flattens into a wider band instead of two sharp dips. The trap is cutting the slot by guesswork: too long a slot pulls the second resonance too far down and kills the gain at band centre, too short and the resonances never merge. I tune slot length in small steps and watch both S11 and realized gain, not S11 alone, because a wide S11 dip with a gain notch in the middle is useless.
Stacked patches are the more forgiving route. A parasitic patch roughly a tenth of a wavelength above the driven patch, slightly different in size, gives a coupled resonance that widens the band to 10% or more in my tests. Key details: the air gap or foam spacer thickness sets the coupling, and the parasitic patch must be aligned carefully, because a millimetre of lateral offset shifts the upper resonance noticeably. The cost is height and a more complex assembly, so I only use it when the product can afford a few extra millimetres.
A practical sequence: start with the plain patch matched at band centre, then add the slot or the stacked element, then re-check the pattern, because both techniques can tilt the beam or raise cross-polarization. If the pattern degrades, back off the coupling before chasing S11 further.
Which do you reach for first when a patch is too narrowband, slotting or stacking, and what bandwidth have you actually shipped with each?
Slotting first. A U-slot cut into the patch creates a second resonance close to the main one. When the two resonances are spaced right, the return-loss curve flattens into a wider band instead of two sharp dips. The trap is cutting the slot by guesswork: too long a slot pulls the second resonance too far down and kills the gain at band centre, too short and the resonances never merge. I tune slot length in small steps and watch both S11 and realized gain, not S11 alone, because a wide S11 dip with a gain notch in the middle is useless.
Stacked patches are the more forgiving route. A parasitic patch roughly a tenth of a wavelength above the driven patch, slightly different in size, gives a coupled resonance that widens the band to 10% or more in my tests. Key details: the air gap or foam spacer thickness sets the coupling, and the parasitic patch must be aligned carefully, because a millimetre of lateral offset shifts the upper resonance noticeably. The cost is height and a more complex assembly, so I only use it when the product can afford a few extra millimetres.
A practical sequence: start with the plain patch matched at band centre, then add the slot or the stacked element, then re-check the pattern, because both techniques can tilt the beam or raise cross-polarization. If the pattern degrades, back off the coupling before chasing S11 further.
Which do you reach for first when a patch is too narrowband, slotting or stacking, and what bandwidth have you actually shipped with each?
