L
Admin
[Technical article / FAQ, rfforum Editorial Team]
A GNSS sky plot can look healthy while the correlator is still chewing multipath. Choke rings and ground-plane size own different parts of that error; treating them as interchangeable "antenna upgrades" is how a survey-grade claim fails on a metal roof or a vehicle roof with a short skirt.
1. What each geometry actually buys
- A choke ring attenuates low-elevation and below-horizon energy that arrives as multipath from the ground or deck. It does not magically enlarge the sky FOV; it shapes the pattern so the correlator sees less delayed energy.
- Ground-plane size and edge treatment set diffraction and image currents. A plane that is too small relative to the lowest frequency lets the feed "see" the vehicle body; a plane that is large but thin-edged can still diffract multipath into the upper hemisphere.
- Phase-centre stability with elevation and azimuth often moves more than raw C/N0 when multipath is present. A pretty sky plot with wandering phase centre is a multipath story, not a constellation story.
2. Freeze list before you blame the receiver firmware
- Antenna type (choke / patch / helix), ground-plane diameter or vehicle-roof mount note, and measured or datasheet phase-centre variation vs elevation
- Mount: height above deck, nearby vertical metal, and whether the plane is bonded or floating relative to the chassis
- Whether the multipath metric (code-minus-carrier, multipath estimator, or baseline residual) moves when you raise the antenna or add a temporary skirt — not only whether satellites stay locked
- Same receiver, same firmware, two mounts: if only the mount changes and residuals jump, stop the firmware hunt
3. Habit
Never freeze "survey accuracy" from a sky plot alone. Archive antenna PN, plane size or choke note, and mount height next to every multipath or residual claim. If raising the antenna 20–30 cm or adding a temporary skirt kills the residual, the constellation was never the patient.
4. Practical five-minute gate
- Photograph the mount and measure plane diameter or choke depth
- Compare residuals before/after a known better mount or temporary skirt
- Write phase-centre vs elevation on the same sheet as the accuracy number
Discussion: on your last stubborn multipath residual, did the fix land in choke depth, plane size, or mount height — and do you forbid publishing accuracy from a sky plot with no mount note?
— rfforum 编辑部 / Editorial Team
A GNSS sky plot can look healthy while the correlator is still chewing multipath. Choke rings and ground-plane size own different parts of that error; treating them as interchangeable "antenna upgrades" is how a survey-grade claim fails on a metal roof or a vehicle roof with a short skirt.
1. What each geometry actually buys
- A choke ring attenuates low-elevation and below-horizon energy that arrives as multipath from the ground or deck. It does not magically enlarge the sky FOV; it shapes the pattern so the correlator sees less delayed energy.
- Ground-plane size and edge treatment set diffraction and image currents. A plane that is too small relative to the lowest frequency lets the feed "see" the vehicle body; a plane that is large but thin-edged can still diffract multipath into the upper hemisphere.
- Phase-centre stability with elevation and azimuth often moves more than raw C/N0 when multipath is present. A pretty sky plot with wandering phase centre is a multipath story, not a constellation story.
2. Freeze list before you blame the receiver firmware
- Antenna type (choke / patch / helix), ground-plane diameter or vehicle-roof mount note, and measured or datasheet phase-centre variation vs elevation
- Mount: height above deck, nearby vertical metal, and whether the plane is bonded or floating relative to the chassis
- Whether the multipath metric (code-minus-carrier, multipath estimator, or baseline residual) moves when you raise the antenna or add a temporary skirt — not only whether satellites stay locked
- Same receiver, same firmware, two mounts: if only the mount changes and residuals jump, stop the firmware hunt
3. Habit
Never freeze "survey accuracy" from a sky plot alone. Archive antenna PN, plane size or choke note, and mount height next to every multipath or residual claim. If raising the antenna 20–30 cm or adding a temporary skirt kills the residual, the constellation was never the patient.
4. Practical five-minute gate
- Photograph the mount and measure plane diameter or choke depth
- Compare residuals before/after a known better mount or temporary skirt
- Write phase-centre vs elevation on the same sheet as the accuracy number
Discussion: on your last stubborn multipath residual, did the fix land in choke depth, plane size, or mount height — and do you forbid publishing accuracy from a sky plot with no mount note?
— rfforum 编辑部 / Editorial Team
