In PCB design, GPS antenna performance depends not only on the antenna itself, but also on antenna placement, RF routing, grounding, shielding, and separation from noisy digital circuitry. GNSS signals reach the receiver at extremely low levels, so small layout choices such as keeping the RF path short, maintaining a continuous reference ground, and controlling winterference can directly affect sensitivity, signal quality, and positioning reliability.
What GPS Antenna Design Involves
A GPS antenna on a PCB is not defined by trace routing alone. The design starts with how the antenna will be integrated into the board, because the received satellite signal is extremely weak, typically about 15 dB below the thermal noise floor when it reaches the Earth’s surface. At that signal level, antenna type, PCB ground plane, RF path length, and nearby interference all influence whether the receiver can preserve usable signal quality.
For this reason, GPS antenna PCB design usually means managing a complete receive path rather than drawing one RF line. The board layout has to account for antenna selection, the continuity and size of the reference ground, separation from noisy digital circuitry, and loss or mismatch along the feed path. If any of these parts is handled poorly, signal quality can drop noticeably before the receiver even begins acquisition and tracking.
GPS Antenna Types
GPS antennas used in PCB design can be viewed from two dimensions. One is the physical antenna form, such as patch, chip, helix, and external antennas. The other is the receive architecture, which separates active antennas from passive antennas.
| Type | Structure | Space Requirement | Ground Plane Dependence | Main Design Note |
|---|---|---|---|---|
| Patch antenna | Flat ceramic radiator mounted over a ground plane | Medium to large | High | Usually offers stronger GPS reception when the PCB can provide enough ground area |
| Chip antenna | Small surface mounted antenna placed directly on the PCB | Very small | High | Saves board space but needs tighter control of clearance, PCB size, and tuning |
| Helix antenna | Three dimensional spiral radiator | Medium footprint with more height | Medium | Useful when the product can accept vertical space and a non planar antenna form |
| External antenna | Antenna placed outside the main PCB through a coaxial cable | Off board | Low on the main PCB | Suitable when the board, enclosure, or local noise prevents good onboard antenna performance |
Patch, chip, helix, and external antennas do not describe the same design tradeoff. Patch and chip antennas are both common onboard options, but they respond very differently to PCB size and ground plane conditions. Helix antennas shift more of the constraint to product height, while external antennas move the radiating element away from the PCB and reduce the dependence on the board itself.
Antenna Architecture
| Type | Structure | Power Requirement | Electrical Characteristic | Main Design Note |
|---|---|---|---|---|
| Passive antenna | Antenna only, with no integrated amplifier | No | No gain added in the antenna path | Requires tighter control of feed length, loss, matching, and receiver side layout |
| Active antenna | Antenna with an integrated LNA, often with filtering | Yes | Adds gain before the receiver input | More tolerant of cable loss and longer feed paths, but requires DC bias and proper input level control |
Active and passive do not replace patch, chip, helix, or external antennas. They describe whether the antenna path includes active front end electronics. A patch antenna can be passive or active, and an external antenna can also be passive or active. Keeping these categories separate makes the later comparison more accurate.
Choosing a GPS Antenna for the PCB
Antenna selection starts with board size and available installation space. If the PCB can provide a larger and more open antenna area, a patch antenna is often the stronger option. A 25 × 25 mm patch is a common size with a balanced combination of cost and performance, while patch antennas smaller than 17 × 17 mm often provide only moderate navigation performance. Patch performance also improves with a larger ground plane, and about 70 × 70 mm is often treated as a favorable ground size when stronger results are expected. When the board is much smaller, a chip antenna or an external antenna is often the more realistic option.
Signal environment and sensitivity target come next. When the antenna must sit close to digital circuitry, clocks, power converters, or other transmitters, interference becomes harder to control. Clock harmonics can extend into the GNSS band region, and strong nearby transmitters can reduce receiver sensitivity. Under these conditions, an active antenna becomes more attractive because the integrated LNA lowers the overall receiver noise figure, and filtered active antennas are less sensitive to interference coupled onto the cable. An active antenna also becomes a more natural fit when the RF cable between antenna and receiver exceeds about 10 cm.
Cost and current consumption then determine whether the added performance margin is worth the added complexity. A passive antenna does not need power and keeps the front end simpler, but it places tighter limits on feed loss, matching quality, and cable length. An active antenna adds front end electronics and often draws about 3 to 20 mA, or roughly 10 to 60 mW, depending on the design. For compact products with short feed paths, tighter budgets, and moderate positioning requirements, a passive solution can still fit well. When higher sensitivity, longer cable runs, or more margin against attenuation and interference is needed, the active option often makes more sense.
Ground Plane Under the Antenna
The ground plane is part of the GPS antenna system, not just a piece of PCB copper. For a patch antenna, ground plane size changes gain, axial ratio, center frequency, and radiation pattern. A ground area of about 50 × 50 mm to 70 × 70 mm is commonly treated as a good range for ceramic patch designs, and a smaller ground plane usually reduces performance margin.
Ground plane requirements also change with antenna type. Patch antennas depend strongly on a solid supporting ground plane, while chip antennas also use the host PCB as part of the antenna system and become more sensitive when the board is small or the ground is broken up. When the ground area is limited or heavily constrained by the enclosure, an external antenna often becomes easier to integrate than forcing an onboard antenna into an unsuitable PCB environment.
Grounding, Shielding and Isolation
Grounding, shielding, and isolation all aim at the same problem: unwanted energy reaching the GNSS receiver input. The received GPS signal is already below the thermal noise floor when it arrives at the antenna, so clock leakage, switching noise, and coupling from digital lines can reduce C over N0 and make acquisition less stable. The receiver side is usually arranged with a cleaner RF ground region, and that RF ground is then tied to the digital ground at a single point.
Shielding is commonly applied around the receiver, the matching network, and any external LNA placed near the input path. A ground plane under the GNSS module reduces interference at the RF input line, and a dense fence of ground vias around the microstrip or coplanar waveguide reduces field leakage into nearby circuitry. These measures do not change the antenna type, but they change how much noise reaches the receiver front end.
Isolation depends on distance and layout partitioning. The antenna feed and receiver front end should stay away from clocks, high speed digital buses, switching regulators, and other radios. Wireless transmitters such as GSM, WCDMA, Wi Fi, and Bluetooth can produce out of band energy at the GNSS input, and that energy can be reduced with cleaner grounding, stronger separation, and input filtering when the design requires it. When the antenna cannot be moved farther from noisy electronics, shielding the electronics side becomes the more effective option.
RF Feed Path and Matching
The RF path between the GPS antenna and the receiver is part of the antenna system itself. Once the signal leaves the antenna, every extra millimeter of trace, every impedance discontinuity, and every mismatch reduces the signal level seen at the receiver input. For this reason, the feed path is usually kept short and designed as a 50 Ω transmission line with a continuous reference ground. Long routing raises insertion loss, and routing across an interrupted reference structure makes the signal path less stable.
Matching is influenced by more than trace width. GPS antennas are commonly intended to work into a 50 Ω load, but the input impedance can shift when the ground plane changes, when the enclosure material alters the surrounding dielectric environment, or when nearby metal and components disturb the antenna near field. The smaller the antenna, the less margin it usually has against this shift. Once the impedance moves away from the intended point, the antenna no longer transfers energy into the receiver as efficiently, and the center response can move away from the GNSS band.
Routing details then decide whether the designed match can still be preserved on the PCB. Unnecessary vias should be avoided in the antenna feed because each via adds an inductive discontinuity, and at GPS RF frequencies one via can add about 10 Ω of impedance. The feed should also avoid extra parasitic capacitance and should stay referenced to solid ground along the path. These layout details matter more with passive antennas, because there is no gain stage in the antenna path to offset feed loss.
When the antenna is very small or the PCB environment makes matching harder, placing an LNA close to the antenna can reduce the impact of feed loss. For receiver chains that support passive antennas, about 15 dB of antenna LNA gain is usually sufficient even with cable lengths up to 5 m, while active antenna use becomes more favorable once the cable between antenna and receiver grows beyond about 10 cm. In other words, matching still matters, but the closer the gain stage sits to the antenna, the less exposed the weak GNSS signal is to loss before amplification.