RF PCB Layout for 2.4 GHz and Sub-GHz Radios
Antenna keep-outs, 50-ohm feed lines, matching networks and radio module integration
RF PCB layout is the part of board design that carries radio signals between a transceiver and its antenna without losing range: controlled 50-ohm feed lines, matching networks, ground treatment and antenna keep-outs. Electuni lays out 2.4 GHz Wi-Fi and Bluetooth LE boards and sub-GHz LoRa boards for teams adding wireless connectivity to a product.
Design → Prototype → Production
Key Facts: RF PCB Layout
- Bands
- 2.4 GHz, Sub-GHz
- Radios
- Wi-Fi, BLE, LoRa
- Feed Lines
- 50 Ω GCPW / Microstrip
- Matching
- Pi / T Network Footprints
- Integration
- Modules or Chip-Down
- Lead Designer
- IPC CID+ Certified
What You Get
- RF placement plan covering radio, matching network, antenna and keep-out zones
- 50-ohm feed line geometry calculated against the agreed stackup
- Matching network footprints (pi or T) for tuning on the built board
- Ground stitching, via fencing and shield-can footprints where needed
- Module integration following the vendor's antenna and keep-out guidance
- Altium Designer layout and fabrication package with impedance notes
How We Approach RF PCB Layout
RF layout is decided at placement, not at routing. The antenna position is set first, usually at a board edge or corner and away from batteries, displays, metal fasteners and any metal in the enclosure. The radio and its matching network are then placed so the feed line is short and direct. Module vendors such as Espressif publish antenna keep-out regions for their ESP32 modules, and those areas stay clear of copper, components and traces on every layer.
The feed line from radio to antenna is designed as a 50-ohm transmission line, most often a grounded coplanar waveguide on the top layer over solid ground on layer 2, calculated with a field solver against the fabricator's stackup. A pi or T matching network sits close to the antenna feed, with footprints for every position even if some are fitted with zero-ohm links, so the match can be tuned on the built board instead of fixed by a respin.
Ground is treated as part of the RF circuit. The plane under the radio stays unbroken, stitching vias tie the top-layer ground pour to the reference plane along the feed line and around the RF section, and switching regulators and clocks are kept away from the antenna and its feed. Where emissions or coexistence with a second radio require it, a shield-can footprint is added over the RF section. The same principles apply at 2.4 GHz and sub-GHz, although sub-GHz antennas need more ground area and physical length.
Technical Constraints and Trade-offs
- Antenna performance depends on the ground plane and enclosure. Chip and PCB antennas are specified for a reference ground size, and a smaller board shifts their tuning.
- Keep-out zones under and around a module antenna take board area that cannot carry copper, components or traces on any layer.
- Grounded coplanar waveguide width depends on dielectric thickness, so a 50-ohm line on a thick two-layer board looks very different from one on a four-layer board with a thin layer 2.
- Matching values depend on the final board, antenna and enclosure, so they are tuned on built hardware rather than fixed in the schematic.
- Two radios on one board, such as Wi-Fi and cellular or LoRa and BLE, need physical separation and careful supply decoupling to coexist.
- Sub-GHz antennas are larger and more sensitive to ground length than 2.4 GHz antennas, which shapes the board outline early in the design.
Fabrication and Assembly Handoff
- Stackup with dielectric thickness and material named, so the fabricator can hold the 50-ohm geometry
- Impedance table listing each controlled RF trace, its layer, reference plane and target
- Gerber or ODB++ data with antenna keep-out areas preserved on all copper layers
- Fabrication notes on solder mask over RF traces, as assumed in the impedance calculation
- BOM marking matching components as tuning positions, with initial fitted values
- Assembly drawing showing shield-can position and orientation where used
From a real project
LoRa sensor node with solar harvesting and 15 kV line sensing
A battery-and-solar powered remote monitoring node for electric fence lines, built around an SX1262 sub-GHz LoRa transceiver and an STM32L052 microcontroller. It sits outdoors with no mains power and no network and must report line condition over long range. To keep the link budget LoRa offers with a small antenna, the matching network, ground plane treatment and antenna keep-out were resolved at layout, alongside high-voltage spacing for a sensing front-end rated for 12-15 kV line pulses.
Read the project on our PCB division's site (opens in a new tab)How an Electuni Engagement Works
Step 1
Discovery
Requirements analysis and feasibility study.
Step 2
Architecture
System design, component selection, and risk assessment.
Step 3
Design
Schematic, layout, firmware, and mechanical CAD.
Step 4
Review
DFM/DFT checks and design review with stakeholders.
Step 5
Prototype
Board fabrication, assembly, and firmware bring-up.
Step 6
Handoff
Production files, BOM, and full documentation package.
Tools & Technologies
- Altium Designer
- Altium Layer Stack Manager
- Saturn PCB Toolkit
- Smith Chart Matching Tools
- 3D Clearance Check
RF PCB Layout FAQ
Where should the antenna go on a PCB?
Usually at a board edge or corner, with its keep-out area either extending off the board or kept clear of copper on every layer. Keep it away from batteries, displays, metal screws, connectors and switching regulators. For radio modules, follow the vendor's placement guidance exactly, because the antenna was tuned assuming that position on a host board.
What is a grounded coplanar waveguide and why use it for RF?
A grounded coplanar waveguide (GCPW) is a trace with ground pour on both sides on the same layer and a solid ground plane underneath. Compared with plain microstrip, it confines the field more tightly, gives a narrower 50-ohm trace on thicker boards and reduces coupling to nearby signals. Rows of stitching vias along both sides keep the side grounds tied to the plane.
Why leave footprints for a matching network?
An antenna's impedance changes with the ground plane, enclosure, nearby components and even the hand holding the product. A pi or T network of small inductors and capacitors between the radio and the antenna lets the match be corrected on built boards using a vector network analyser. Without the footprints, a mismatch can only be fixed with a board respin.
Should I use a pre-certified radio module or a chip-down design?
A pre-certified module, such as an ESP32 or nRF module with an integrated antenna, carries its own RF layout and modular certification, which reduces, but does not remove, the end product's regulatory testing. Chip-down design lowers unit cost and board size at volume but makes you responsible for RF layout, matching and full certification. Many products start with modules.
Is RF layout for sub-GHz LoRa different from 2.4 GHz?
The principles are the same: a 50-ohm feed, short traces, solid ground and a clear antenna area. The scale differs. Sub-GHz wavelengths are roughly two and a half to three times longer, so antennas are larger and more dependent on ground size. SX126x-class transceivers commonly use separate transmit and receive paths joined by an RF switch, each needing its own matching.
Does my board need an RF shield can?
Not always. A shield can over the RF section helps when a board has noisy digital circuits, more than one radio, or must pass emissions testing with margin. The footprint costs little board area and keeps the option open, so fitting the can itself can be decided after testing. Pre-certified modules usually include their own shield already.