433 MHz PCB Antenna Design: A Practical Guide

A 433 MHz PCB antenna design looks simple on paper and defeats a surprising number of otherwise solid products. The quarter wave length is easy to calculate. Everything after that is where designs fail.

This guide covers the decisions that actually determine whether your radio hits its link budget.

Why 433 MHz is harder than 2.4 GHz

At 2.4 GHz a quarter wave monopole is about 31 mm. At 433 MHz it is roughly 173 mm in free space. Most handheld or wearable products do not have 173 mm to spare.

That single fact drives every other choice. You will be working with a compromised, electrically small antenna, and the job becomes managing that compromise rather than avoiding it.

The upside is worth it. Sub-GHz propagates further, penetrates foliage and structures better, and tolerates body blocking more gracefully than 2.4 GHz.

Choosing an antenna topology

Three options cover most sub-GHz products.

Wire monopole. A simple wire or whip soldered to the board. Best efficiency, worst mechanical integration. Use it when range matters more than form factor.

Meandered PCB trace. The trace folds back on itself to fit the quarter wave into a smaller footprint. Costs nothing to manufacture. Expect 30 to 60 percent efficiency depending on how aggressively you meander.

Chip antenna. A ceramic part with a published matching network. Small and repeatable, but at 433 MHz the good ones are still physically large and efficiency is modest.

For most sub-GHz products the meandered PCB trace is the right starting point. It is free, it is repeatable across production, and its weaknesses are predictable.

The ground plane is part of the antenna

This is the single most common mistake in sub-GHz design.

A monopole works against a ground plane. The ground plane forms the other half of the radiating structure. If your ground plane is much shorter than a quarter wavelength, it becomes the dominant limit on efficiency, and no amount of matching network design tuning will recover it.

Practical rules:

• Give the ground plane as much length as the enclosure allows, ideally in line with the antenna
• Keep a clear keepout under and around the antenna trace, with no copper, no components and no traces on any layer
• A keepout of at least 5 mm on each side is a reasonable starting point at 433 MHz
• Place the antenna at a board edge or corner, never in the middle

If your product has a battery, remember that the battery is a large conductor. It will detune the antenna. Model it or measure with it in place.

Matching network topology

Plan for a pi network footprint even if you hope not to need it.

Lay out three placeholders: series inductor or capacitor, and two shunt positions to ground. Populate them with 0 ohm links and DNP parts initially. When the prototype comes back and the impedance is not 50 ohms, and it will not be, you have somewhere to put the correction.

Skipping this costs a board spin. Including it costs three pads.

Measuring with a NanoVNA

Simulation gets you close. Measurement tells you the truth.

Calibrate first. Open, short, load at the end of the same cable you will use for the measurement. Skipping calibration makes the measurement meaningless.

Watch the cable. On an electrically small antenna the coax braid will radiate and shift your reading. Add ferrite beads on the cable and use the shortest practical length.

Measure in the enclosure. Plastic housings shift resonance downward, often by several megahertz. A design tuned bare and assembled later will be off frequency.

Target return loss, not perfection. S11 below minus 10 dB across your band is a good working target. Chasing minus 20 dB on a small antenna is usually wasted effort.

Sharing one antenna between two bands

Some designs need a single antenna to serve both a sub-GHz radio and a GPS receiver at 1575 MHz. This is achievable with an LC diplexer.

The diplexer presents a low impedance path at 433 MHz toward the radio and a high impedance path toward the GPS front end, and the reverse at L1. The design constraints are:

• Insertion loss in the GPS path directly reduces your position fix sensitivity, so budget it carefully
• The sub-GHz path must handle full transmit power without desensitising or damaging the GPS LNA
• Component tolerance matters more than in a single band design, so specify tight tolerance parts

An external LNA on the GPS path can recover some of the diplexer loss, at the cost of current draw.

Transmit power and matching at +22 dBm

At full output power the matching network sees real current. Two things change:

Component ratings. Check the voltage and current ratings on your matching components, not just their values. A 0402 capacitor that is fine at 0 dBm can fail at +22 dBm.

Harmonics. Higher output power raises harmonic content. A low pass filter after the matching network is usually necessary to pass regulatory emissions limits, and it is far cheaper to design in than to retrofit. For layout guidance on passing emissions, see our guide on EMC compliant PCB design.

A workable design sequence

1. Fix the enclosure dimensions and the available ground plane length 2. Choose the topology that fits, usually a meandered trace 3. Lay out with a generous keepout and a pi network footprint 4. Simulate to get within a few percent 5. Build, then measure with a calibrated NanoVNA inside the real enclosure 6. Tune the pi network to bring S11 below minus 10 dB across the band 7. Verify harmonics against your regional emissions limits before certification

Common failures worth avoiding

• Tuning the antenna on a bare board, then finding it off frequency once assembled
• Running a trace or pour under the antenna keepout on an inner layer
• Omitting the matching network footprint to save space
• Ignoring the battery and cabling during measurement
• Treating a chip antenna's datasheet matching values as final rather than as a starting point

Getting help with sub-GHz RF

Sub-GHz antenna work rewards measurement discipline more than clever theory. If your link budget is not closing, the problem is usually the ground plane or the enclosure rather than the radio.

SRQ Robotics designs and characterizes sub-GHz RF front ends, including balun design, STM32WL based radios at full transmit power, LC diplexers for shared antenna paths, and NanoVNA verification in the final enclosure. Learn more about our PCB design services or get in touch if you are stuck on a range problem.