Choosing an Antenna for a Meshtastic Node

[Pavel]

Short answer: Choose an antenna resonant for the node's regional band, with the correct connector and polarization, then place it clear of the enclosure and body. Advertised gain is less important than honest band tuning and a usable radiation pattern.

Before you start​

• Set the correct regional radio profile before testing; frequency plan, permitted power and duty-cycle constraints depend on location.
• Use the same modem preset, channel settings and channel key on nodes that are expected to communicate.

Step-by-step method​

1. Step 1: Identify EU868, US915 or the actual regional frequency and reject antennas intended for another LoRa band even if the connector fits.
2. Step 2: Check SMA versus RP-SMA and pigtail type. Avoid stacks of adapters and route the feedline without sharp bends or pressure on the board connector.
3. Step 3: For handheld and fixed nodes, begin with vertical polarization. Keep the radiating element away from battery packs, metal cases and the ground plane conditions it was not designed for.
4. Step 4: Compare antennas on a fixed route with unchanged power and modem preset. Record delivery rate, RSSI and SNR over many packets.

Concrete example​

A high-gain long whip can have deep elevation nulls and perform worse for nearby hilltop paths than a shorter antenna. A VNA match check helps, but field packet delivery remains the final measure.

How to judge the result​

The selected antenna gives repeatable bidirectional delivery and stable link margin in the node's actual orientation, not only a low bench SWR.

What to record​

• Record RSSI, SNR, hop count, delivery success and antenna height for a fixed route instead of quoting one maximum-distance message.
• Line of sight, Fresnel clearance and antenna placement usually improve range more effectively than additional transmit power.
• For solar nodes, calculate daily energy use and worst-month solar input, then include battery reserve for poor weather.

Common mistakes​

• Changing several hardware, software or RF variables at once, which removes the controlled comparison needed to identify the cause.
• Treating one autoscaled screenshot or one unusually good result as proof without recording the settings and repeating the test.
• A physically compatible 915 MHz antenna on an EU868 node can be mistuned enough to reduce range and stress the transmitter match.

Final check​

Power-cycle both nodes and repeat the same bidirectional route test. Save region, preset, antenna, RSSI, SNR and delivery rate while keeping channel keys and precise locations private.