Which Meshtastic Antennas Are Actually Good? $8,000 Analyzer Test

Every mesh antenna on the market claims some kind of gain and some kind of frequency range. Very few of them come with proof. So we took a stack of the antennas people actually put on their Meshtastic and MeshCore nodes and hooked them, one at a time, to an Anritsu S331L Site Master, a professional handheld cable and antenna analyzer that covers 2 MHz to 4 GHz.

The lineup: a set of four Amazon antennas, two muzi works antennas (the ones the forums tend to recommend), the antenna that comes in the Heltec expansion kit, the tiny stock antenna on a Heltec V3, the antenna that ships with the Seeed Studio Solar Node, a Slinkdsco 5.8 dBi fiberglass antenna, and a few commercial 900 MHz antennas including a big Yagi somebody gave me.

For each one we looked at SWR at 910 MHz, which sits in the middle of the US 902 to 928 MHz band the mesh networks use, and then flipped to the Smith chart to see how close the antenna actually was to a 50 ohm match. Some performed exactly how we expected. Others did not.

A Quick Word on the Test Setup

This was not a lab. We were sweeping antennas on a bench, and several of them needed N type to SMA adapters to connect to the analyzer, which is not ideal. Some of these antennas technically want a ground plane, and real gain and pattern testing needs an anechoic chamber. What this setup does tell you, reliably, is where an antenna is resonant and how well it matches 50 ohms at the frequency you care about.

Somebody asked whether being close to metal was skewing the results. I don't think so. Most of these are dipole style antennas with their own radiating element and ground, so they are essentially free space antennas. We could have taken them out to the middle of the yard, but the results weren't bad enough to make me suspect the setup.

The Commercial 900 MHz Antennas

I also brought some commercial antennas, the kind you'd find on LMR installations like an 800 MHz public safety system. The first was a unity gain Mobile Mark antenna rated for 806 to 960 MHz. At 910 it read about 1.9:1, with the SWR dropping off toward the higher frequencies, so the mesh band is just on the edge of what it's designed for. The Smith chart was almost a perfect circle, but centered around 100 ohms, so it's going in a different direction from 50 ohms entirely.

The next commercial antenna also drew a clean circle on the Smith chart. Still not 50 ohms, but at least a circle, and it read about 1.6:1. Considering the gain you get from that style of antenna, that might be a good compromise.

Then there's the giant Yagi. Someone randomly handed it to me and said I might need it someday, and if you're a junkyard guy like me, you never throw that stuff away. It most likely came off a SCADA or old LMR system running data in the 900 MHz range. It had a much nicer, flatter curve, with its lowest point around 930 MHz and about 1.4:1 at 910. On the Smith chart it was about half a circle. If you know how to read that properly, leave a comment. Veritasium also put out a great video on reading Smith charts if you want to learn.

Anything below 2:1 will work. But if you're building a repeater and want the most out of it, sweep it and get it as close as you can.

Pro Analyzer vs NanoVNA

So how much better is something like the Anritsu than my NanoVNA? Think of it like multimeters. A cheap one from the discount store will show you a voltage, but do you trust it? With a Fluke, you never doubt the reading. Fluke is the industry standard for meters, and Anritsu is the industry standard for this kind of equipment.

The S331L can be sent back to Anritsu for calibration and certification, which gives you traceability to NIST. When it says 1 dB, it really is 1 dB. You won't get that with a NanoVNA, but the NanoVNA does essentially the same job, and for checking a mesh antenna its results will likely line up. It's a level-of-trust thing more than a different answer.

SMA vs RP-SMA: Why Two Connectors Exist

While we were swapping adapters, connectors came up. Here's the short version, and don't quote me on the exact rule: the FCC requires that certain certified unlicensed devices not have an antenna that's easy to swap in the field (the rule usually cited is Part 15.203). FRS radios are a good example, since they're certified with a fixed antenna. Manufacturers who still wanted to let customers change antennas got around it with a connector that isn't the common one, and that's a big part of why reverse polarity SMA exists alongside regular SMA. Ham gear is allowed interchangeable antennas, so it typically uses regular SMA.

The two look nearly identical, so check which one your node has before you order an antenna. That's close enough to get you started on your own research.

Analyzer and Antenna Specs

Conclusion: Trust the Sweep, Not the Label

Most of what we tested was usable. The Slinkdsco was the best base antenna of the day, the Seeed Solar Node antenna punched above what you'd expect from an included antenna, and the commercial antennas showed why people pay more for them. The Heltec V3 stubby was exactly as bad as its reputation.

The real lesson is consistency. Budget antennas aren't built to the same quality and standard as commercial models, and two antennas from the same listing can measure differently. You don't need a professional analyzer to find that out. A NanoVNA will do it.

If you've got a NanoVNA sitting in a drawer, pull it out and sweep your mesh antennas before the next one goes up on a mast. And if you can read a Smith chart better than I can, the comments are open.