RF Connectors for Makers: SMA, BNC, UHF, and N-Type Explained
This site's wiring reference covers AWG, JST, Dupont, and XT60 — the connectors that move power and low-frequency signals between boards, motors, and batteries. RF connectors are a different animal entirely, and treating them like ordinary wire connectors is how people end up with 3 dB of loss they can't explain, a cracked SMA center pin, or a "broken" antenna that was actually just a mismatched connector pair. If you're setting up a NanoVNA, building an antenna for the ATS Mini or a ham radio, wiring a GPS or LoRa module's external antenna pigtail, or doing anything with an RTL-SDR beyond the stock dongle, you need to know this family of connectors specifically.
Why RF Connectors Are Different
An ordinary wire connector just needs to pass current without excessive resistance. An RF connector has to maintain a controlled impedance (almost always 50 ohms in maker-relevant gear) right through the connector body, with minimal reflection at the junction — any discontinuity in that impedance shows up as signal loss and standing waves, which is exactly what a NanoVNA's SWR sweep is measuring. That's why RF connectors are precision-machined, why mismatched connector families don't mate even when they look similar, and why a cheap connector with sloppy dimensional tolerance can measurably hurt performance in a way a cheap banana plug never would.
The Connector Families You'll Actually Encounter
ConnectorTypical use on this siteNotes SMA (and RP-SMA)RTL-SDR dongles, ESP32/LoRa module antennas, NanoVNA ports, handheld ham radiosThreaded, screw-on, good to several GHz. RP-SMA ("reverse polarity") swaps the pin/socket gender convention specifically to be incompatible with standard SMA on consumer WiFi gear — a huge source of "why won't this screw on" frustration. Check polarity, not just thread size, before buying adapters. BNCOscilloscope probes, older bench RF gear, some ham radio accessoriesQuarter-turn bayonet lock instead of threads — fast to connect/disconnect, which is why it's the default on scope probes where you're plugging and unplugging constantly. UHF (PL-259/SO-239)CB radios, many base-station HF ham radio antennasNot actually constant-impedance and not great above ~150 MHz, but it's the de facto standard on HF gear for historical reasons. Expect to need a PL-259-to-SMA or PL-259-to-BNC adapter when connecting HF gear to SDR test equipment. N-typeHigher-power or outdoor antenna runs, CO2 laser-adjacent RF (not applicable here, but common on serious ham/WiFi installations)Larger, weatherproof-capable with the right boot, handles more power than SMA. Shows up when a maker project graduates from a dongle antenna to a real outdoor installation. MCX / MMCX / U.FL (IPEX)GPS modules, internal antenna pigtails inside enclosures, some compact SDR hardwareSnap-fit, not threaded, and genuinely fragile — U.FL connectors are rated for a low number of mate/unmate cycles (often cited around 30) before the connection degrades. Treat these as "plug once and leave it" connections, not something you disconnect routinely.Reading a Connector: Gender and Polarity Without Guessing
RF connector gender is defined by the center pin, not the outer shell shape the way it might be with a plug/socket pair elsewhere — a connector with a center pin sticking out is "male" regardless of whether the surrounding shell looks like it should be the opposite. Reverse-polarity variants exist specifically to break this rule on purpose (RP-SMA is the big one), so when buying an adapter, specify both the connector family and gender/polarity explicitly, and when in doubt, measure: a standard SMA male has a pin, a standard SMA female has a socket; RP-SMA male has a socket where you'd expect a pin.
Soldering RF Connectors: What's Actually Different
Soldering an SMA or BNC connector onto coax isn't like soldering a JST pigtail. A few points that trip people up the first time:
- Strip to spec, not by eye. Each connector is designed for a specific coax diameter and dielectric thickness (RG-316, RG-174, RG-402, etc.). Too much exposed shield or center conductor changes the local impedance right at the connector and creates a measurable reflection — this matters more than it looks like it should at HF/VHF and a lot more at UHF and above.
- Solder the center pin before you solder the shield. Working inside-out avoids melting solder joints you just made on the next step, and lets you trim the center pin flush before it's buried under the shell.
- Heat the shield braid quickly and move on. Letting heat linger on the shield-to-shell solder joint can melt the connector's internal dielectric (especially on smaller connectors like SMA and MCX), which changes the impedance permanently and is not something you can fix after the fact.
- Crimp connectors exist and are often the better choice. For RG-58/RG-8-class coax, a crimp-on SMA/BNC/N connector with the correct crimp tool produces more consistent results than hand-soldering, especially for someone doing this occasionally rather than daily.
Adapters, Barrels, and the Hidden Cost of Chaining Them
It's tempting to solve a connector mismatch by stacking two or three adapters — SMA-to-BNC, then BNC-to-UHF, say. Each junction is a small impedance discontinuity, and at HF this barely matters, but by VHF/UHF (where most maker SDR work and WiFi/LoRa modules live) a chain of adapters measurably adds loss and can show up directly on a NanoVNA sweep as unexpected SWR bumps. When troubleshooting a "bad antenna" reading, check the adapter chain first — a single direct adapter is usually fine, three stacked together is a real suspect.
Buying Guidance
For antenna pigtails and dongle connections, a basic SMA-male-to-SMA-female (or RP-SMA, check your hardware) adapter cable from a reputable brand is cheap enough that buying a 2-3 pack of known-good ones beats cycling through unlabeled bags of mixed connectors. For anything feeding a NanoVNA calibration or a serious antenna build, spend the extra few dollars on name-brand connectors (Amphenol, Pasternack, or equivalent) — connector-to-connector dimensional consistency is exactly what you're paying for, and it's the difference between a calibration that holds and one that drifts.
None of this is complicated once the mental model clicks: RF connectors aren't fasteners, they're precision impedance transitions that happen to also hold two cables together. Get the family and polarity right, respect the strip-length spec when soldering, and keep adapter chains short, and the connector stops being the mystery variable in every RF measurement you make.