Picking the wrong wireless module can sink your entire product before it ever reaches customers.
So I'm going to walk you through nine modules that seem tempting in tutorials but will cause serious problems in production, starting with number nine and working our way to the most critical one.
Click here to read or watch the full version
Module #9: HC-05 and HC-06
These are Bluetooth Classic modules built around something called Serial Port Profile, or SPP, and they're the wrong choice for almost any consumer product that talks to a smartphone. iOS doesn't natively expose SPP to third-party apps, and the HC-05 and HC-06 can't support Apple's MFi program at all, so there just isn't a path to iOS compatibility with these modules.
The market is also flooded with counterfeits, so two modules with the same marking can behave completely differently in the field. Most units sold have no legitimate FCC modular approval, which means you have no usable grant to inherit when it's time to certify your product.
Instead, use a real BLE module built on Nordic or Silicon Labs chips, with a verifiable FCC ID and datasheet-backed specs. I personally tend to lean toward Nordic, and my favorite vendor for Bluetooth products is Fanstel, though Raytac and Laird are solid options too.
Module #8: HM-10
The HM-10 is a BLE 4.0 era part with serious clone problems lurking behind it. The chip inside may or may not be the TI CC2540 or CC2541 the original design used, because the market is full of clones running completely different silicon.
The AT command set everyone designs around is proprietary to the original manufacturer's firmware, and clones don't always replicate it correctly. Picture this: you design around a specific set of AT commands, you scale to production, and your second batch behaves differently because the manufacturer sourced from a different clone supplier.
That problem shows up after you've already shipped. For BLE modules, Fanstel is my go-to recommendation, but Raytac and Laird are solid choices too, all built on modern Nordic nRF52 silicon with real certification.
Module #7: Legacy BLE Modules
Module #7 covers the whole category of legacy BLE modules, including parts like the Microchip RN4020, the Silicon Labs Bluegiga BLE112 and BLE113, and older TI CC2540-based modules. I used to love a lot of these modules, and I designed with them frequently, but that was quite a while ago, and they're all outdated now with better alternatives available.
They're BLE 4.0 or 4.1 era parts missing features that newer products often need, and many are marked not recommended for new designs or flat out discontinued with SDK support winding down. You design in something like the RN4020 because a tutorial used it or it was in an old BOM, and two years later you can't source the part and your firmware is tied to that module's stack.
Before you design in any BLE module, check what silicon is actually inside it, look at the vendor's lifecycle status, confirm which BLE spec version it supports, and see how actively the SDK is being maintained. For new designs, stick to current BLE 5.x modules built on modern silicon from a serious vendor.
Module #6: ESP-01
The ESP-01 is a Wi-Fi module built on Espressif's ESP8266 chip, and it's everywhere in tutorials and maker projects. It's great for learning Wi-Fi basics, but it's a terrible choice for a real product.
GPIO is minimal and the antenna is a poorly-tuned trace, but the real problem is that the entire form factor lacks an RF shield. An RF shield is a strict requirement for FCC modular approval, so no standard ESP-01, whether genuine or clone, can hold a valid modular FCC grant.
Even though Espressif's 15-year longevity commitment for the 8266 runs through 2029, there are better, more capable, more secure options available today.
For a new Wi-Fi product, skip the 01 and use a properly shielded module in the WROOM or MINI form factor built on modern ESP32 silicon like the S3, the C6, or the C3, which actually carry FCC modular approval and are well-documented for production.
Module #5: Generic nRF24L01
Module #5 is the generic nRF24L01, and I'm talking about the cheap clones flooding the market, not legitimate Nordic silicon used properly.
These modules use a proprietary 2.4 GHz protocol, not BLE, not Zigbee, not Thread, which means you're building an island that can't talk to anything else.
The clones have no inheritable approvals or trustworthy compliance documentation, the range claims on the listings are basically fiction, and there's no standardized ecosystem for interoperability.
If you're reaching for an nRF24L01, it usually means you haven't decided on a real protocol yet. Protocol selection is actually a bigger decision than module selection, so get the protocol right first and then pick the module to match.
For most point-to-point or small-mesh applications, plain BLE or BLE mesh is probably your best answer.
Module #4: Antenna and Certification Issues
This is a whole category of modules where the antenna and certification story falls apart in practice. This includes raw Semtech SX1276 or SX1262 breakouts and modules with an RF trace pin where the modular grant only stays valid if you replicate the vendor's exact microstrip layout and stackup, which beginners almost always deviate from.
It also includes modules with a U.FL or SMA connector where the FCC grant lists specific antenna types and maximum gains that integrators either ignore or don't bother to check before picking a completely different antenna.
The moment you deviate from a module's FCC grant antenna conditions, the modular approval no longer covers your integration.
That can mean full intentional radiator testing for the US alone running roughly eight to fifteen thousand dollars per radio, with multi-region compliance pushing past thirty thousand.
Antenna matching on RF trace pin modules usually requires a vector network analyzer and RF expertise that most product creators just don't have, plus impedance-controlled traces and proper antenna placement.
So the rule is, if a module doesn't ship with an integrated antenna or an FCC grant listing antenna types and gains you intend to use, where you must match the type and stay at or below the listed gain, you're buying yourself an RF project. Use modules with integrated PCB or chip antennas and valid modular FCC approval instead.
Module #3...
Click here to finish reading or watching the full version
Talk soon,
John
P.S. If you want help with wireless design decisions so you can avoid these pitfalls, you can get help from me and other experienced hardware experts inside the Hardware Academy. We cover exactly this kind of practical hardware selection and certification strategy.