The most expensive mistake in a wireless product isn't the radio design — it's deciding to design your own radio at all when you didn't need to. FCC intentional radiator certification is where that decision shows up on the invoice, and most hardware teams underestimate it by an order of magnitude.
Three ways to put a radio in your product
When you ship anything that intentionally transmits RF — BLE, Wi-Fi, LoRa, sub-GHz — the FCC treats it as an intentional radiator under Part 15. You have three paths:
- Chip-down design. You lay out the radio IC, the matching network, and the antenna yourself. Maximum control, lowest BOM cost at volume, and a full intentional-radiator test you own completely.
- Pre-certified module with chip antenna. You drop in a module that already holds an FCC ID.
- Pre-certified module with an external antenna from the module's approved antenna list.
The difference between path 1 and paths 2–3 is thousands of dollars and several weeks.
Why a pre-certified module saves the full test
A pre-certified module carries modular approval — the FCC has already tested the radio against the emissions limits. When you integrate it under the module's grant conditions (correct antenna gain, proper trace layout, the right shielding), you inherit that grant. You still do a small amount of testing — unintentional radiator / Part 15B for your host product — but you skip the expensive intentional-radiator transmitter measurements.
A full intentional-radiator test for a custom radio can run $10,000–20,000 and add 4–8 weeks. Riding a module's modular approval can cut the RF portion of that to a fraction. For a startup shipping a few thousand units, the chip-down BOM savings of a dollar or two per unit rarely beats the certification cost and schedule risk.
The antenna is where modular approval gets broken
Here's the trap: modular approval is only valid if you respect the module's antenna conditions. The grant specifies a maximum antenna gain and often a specific antenna or antenna type. Bolt on a higher-gain external antenna and you've voided the modular grant — now you're back to a full permissive change filing or a fresh intentional-radiator test.
Concrete example: a 200-unit industrial gateway used a pre-certified sub-GHz module rated for a 2 dBi antenna. The first mechanical prototype shipped with a 5 dBi whip because it "fit the enclosure better." That single swap would have forced a Class II permissive change. Catching it at the sample stage — before tooling — kept the project on the module's existing grant and saved the full re-test.
Tuning happens at the factory, not in CAD
Even with a module, your antenna performance depends on the final enclosure, ground plane, and nearby metal. A chip antenna tuned on the bare board will detune once it's inside a plastic case next to a battery. So real tuning happens on production-representative samples at the factory: measure return loss (aim for better than -10 dB at your band), adjust the matching network, and re-measure on units that match what ships.
If you don't have RF engineers near the line, this is exactly the kind of thing that benefits from an engineering-led partner on the ground. An agent like China Sourcing Agents, run by a hardware engineer who reads the schematic and checks the antenna placement against the module's grant conditions, can stop a 5 dBi-instead-of-2 dBi mistake before it costs you a re-certification.
Actionable checklist
Before you commit to a radio path:
- Default to a pre-certified module unless your volume genuinely justifies a chip-down design.
- Pull the module's FCC grant and read the antenna gain and type limits before mechanical design.
- Pick your antenna from the approved list; never exceed the rated gain.
- Tune and verify return loss on production-representative samples inside the real enclosure.
- Budget Part 15B host testing even when you ride a module's grant.
Get the path right at the start and intentional-radiator certification stops being the scary line item on your bring-up schedule.
Top comments (0)