Marketing names blur three different products. A Bluetooth adapter adds wireless capability to a host that lacks it. A Bluetooth receiver accepts wireless audio and drives a wired speaker path. A Bluetooth transmitter does the reverse: it encodes a wired source and radiates to headsets or speakers. Product managers who treat those labels as synonyms force the wrong PCB architecture, BOM, and test plan.
For overseas teams building or sourcing these boards in China, the honest PCB angle is RF layout, audio integrity, module choice, and functional test--not another glossary of retail dongles.
Signal direction sets the block diagram
| Role | Dominant path | Hardware emphasis |
|---|---|---|
| Adapter | Bidirectional host link | Host interface (USB/UART), SoC or module, antenna, ESD, firmware access |
| Receiver | RF in, analog/digital audio out | Antenna, Bluetooth device, DAC/codec, low-noise analog routing, output connectors |
| Transmitter | Wired audio in, RF out | Input interface, ADC/codec, latency/codec support, RF path, antenna |
Dual-mode "2-in-1" boards implement both audio directions plus a mode switch. Complexity jumps: shared antenna, shared codec, firmware modes, and a test matrix that must exercise TX, RX, and switching--not a single happy-path pairing demo.
Adapter boards: treat them as tiny RF systems
USB or UART adapters fail more often from power noise, ESD at the connector, and antenna keep-out violations than from "wrong Bluetooth version" marketing. Local decoupling, continuous return under RF, and datasheet keep-outs around PCB or chip antennas are non-negotiable even when the RF silicon sits inside a certified module.
Host interface choice changes ESD strategy and mechanical stress on connectors. Firmware pads and test points should exist before the enclosure freezes.
Receiver and transmitter: audio is half the board
Receiver layouts lose perceived quality when DAC outputs share return paths with DC-DC switchers or sit under RF matching. Keep analog away from clocks, USB pairs, and switching regulators. Transmitter designs add input conditioning and codec/latency requirements; TV and gaming use cases need latency verification with the actual headset ecosystem, not a bench speaker that hides delay.
Codec marketing (SBC, AAC, aptX-family, LDAC, and others) is a system claim. Both ends must support the mode you advertise, and PCBA test should prove it when latency or quality is a selling point.
Module procurement realities for China builds
Bare SoCs look cheaper on a spreadsheet. They move RF matching, antenna tuning, and often certification burden onto your team. Pre-certified modules cost more per unit but shrink NPI risk and protect antenna keep-out rules already validated by the module maker--if you follow their layout notes.
Do not casually substitute RF matching parts or antennas during purchasing "equivalents." Those substitutions change range and spurious behavior. Lock MPNs for the RF section the same way you lock a crystal.
Layout and assembly notes that survive fab quotes
Prefer short 50-ohm RF routes, unbroken reference under the RF section, and antenna placement at the board edge away from batteries, shields, and metal connectors. Four-layer stackups usually pay for themselves on receiver/transmitter boards. SMT must respect module reflow profiles; AOI plus pairing/range/audio functional tests beat visual inspection alone.
Related reading on RF-minded board work: antenna feed network PCB design. When you are ready to quote modules, BOM, and stackup together, use how to place an order.
Name the product by signal direction first. The PCB and the test plan follow that decision; the retail label does not.