A custom IoT antenna should fit the device’s operating bands, ground plane, enclosure and mounting position. Judge the result by measured radiation efficiency, coverage in the required directions and energy per successful message. A small radiator or a good impedance match alone cannot establish battery life or reliable connectivity.
Start with the complete device
Supply the antenna designer with the PCB outline, stackup, battery, display, shields, cables, enclosure materials and installation orientation. Include the required regional bands and the modem’s antenna-port specification. A layout tuned on an open development board can change when installed inside the finished product.
Define the use cases separately. A sensor attached to a steel cabinet, a wearable against the body and a plastic gateway on a desk place different materials near the antenna. Measure representative configurations instead of applying one free-space result to all three.
For scale, the free-space wavelength at 2.4 GHz is about 125 mm, and a quarter wavelength is about 31.2 mm. At 868 MHz, those values are about 345 mm and 86.4 mm. These calculated lengths describe electrical scale; they are not mandatory dimensions for every antenna design.
Choose a format by its integration requirements
| Format | What to check in the device |
|---|---|
| PCB trace | Board-edge placement, copper clearance, ground dimensions, stackup and the cost of changing the main PCB. |
| Ceramic chip | The manufacturer’s reference layout, required ground plane, matching network and nearby components. Package size excludes its surrounding keep-out area. |
| FPC | Installed position, adhesive, bends, cable routing and separation from the battery or metal enclosure. |
| LDS or formed conductor | Housing geometry, dielectric properties, feed contacts, dimensional tolerance and production inspection. |
| External antenna | Connector, cable loss, strain relief, mounting surface and exposure to the operating environment. |
No format has a fixed efficiency advantage over the others. Compare candidates in the same enclosure, at the same frequencies and with the same measurement definition. Include the ground plane and clearance volume when comparing size.
For an external RF connection, select coaxial cable connectors for the cable type, frequency and mechanical arrangement. Check the installed cable loss; it can consume part of the improvement obtained from the antenna.
Separate impedance matching from radiation efficiency
S11 in decibels is normally negative for a passive antenna. Return loss uses the opposite sign: S11 = −10 dB corresponds to a return loss of 10 dB. State which quantity a specification uses.
The reflected fraction of incident power is 10S11(dB)/10. The following values are calculated at one reference plane:
| S11 | Return loss | Reflected power | Mismatch loss |
|---|---|---|---|
| −6 dB | 6 dB | 25.1% | 1.26 dB |
| −10 dB | 10 dB | 10.0% | 0.46 dB |
| −15 dB | 15 dB | 3.16% | 0.14 dB |
| −20 dB | 20 dB | 1.00% | 0.044 dB |
Radiation efficiency describes how much of the power accepted by the antenna is radiated. Total efficiency also includes input mismatch. For a single-port antenna at the stated reference plane, total efficiency equals radiation efficiency multiplied by the accepted-power fraction.[1]
For example, 60% radiation efficiency and S11 = −10 dB give 0.60 × 0.90 = 54% total efficiency. Improving S11 to −20 dB while leaving radiation efficiency unchanged gives 59.4%, an improvement of about 0.41 dB. That is not a measured battery-life gain.
Match the antenna in its final assembly. Calibrate the vector network analyzer to the intended reference plane, account for fixtures and check whether the measurement cable disturbs the antenna currents. 50 Ω RF adapters used in the test path must suit the connector and frequency range; adding an adapter can change the measurement unless it is included in the calibration or correction.
Calculate power consumption from the operating cycle
Antenna improvements may reduce retries, time spent searching for service or the transmit power needed for a link. The modem, protocol and network determine whether those changes occur. Do not convert an antenna-efficiency percentage directly into the same percentage of battery savings.
Measure charge per successful application message, including wake-up, network access, transmission, receive windows, retries and the return to sleep. Keep payload, reporting interval, firmware and network conditions consistent when comparing antennas.
For a simple two-state calculation, average current is:
Iavg = [Iactivetactive + Isleep(T − tactive)] / T
Consider an illustrative 600-second cycle with 120 mA active current for 2 seconds and 0.010 mA during the remaining time. Average current is about 0.410 mA. Dividing a nominal 2500 mAh capacity by that current gives about 254 days. If measured active time falls to 1 second with all other assumptions unchanged, the calculation becomes about 0.210 mA and 496 days.
Those are idealized arithmetic results, not product-life predictions. The calculation excludes self-discharge, capacity loss with temperature, pulse-load limits, battery cutoff and other device loads. Use measured current traces and usable battery capacity for a design decision.
For cellular IoT, review the network’s supported power-saving behavior alongside the antenna. PSM can retain registration while the device is unavailable for normal paging; eDRX lengthens sleep intervals between monitoring opportunities. These mechanisms trade immediate downlink availability against energy consumption.[2]
Check connectivity in the required directions
Peak gain describes one direction. A device with high peak gain can still have a weak direction that matters after installation. Review three-dimensional patterns or representative orientation tests, including the intended mounting surface and any body loading.
Doubling total antenna efficiency from 20% to 40% adds about 3 dB to the link budget when the other factors stay unchanged. It does not establish a fixed reduction in packet errors or guarantee service inside a closed metal cabinet. Check the complete path, receiver performance and available signal margin.
Measure reception while other radios and noisy circuits are active. Compare packet success, receive sensitivity and current with the interfering subsystem on and off. Include cellular transmission, Wi-Fi traffic, display activity, DC-DC converters and relevant processor operating modes.
Antenna separation, polarization and decoupling structures are design options, not fixed isolation guarantees. In a multi-antenna device, review efficiency, coupling, correlation and OTA performance together. A low envelope correlation coefficient alone cannot promise a particular MIMO data rate.
Keep NFC and far-field antenna design separate
NFC operates at 13.56 MHz and uses close-range coupling.[3] A ferrite layer selected for an NFC coil near metal is not evidence of improved performance for a 700 MHz cellular antenna or a 2.4 GHz radiator. Select materials using their properties at the actual operating frequency and verify the assembled antenna.
Likewise, an LDS process setting, substrate trade name or bend-radius claim does not establish RF performance. Obtain the selected material’s frequency-dependent properties and the supplier’s mechanical limits; test the shape and assembly that will enter production.
Specify the evidence required before production
Request full-band impedance data, radiation or total efficiency with the definition stated, patterns in the relevant configurations, and system-level transmit and receive results. CTIA’s OTA programs distinguish single-antenna, MIMO and converged cellular/Wi-Fi testing; their purpose extends beyond an impedance sweep.[4]
- Identify the hardware revision, antenna location, matching values and cable arrangement used for each result.
- Record measurement uncertainty and distinguish simulation from prototype measurements.
- Repeat relevant checks after enclosure, battery, cable or mounting changes.
- Compare several production-representative samples to expose assembly and material variation.
- Agree on the applicable operator, regulatory and product test requirements separately; good S11 is not a compliance certificate.
Release the antenna with a controlled drawing, allowed materials and assembly positions, matching-component values and measurable acceptance limits. Retain the current-consumption traces used for any battery-life statement.
Need to define the RF connection or antenna test interface?
Send Dolph the operating bands, connector and cable details, mounting constraints and available test results. Identify whether the requirement concerns the antenna, an RF adapter or the measurement path so the applicable component scope can be confirmed.
