Panel For Example Panel For Example Panel For Example

PCB Design Considerations for Wearable Devices

Author : Adrian April 07, 2026

 

Overview

Because wearable devices are very small, there is no single standard printed circuit board for the growing Internet of Things market. Until standards emerge, designers rely on board-level experience and manufacturing practices and consider how to apply them to the unique challenges of wearables. Three areas require special attention: board substrate materials, RF and microwave design, and RF transmission lines.

 

PCB Materials

PCB stackups are typically built from layers of materials such as fiber-reinforced epoxy (FR4), polyimide, Rogers materials, or other laminates. The insulating material between layers is called prepreg.

Wearable devices demand high reliability, so designers often face a choice between FR4 (the most cost-effective PCB material) and more advanced, more expensive materials.

If a wearable application requires higher speed or higher frequency performance, FR4 may not be optimal. The dielectric constant (Dk) of FR4 is about 4.5, Rogers 4000-series materials are around 3.55, and Rogers 4350 is about 3.66.

Multilayer PCB stackup showing FR4 and Rogers 4350 and core thickness

Figure 1: Multilayer PCB stackup illustrating FR4, Rogers 4350, and core thickness.

The effective dielectric constant of a stackup is the ratio of capacitance or stored energy between a pair of conductors in the stackup to the capacitance or energy between the same conductors in a vacuum. At high frequencies, lower loss is preferred; therefore Rogers 4350 with Dk 3.66 is better suited for higher-frequency applications than FR4 with Dk 4.5.

Typical wearable PCBs range from 4 to 8 layers. For an 8-layer board, the common practice is to provide sufficient ground and power planes and to sandwich routing layers between them. This reduces crosstalk ripple effects and significantly lowers electromagnetic interference (EMI).

During layout, large ground planes are usually placed adjacent to power distribution layers. This arrangement minimizes ripple and reduces system noise close to zero, which is especially important for RF subsystems.

Compared with Rogers, FR4 has a higher dissipation factor (Df), particularly at high frequencies. High-performance FR4 stackups may have Df around 0.002, an order of magnitude better than standard FR4, while Rogers stackups can have Df of 0.001 or lower. Using FR4 in high-frequency applications produces noticeable differences in insertion loss. Insertion loss is defined as the power loss when a signal transmits from point A to point B using FR4, Rogers, or other materials.

 

Manufacturing Issues

Wearable PCBs require tighter impedance control to ensure cleaner signal transmission. Historically, trace tolerance was ±10%, which is insufficient for modern high-frequency, high-speed designs. Current requirements are typically ±7%, and in some cases ±5% or tighter. These tolerances and other variables make manufacturing controlled-impedance wearable PCBs more challenging and reduce the number of suppliers capable of producing them.

Rogers high-frequency stackups usually maintain dielectric constant tolerance around ±2%, with some products reaching ±1%. In contrast, FR4 stackup Dk tolerances can be as high as ±10%. Consequently, Rogers offers much lower insertion loss. Transmission loss and insertion loss in Rogers stackups can be about half that of conventional FR4.

Cost is often a primary concern. Rogers can provide relatively low-loss, high-frequency stackup performance at acceptable prices. For commercial applications, hybrid PCBs combining Rogers layers and epoxy-based FR4 layers are common: some layers use Rogers, while others use FR4.

When selecting Rogers, frequency is the main consideration. Above roughly 500 MHz, designers tend to choose Rogers materials, especially for RF and microwave circuits where strict impedance control is required.

Compared with FR4, Rogers offers lower dielectric loss and a more stable dielectric constant over a broad frequency range, delivering the low insertion loss needed for high-frequency operation.

Rogers 4000-series materials have a coefficient of thermal expansion (CTE) and dimensional stability superior to FR4. This helps maintain stable board dimensions when the PCB undergoes cold, hot, and high-temperature reflow cycles at higher frequencies and temperature cycling.

In hybrid stackups, standard manufacturing processes can mix Rogers and high-performance FR4, enabling good yield without special via preparation steps for Rogers stackups.

Standard FR4 may not meet stringent electrical performance for some applications, but high-performance FR4 with higher glass transition temperature (Tg) offers reliable properties, relatively low cost, and applicability across designs from simple audio to complex microwave.

 

RF and Microwave Design Considerations

Portable technologies and Bluetooth have enabled RF and microwave functions in wearable devices. Operating frequencies are becoming increasingly diverse. A few years ago, very high frequency ranges were treated differently, but now ultra-high-frequency applications span from about 10 GHz up to 25 GHz.

For wearable PCBs, RF sections require careful routing, keeping high-frequency traces isolated and routed away from ground where appropriate. Other considerations include providing bypass filters, sufficient decoupling capacitors, proper grounding, and designing transmission and return paths to be closely matched.

Bypass filters suppress noise content and crosstalk ripple. Decoupling capacitors should be placed as close as possible to the power pins of the devices they support.

High-speed transmission lines and signal return networks require a dedicated ground plane between power and signal layers to smooth jitter caused by noise. At higher signal speeds, small impedance mismatches cause unbalanced transmission and reception and result in distortion. Therefore, impedance matching for RF signals demands close attention because RF signals operate at high velocity with tight tolerances.

 

RF Transmission Lines

RF transmission lines require controlled impedance to carry RF signals from a specific IC substrate onto the PCB. These transmission lines can be routed on outer layers, top and bottom layers, or internal layers.

Common transmission line structures used in PCB RF layout include microstrip, suspended stripline, coplanar waveguide, or grounded variants. A microstrip consists of a conductor trace of fixed length and a full or partial ground plane directly beneath it. Characteristic impedances for typical microstrip structures range from 50 Ω to 75 Ω.

Coplanar waveguide providing better isolation between RF traces and nearby routing

Figure 2: Coplanar waveguide can provide better isolation between RF lines and nearby traces.

Suspended stripline is another routing and noise-suppression method. This structure uses a fixed-width trace on an inner layer with large ground planes above and below the center conductor. With the ground plane sandwiched in the middle of the power layers, suspended stripline provides very effective grounding and is often preferred for RF routing on wearable PCBs.

Coplanar waveguide offers improved isolation when traces need to be routed close to each other. This geometry consists of a center conductor with ground planes on both sides or below. The preferred methods for transmitting RF signals are suspended stripline or coplanar waveguide, as both provide superior isolation between signal traces and RF routing.

For coplanar waveguides, it is recommended to use a via fence on both sides of the trace. A via fence is a row of grounded vias placed along each metal ground region adjacent to the center conductor. The main trace running between fences then has a short return path to the underlying ground layer on each side. This approach reduces noise levels associated with RF ripple effects. The dielectric constant of 4.5 corresponds to prepreg FR4 materials, while the prepreg dielectric constant for microstrip, stripline, or offset stripline is around 3.8 to 3.9.

Via fence recommended on both sides of a coplanar waveguide

Figure 3: Via fence recommended on both sides of a coplanar waveguide.

In some designs that use extensive ground planes, blind vias may be used to improve decoupling performance and provide a shunt path from the device to ground. Shortening the via length achieves two goals: it creates a shunt to ground and reduces the transmission distance for small ground islands, which is an important RF design consideration.

Source: Electronic Engineering Journal