Every wearable starts the same way. The industrial designer hands you a beautiful, curved enclosure the size of a matchbox. You look at the block diagram — SoC, PMIC, sensors, antenna, battery, display — and realise none of it fits on one flat board.
So you fold the electronics. And the moment you fold them, you have to decide: two rigid boards and a flex cable with connectors, or one rigid-flex PCB?
This post is about making that call honestly, and then designing the board so it survives a few thousand wrist flicks instead of cracking at the first fold.
Why Wearables Keep Ending Up on Rigid-Flex
A rigid-flex PCB is a single board where rigid sections (where your components live) are joined by flexible polyimide layers that are part of the same stack-up. No connectors between sections, no separate cable — the interconnect is laminated in.
For wearables, that matters because three things are always in short supply: volume, weight and reliability margin. Rigid-flex attacks all three at once. It also costs more and asks more of your layout discipline. So let’s be clear about when it pays for itself.
When Rigid-Flex Is Worth It (and When It Isn’t)
1. Space and 3D Packaging
Why it matters: Board-to-board connectors and their mating cables eat height and area, and in a wrist or patch device every millimetre of height is fought over.
Look for:
- Designs where boards must stack or wrap around a battery or sensor
- Multiple PCBs that need to sit at different angles inside a curved enclosure
- Connector height that is driving the overall product thickness
Real-world gotcha: A connector that looks tiny in the datasheet still needs mating clearance, a keep-out and an assembly tool path. Rigid-flex removes all of that.
2. Reliability
Why it matters: Connectors are one of the most common failure points in portable electronics. Vibration, drops, sweat and thermal cycling all work on contact interfaces. A laminated flex has no mating contacts to fret, loosen or corrode.
Real-world gotcha: Fewer connectors also means fewer solder joints and fewer places for an assembler to seat a cable at a slight angle — a classic intermittent fault that only shows up in the field.
3. Assembly and the Cost Crossover
Why it matters: At the bare-board level, rigid-flex is more expensive than rigid boards plus a cable. But total product cost includes connectors, cables, assembly labour, test time, yield loss and field returns.
Look for:
- Interconnects needing three or more connector pairs
- High pin-count or high-speed links between boards
- Products where a field failure is expensive (medical, safety, premium consumer)
Once those add up, rigid-flex often wins on total cost even though the PCB quote is higher.
When Flex + Connector Is the Better Choice
- Two boards, a low pin count, and plenty of room
- Early prototypes where the mechanical design is still changing every week
- Modules that must be replaceable in service
- Very cost-sensitive products with no space pressure
How to Get It Right
1. Start With the Stack-Up
Why it matters: The stack-up decides flexibility, impedance, cost and manufacturability before you route a single trace.
Look for:
- A symmetric stack-up, so the rigid sections don’t warp
- Flex layers placed in the middle of the rigid stack, so they exit the rigid section on the neutral axis
- Adhesiveless polyimide laminates: thinner, more flexible and more dimensionally stable than adhesive-based ones, with less Z-axis expansion stressing plated holes
- Rolled annealed (RA) copper in areas that flex repeatedly; it handles bending better than electrodeposited (ED) copper
- Unbonded (air-gap) flex layers where several flex layers must bend together
Fig 1. Typical rigid-flex stack-up: the flex layers run through the middle of the rigid sections, on the neutral axis.
2. Respect the Bend Radius
Why it matters: Bend too tightly and the copper work-hardens and cracks. It rarely fails on day one. It fails months later.
Commonly used rules of thumb, drawn from IPC-2223 guidance (always confirm with your fab):
- Static bend (folded once at assembly): about 6–10× the flex thickness for single- and double-layer flex, and 15–20× or more for multilayer flex
- Dynamic bend (flexes in use): stay single-layer if you can, and use a much larger radius, often 20–40× or more; many designers aim for 100×
Real-world gotcha: “Static” must really be static. If the strap or hinge moves with every wrist turn, design it as dynamic.
Fig 2. Bend radius rules of thumb from IPC-2223 guidance. Confirm the final numbers with your fab.
3. Keep the Bend Zone Boring
Why it matters: The bend zone should contain copper traces and nothing else.
- No vias, pads or components in the bend area
- Keep plated holes a safe distance from the rigid-to-flex transition (get the exact number from your fab)
- No sudden changes in trace width or copper density inside the bend
4. Route Like Copper Has Feelings
- Run traces perpendicular to the bend line
- Use curved traces and corners, not sharp 90° or 45° turns
- Stagger traces on adjacent layers instead of stacking them, to avoid an I-beam effect that stiffens the flex
- Add teardrops where traces meet pads and vias to spread the stress
- Use hatched copper planes in the flex instead of solid pours, to keep it flexible
Fig 3. Bend zone routing: what cracks and what survives.
5. Coverlay vs Flex Solder Mask
Why it matters: The flex area needs a protective layer that can bend without cracking.
- Coverlay (polyimide film with adhesive) is the standard for bend areas: tough and flexible
- Flexible photoimageable solder mask allows finer openings around fine-pitch parts, but it is less suited to repeated bending
Use coverlay in the bend and flex mask only where you need fine-pitch openings.
6. Use Stiffeners Wisely
Why it matters: Components, connectors and ZIF fingers on flex areas need local support.
- FR-4 or polyimide stiffeners under components on the flex
- Stiffeners under ZIF contacts to reach the connector’s specified thickness
- Stainless steel or aluminium where you also want heat spreading or a rigid mounting point
Real-world gotcha: Ending a stiffener exactly where the flex starts bending creates a stress point. Keep the stiffener edge clear of the bend.
7. Control Impedance Across the Transition
Why it matters: The dielectric thickness changes between rigid and flex sections, so a trace width that gives 50 Ω or 90 Ω differential in the rigid section will not give the same value in the flex.
Look for:
- Separate trace widths calculated for rigid and flex sections
- A continuous reference plane across the transition
- The effect of hatched ground on impedance (it raises it; your fab’s field solver should account for it)
- Impedance test coupons for both sections
8. Talk to Your Fab Early
Why it matters: Rigid-flex rules vary more between fabs than rigid PCB rules do.
Share your stack-up, bend locations and outline before layout, not after. Ask about material availability, minimum bend radius, transition clearances, coverlay registration and panel use. A paper mock-up or a 3D fold in your MCAD tool will catch most mechanical surprises early.
9. Test Like It’s Going on Someone’s Wrist
- Electrical test and impedance coupon checks (TDR) on every lot
- Bend-cycle testing for dynamic areas
- Thermal cycling, with cross-sections of plated holes near the transition
- Qualification to IPC-6013, the performance specification for flexible and rigid-flex boards
What Drives the Cost
- Total layer count and the number of flex layers
- Number of rigid sections and flex arms
- Adhesiveless materials and RA copper
- Panel use (odd flex outlines nest poorly on a panel)
- Stiffeners, impedance control, and air-gap constructions
Simplify the outline, minimise the flex layer count, and keep the rigid sections a standard thickness. Together these deliver the biggest savings.
Common Mistakes
- Placing vias or pads in the bend zone
- Treating a dynamic bend as static
- Using solid copper pours in the flex area
- Routing traces at an angle across the bend
- Forgetting that impedance changes between rigid and flex
- Leaving the fab out until the design is “finished”
- Skipping a physical fold mock-up
Final Thoughts
Rigid-flex is not a luxury. For many wearables it is the only way the product fits, and the most reliable way to keep it working. On our Wearable Ultra Compact Medical Device, a wrist-mounted Qualcomm SoC platform, a stacked 6-layer, impedance-controlled rigid-flex design is what made the compact form factor possible. But rigid-flex rewards discipline: plan the stack-up, respect the bend, keep the bend zone boring, and bring your fab in early.
Bend it on paper first. Your board will thank you.
Written by: Prit Varmora @ Ardra Lab
Designing a wearable that needs to fold into a tight space? Talk to our hardware team.