June 29, 2026
Board type is rarely a preference. Wearables, folding mechanisms and high-power LEDs each force a particular substrate — and each of those narrows the list of factories that will take a small batch.
Choosing rigid versus flex is rarely a process preference; the product form decides it. Wearables with minimal internal space and body-contouring surfaces need flex. Folding mechanisms, lifting screens and any moving connection make rigid-flex close to unavoidable. LED lighting and switching power supplies need metal-core boards to move heat to the back surface. RF and high-speed links depend on the dielectric choices of high-frequency material, which drive signal integrity directly. Choose the wrong board type and every later design effort is compensating for the first mistake.
What matters here is that board-type selection looks like a design-stage decision, but it directly determines whether anyone will take your order. The narrower the process, the fewer factories that can do it; the smaller the quantity, the fewer factories willing to adapt tooling and curves. Stack the two together and special board plus small batch becomes the hardest category of order to place in this industry.
Flex and rigid-flex boards have a much narrower placement window than rigid. Flexible substrates deform during solder printing and need dedicated support fixtures. Reflow profiles must be set separately — too hot and the board wrinkles, too cold and joints go cold. Incoming dimensional stability of the substrate makes the locating reference hard to choose. Metal-core boards face different problems: heat draws away quickly and solder joints cool at a different rate, which changes wetting behaviour.
All of that adaptation takes fixtures, takes curve development and takes line time. For a high-volume plant, stopping a standard line to develop for one special board is uneconomic unless the order volume amortises it. So most lines only take standard rigid boards and decline flex and metal-core outright — not because they cannot, but because it does not schedule. In that filtering process a small-batch order has almost no negotiating room, and the usual outcomes are a compromised board design or a long detour through a rigid-board workaround.
Our plant is unusual here: flex, rigid, rigid-flex, metal-core and high-frequency boards all sit inside its process envelope, single-sided and panelised alike; it places components down to 1005 and 03015 packages and inspects down to 01005, with micro components and flexible boards completed in the same plant. Solder printing, SPI, reflow, wave soldering, conformal coating and full assembly are all under its own control, with data from material to finished product held in its own system.
It makes this routine rather than exceptional because small and mixed volume is its main business — frequent changeovers and process adaptation are amortised by order density. For the customer, that means a flex board does not have to be split across three factories, and a sound board-type decision does not have to be compromised to fit someone's capacity. Process breadth is not a party trick; it is what keeps your design choices free.
If you have a special-board requirement, a few specific questions at quotation stage go a long way. What support fixture is used for flex boards, and who sets the temperature curve? What is the solder-joint acceptance standard on metal-core? What is the placement and inspection capability for 01005-class components? What coating, film thickness and verification method for conformal coating? The more specific the answers, the more real the process experience; a flat yes usually means limited practice.
The quicker route is to look at its process list and active cases. On our side, drawings or sample requirements submitted return a solution and quotation within twenty-four hours, with process boundaries and the adaptation approach set out in writing — whether it can be done, how, and what to watch for. Knowing that before the line runs is worth considerably more than discovering it at the line.
Flex, rigid-flex and metal-core boards each move a different constraint. Flex design is limited by bend radius and by where copper can cross the fold; rigid-flex adds the question of how many layers are allowed to cross the flex-to-rigid transition; metal-core boards are about thermal path and the handling damage that thick substrates invite. A supplier who answers a general enquiry with a general capability sheet is not answering the question. The useful version is specific: minimum inner bend radius for your stack-up, the number of signal layers permitted across the transition, and the thermal resistance per square of the metal substrate you have in mind.
Small volume is exactly where these questions get skipped, because the volume does not justify long engineering discussions. But the failure modes are design failures, not volume failures — a board that cracks at the fold fails identically at fifty units and fifty thousand. We state the process limits against the design before quoting, and flags where the design has to move rather than adapting the spec silently.
On flex, rigid-flex and metal-core boards, a high quote on a small order rarely reflects the process itself. It reflects setup spread across a few dozen units: a dedicated fixture, a tuned thermal profile, an adapted transfer mode. These one-off costs are invisible on a large order and can account for a third of the price on a small one.
So design choices drive the cost. The same flex board can avoid an entire fixture set by converting a dynamic bend to static, opening the bend radius by half a millimetre, or replacing a metal core with thickened local copper. Each of those is a concession made in design, and each saves money silently.
At enquiry stage, open the process constraints up: which requirements are hard, and which are negotiable. A supplier willing to take small special-board orders will usually help with that conversation. One that replies can do it and then builds every fixture for the strictest reading of the spec will certainly be expensive.
Tell us your project scenario and requirements, and we will match the right solution and provide a quote within 24 hours.
Get a Free QuotePCBA assembly, SMT placement and THT soldering under one roof: 95,000 CPH capacity that also takes 50-board engineering builds, placement down to 1005/03015 packages, components sourced through authorised channels with a zero-counterfeit commitment and released to the line only after incoming inspection, and IATF 16949:2016 (TS-1051701) plus ISO9001 and RoHS certification. Sampling, small batch and volume production run by the same people on the same parameters.
Explore the Electronics Manufacturing Solution →