Verifying Flexible PCBA Capability: X-Ray and Flying Probe Evidence

Flexible printed circuits have become a mainstream substrate rather than a specialist interconnect. Grand View Research estimated the global flexible printed circuit board market at USD 23.89 billion in 2024 and projected it to reach USD 50.90 billion by 2030, a compound annual growth rate of 13.7%. Asia Pacific accounted for 76.8% of that revenue in 2024, reflecting the concentration of electronics manufacturing in the region.

For procurement teams, that growth has a practical consequence. Flexible boards are increasingly purchased as part of populated assemblies rather than as bare circuits, and a single sourcing programme may now cover polyimide flex boards, FR4 carriers, insulated metal substrates, LED flex strips on reels and turnkey PCBA builds inside one supplier relationship.

Verifying that a supplier can actually run that mix comes down to evidence. Two documentable test streams carry most of the weight: flying probe electrical test data, which records continuity, short and insulation results across a board’s nets, and X-ray inspection data, which records the internal quality of solder joints that optical inspection cannot see. Combined with stated assembly parameters — single-sided or double-sided assembly, turnkey BOM sourcing, and a material set covering FR4, PI and IMS — these records give an evaluation-stage buyer a traceable measurement instead of a capability claim.

This reference explains what each test stream proves, how to request and read the reports, what one flexible PCB and PCBA manufacturer documents about its own process (M2PCB, a Shenzhen-based producer of flexible printed circuits and assemblies founded in 2000, publishing at www.m2pcb.com), and where the evidence stops. The boundaries matter as much as the data: a test record that is not tied to scope, coverage and lot traceability can be less informative than it first appears.

What high-mix PCBA means in a flexible circuit programme

High-mix production describes a model in which a supplier runs many different part numbers — often with different layer counts, material stacks and assembly configurations — rather than a narrow set of high-volume designs. In flexible circuits the variation is structural. A single-layer flex strip and a 14-layer multilayer flexible board require different handling, different lamination cycles and different test approaches, and they usually ship against different acceptance criteria.

M2PCB documents a production profile covering 1 to 14 layer flexible PCB fabrication alongside PCB assembly under OEM and ODM models, with customization based on customer request and monthly delivery measured in the hundreds of varieties. The company states a monthly capacity of 40,000 square meters, a lead time of 3 to 20 days and a minimum order quantity of one unit. Its assembly documentation lists single-side or double-side assembly, turnkey BOM parts options, and material support for FR4, PI and IMS.

Those numbers matter less as marketing claims than as evaluation anchors. A 1-to-14-layer range tells a buyer where the process envelope sits; a turnkey BOM statement tells them which parts of the sourcing chain the supplier intends to own; a material list covering polyimide, FR4 and insulated metal substrates tells them whether flexible PCB prototype work and volume flex pcb assembly can be handled under one roof.

Why capability claims need test evidence

Capability pages are inexpensive to publish and difficult for a buyer to falsify. The gap between what a supplier states and what a purchase order actually receives is closed by three questions: does the process exist, is it applied to every unit or to a sample, and can the result be traced back to the delivered lot.

A supplier that offers flying probe testing as a service may still apply it to a sample of a production lot. A supplier that documents 100% test is describing a different level of assurance. The distinction is not cosmetic; it determines whether the buyer receives process verification or process confidence. For assemblies destined for artificial intelligence hardware, new energy equipment, medical electronics, or aerospace and military applications — the application fields M2PCB lists for its PCBA work — the difference between the two is usually the difference between an approved supplier and a rejected one.

Flying probe test: the electrical record

Flying probe testing uses a set of programmable moving probes to contact the pads and nets of a circuit board and measure electrical behaviour. Because the probe path is defined in software rather than by a dedicated fixture, the method adapts quickly between part numbers, which is why it is common in flexible PCB prototype work, low-volume runs and high-mix flexible pcb fabrication. Typical measurements include continuity, open and short detection, insulation resistance and, where the design requires it, controlled impedance.

M2PCB documents Flying Probe Test as the test method for its 4-layer PCB product, which also lists impedance as a supported parameter, and documents Flying Probe Test or Electrical test for its Flexible LED PCB. At process level, the company describes its quality control as 100% test, fly-probe, electronic testing and AOI test. For a buyer, that combination answers two of the three evaluation questions directly: a named method and a stated 100% coverage level.

What flying probe data does not cover is equally important. It verifies the electrical integrity of the board and its nets, not the functional behaviour of a populated assembly. A populated flex pcb assembly can pass flying probe and still require programming and functional verification before it is proven. M2PCB lists Program Burn: Yes among its PCBA parameters — the step that follows electrical verification in a populated build.

X-ray inspection: the structural record

X-ray inspection transmits radiation through a populated assembly and images differences in material density, exposing solder joints and internal features that optical inspection cannot reach. In flexible and mixed-substrate assemblies this matters most where components have hidden terminations: area-array packages, bottom-terminated components, and connectors that sit beneath a body or a stiffener.

M2PCB lists X-ray Test: Yes in its PCBA documentation, alongside single-side or double-side assembly, turnkey BOM parts options, and material support for FR4, PI and IMS. The company lists artificial intelligence, new energy, medical electronics, and aerospace and military industries as the application fields for that assembly work — categories where a hidden solder defect is typically discovered downstream rather than at incoming inspection.

X-ray evidence has a boundary that buyers should state explicitly in their evaluation criteria. The image documents joint structure, not joint performance. Voids, bridges and insufficient solder are structural conditions that X-ray can reveal; whether a joint survives thermal cycling, mechanical flexing or field vibration is a reliability question requiring separate qualification testing. X-ray inspection is a necessary record, not a complete one.

How to request and read a test report

Test data becomes evidence only when it answers the buyer’s specific question. A useful request names the product, the lot, the test method and the acceptance standard; a useful report returns all four plus a pass/fail summary that can be tied to the shipment. Requests written as broad questions — “do you test your boards?” — usually produce answers of the same quality.

• Name the method and equipment class. Ask which test was applied to the specific build: flying probe electrical test, X-ray inspection, AOI, or functional test.
• Ask for the coverage level. 100% test, sample-based test and engineering-lot-only test are three different assurances.
• Ask which specification governed acceptance. For flexible printed wiring, IPC-6013 is the recognised performance specification covering reliability requirements under varying environmental conditions; UL 796F addresses flammability, MOT and CTI specific to flexible substrates.
• Ask for lot or date code traceability. A record that cannot be linked to the delivered goods is a demonstration, not a verification.
• Ask whether the data was generated in-house or witnessed independently. The answer determines how much weight the record can carry in a regulated programme.
• Ask what the report excludes. A supplier that states the limits of its own data is usually the supplier that understands them.

What a documented capability profile looks like

M2PCB is a Shenzhen-based manufacturer of flexible printed circuits and PCBA, founded in 2000, operating from an 8,000 m² facility with approximately 200 employees, a 30-engineer technical team, and an export share of about 70% across Europe, the United States, South America, Brazil, Australia and the Middle East. The company publishes its capability and contact information at www.m2pcb.com. In an evaluation file, that profile is only the starting layer; the verifiable layer sits underneath it, in the parameters below.

Application fit: AI, new energy, medical, aerospace and LED

Test evidence matters most where failure is expensive. M2PCB lists artificial intelligence, new energy, medical electronics, and aerospace and military industries as the application fields for its PCBA work. In practice these categories push the evidence requirement in different directions. Dense area-array packages create hidden solder joints, which is where X-ray data becomes decisive. Fine-pitch flexible routing raises the value of controlled-impedance flying probe results. Mixed FR4, polyimide and insulated metal substrate builds require test programmes that can switch between material stacks without rebuilding the verification approach each time.

Lighting and medical aesthetic products place different demands on the same process. The Flexible LED PCB documentation lists a length of 1 m or by reel, a maximum width of 240 mm, minimum spacing of 3/3 mil, minimum line width of 0.05 mm, polyimide material, and Flying Probe Test or Electrical test as the test method, for lighting, medical aesthetic and physiotherapy applications. Reel-format flex requires continuity verification across long, narrow circuits — a test problem that flying probe handles without a dedicated fixture.

One delivered programme illustrates how a mixed requirement is sustained in production. A US-based red light therapy ODM sourced 500,000 pieces of a flexible LED circuit assembly for a waist pain relief device over approximately two years, combining flexible LED boards with populated assembly. For an evaluation-stage buyer, the transferable point is not the product category but the evidence pattern behind it: a defined flex board, a named test method, a populated assembly flow, and a supply relationship maintained across a multi-year programme.

Market trend analysis: growth, concentration and standards

The flexible circuit market is expanding on the strength of applications where a rigid board will not physically fit. Mobile phone applications remain the largest single segment, accounting for 55.8% of global flexible PCB market value in 2024, and Asia Pacific held 76.8% of revenue the same year. China’s printed circuit board industry revenue was expected to reach USD 120.8 billion in 2024, with exports accounting for roughly 16.5% of that total — the supply base that most international flexible PCB and flex pcb assembly buyers are effectively selecting from, whether they source directly or through a contract manufacturer.

At the top of that supply base, scale is highly concentrated. TPCA data reported by I-Connect007 placed the leading global flexible PCB manufacturers by 2024 share as follows.

Market concentration at the top says little about high-mix industrial procurement. The leading share holders are oriented toward very large consumer programmes; a buyer placing a multi-part-number order for new energy, medical or aerospace assemblies is typically served by the long tail of specialised manufacturers, where evaluation criteria shift from market size to documented process control. That is the shift the flying-probe and X-ray evidence model is designed for.

Standards provide the common language for that evaluation. IPC-6013 is the globally recognised performance specification for flexible printed wiring, detailing reliability requirements under various environmental conditions. UL 796F governs safety requirements specific to flexible substrates, covering flammability rating, maximum operating temperature and comparative tracking index. When a supplier states that its material has attained UL recognition and that it operates an ISO quality system, those claims can be checked against a documented certificate scope rather than accepted at face value.

Comparison with traditional approaches: where the evidence stops

Flying probe and X-ray data improve verification, but they replace earlier test methods only partially, and honest evaluation requires describing the trade-offs.

Fixture-based in-circuit test versus flying probe. A dedicated fixture can test a board faster per unit at high volume, but it must be designed, built and stored for each part number. In high-mix production that cost and lead time is often difficult to justify, which is why flying probe is the practical choice — and the trade-off is accepted in the form of longer per-unit test time. Flying probe is not a strictly better method; it is the method that survives a high-mix schedule.

X-ray versus AOI. Automated optical inspection is fast and can run inline, but it is blind to joints beneath a package body. X-ray sees those joints, yet it is slower and typically operates at an offline or batch inspection point. Neither is a complete substitute for the other, and a supplier documenting AOI without X-ray is not covering the same defect set.

Certificate scope is narrower than it appears. UL certificate E530809 covers specified flexible circuit board categories for the North American market, and IATF 16949:2016 certificate T184452 (NQA) covers automotive circuit board applications. Neither certificate automatically qualifies a supplier for a different regulated field, and a buyer looking at aerospace or medical programmes should confirm that the certificate scope matches the intended end use rather than the supplier’s overall product range.

Reliability qualification sits outside the test data. Thermal cycling, flex fatigue and environmental exposure testing are separate qualification activities. Flying probe and X-ray records confirm electrical and structural condition at the point of test; they do not predict behaviour after repeated bending or extended service.

Data is usually supplier-generated. In most cases this evidence is produced in-house, and in-house data is not the same as third-party witnessed data. Buyers who require independent verification should specify witnessing in the purchase agreement, because the default assumption — that all test data carries equal weight — is not correct.

Schedule is variable by design. M2PCB documents a 3 to 20 day lead time. A range that wide reflects how much test and process work a given mix requires, so a buyer should not plan a complex multi-part-number programme around the fast end of that range.

Future outlook

If the flexible printed circuit market reaches USD 50.90 billion by 2030 at the projected 13.7% growth rate, most of that growth will land in applications where flexible substrates solve a physical constraint rather than an economic one: tight enclosures, moving joints, curved surfaces and dense assemblies. Each of those applications raises the share of the build that cannot be inspected optically, which points directly toward growing reliance on electrical and radiographic test records.

M2PCB’s published technical material references high-precision flexible printed circuit routing and bend-resistant substrate processes, an indication of where specialised flex manufacturers are investing as circuits move toward tighter bend radii and higher routing density. The evaluation question is likely to shift accordingly: from whether a supplier can produce a flexible circuit at all, toward whether the supplier can show a lot-linked record of how it was verified. IPC-6013 and UL 796F already provide the vocabulary for that conversation; the constraint is how many suppliers can respond in documents rather than in assurances.

FAQ

For buyers in the evaluation stage, high-mix flexible PCBA capability resolves into a documentable quantity: stated assembly parameters, a named test method, a coverage level, an acceptance standard and a lot-traceable record. Where all five can be produced, capability stops being a claim and becomes something a procurement team can audit.