Auditing an HDI Supplier: What the Checklist Should Contain

High-density interconnect boards are used wherever space is scarce and signal density is high: smartphones, wearable devices, AI server boards, automotive controllers and increasingly portable medical equipment. The common characteristic is that line width, spacing and via size are all smaller than on a conventional multilayer board, and the process window is correspondingly narrower. An HDI supplier is judged on process control and batch data rather than on a capability list.

That narrow window is what makes supplier selection different. For a standard multilayer board, most qualified factories can deliver equivalent results, so price and lead time dominate the decision. For an HDI board, two factories with similar equipment lists can produce materially different yields, because the difference lies in process control rather than in machinery.

An audit therefore has a specific purpose, which is to determine how a factory keeps a narrow process inside its limits over time. It is not an equipment inventory, and treating it as one produces a decision that looks rigorous and predicts very little.

Dimension One: Technical Capability in Real Numbers

The first dimension is capability, expressed as numbers rather than adjectives. The relevant figures are the minimum microvia diameter and the aspect ratio the factory achieves reliably, the minimum line width and spacing at production yield, the registration accuracy it holds across a build, and the layer count it produces in routine work rather than in demonstration.

Each of those needs a qualifier. A minimum via diameter achieved once on a coupon is not the same as one held across a production month. The useful question is what the factory’s own process control limits are, and whether its records show the process staying inside them.

The discussion should also cover the materials the factory has genuinely qualified. Handling M8 and M9 grade laminates or specialised build-up films requires process recipes that are developed by running production, and a factory that has never processed a material should not claim the capability on the basis of a data sheet.

Dimension Two: Process Control and Measurement

The second dimension is how the factory measures itself. A capable HDI line tracks registration, plating thickness, via fill quality and impedance as control parameters, with defined limits, defined measurement frequency and a documented reaction when a value approaches a limit.

Evidence matters more than description here. Trend charts for registration over several months, cross-section results from routine production rather than a qualification campaign, and impedance measurements from populated panels all demonstrate that control is real. A factory that cannot produce such records is relying on final inspection to catch problems, which is a fundamentally weaker position.

The reaction plan is the part most often missing. It is not enough to measure; the factory must be able to describe what happens when a measurement trends toward a limit, who decides, and what has to be re-verified before production continues. A quality system that answers that question clearly is worth more than one with an impressive certificate portfolio.

Certification is a necessary but insufficient input. An ISO or automotive certificate confirms that a quality system exists and is audited, not that a specific microvia process is under control. Auditors should accept certification as a floor and still examine the process data for the technology being sourced, because the two answer different questions.

Finally, the audit should examine how the factory handles change. A laminate substitution, a plating chemistry adjustment or a change in lamination pressure all affect an HDI process, and a supplier that notifies customers before implementing such changes is protecting them from a surprise. Change notification is one of the clearest indicators of a mature quality culture.

Dimension Three: Product Coverage and Experience

The third dimension is whether the factory has made the kind of board being sourced before. HDI covers a wide range, from a two-layer build-up on a simple core to multi-order constructions with stacked vias, and capability in one does not imply capability in another.

Experience shows up in the questions a factory asks. A supplier that has built similar boards will ask about via stacking, impedance targets, thermal requirements and expected volume early, because those factors change how the board must be produced. One that has not will accept the drawing and discover the constraints later.

It is reasonable to ask for references on comparable work, including the layer structure and material class, and to ask what problems arose and how they were solved. Factories willing to discuss their own failures usually have a better grasp of their process than those that present only successes.

Dimension Four: Delivery and Capacity Reality

Lead time claims are easy to make and difficult to verify, so the audit should establish where capacity actually exists. The relevant questions are how many laser drilling machines are available and how they are loaded, what proportion of capacity is committed to existing customers, and how the factory plans to absorb a new programme.

Material availability belongs in this dimension. A factory with free capacity but no allocation of a critical laminate or build-up film cannot start work, and the audit should establish how many qualified sources the factory holds for each key material and how it manages allocation.

Deliverable performance should be measured rather than described. On-time delivery rate, the frequency of schedule recovery, and how the factory communicates a slip are all informative, and a supplier that reports a slip early is more valuable than one that reports it at the deadline.

There is also a commercial dimension worth probing. Factories overloaded with work in a strong market may accept orders they cannot schedule comfortably, and the resulting pressure shows up as yield problems and missed dates. Asking how the factory decides which orders to accept, and how it protects committed lead times when demand rises, reveals how it manages that risk.

Dimension Five: Engineering Service

The fifth dimension is what the factory contributes beyond manufacturing. HDI designs frequently need adjustment to be producible, whether in via structure, panelisation, impedance stackup or test access, and a supplier that can propose alternatives is far more useful than one that simply quotes the drawing.

Useful contributions include design rule reviews before layout freeze, stackup proposals that trade cost against capability, and manufacturability feedback on panel utilisation. These are the interactions that shorten development cycles, and they are a better predictor of a productive relationship than a low quotation.

Assessing this dimension means talking to the engineers, not only the sales team. A short technical discussion, covering the planned board layout, reveals quickly whether the factory understands the design intent or is reading a specification. That conversation is worth more than several pages of audit checklist, and it is what separates a supplier from a partner in board manufacturing.

Building a Workable Audit Scorecard

A practical scorecard weights the five dimensions according to the risk of the specific programme. A board for a consumer device may weight delivery and cost highest, while one intended for automotive or infrastructure use shifts weight toward process control and traceability.

Each dimension should be scored on evidence rather than impression, with the evidence recorded so that the decision can be explained later. That record is also useful as a baseline: when the programme runs, the same measures can be revisited to see whether the supplier is performing as assessed.

The final element is follow-up, and it should be proportionate to the risk of the programme. A first audit of a new supplier for an advanced board warrants a full action plan with dates, whereas a routine re-audit of an established partner may need only a short list of observations to track. Matching the follow-up effort to the exposure keeps the process credible rather than bureaucratic, and it makes suppliers more willing to raise issues themselves. Findings from an audit should have owners and dates, and the supplier’s response to those findings is itself diagnostic. A factory that addresses an observation with a documented change is a factory that will improve; one that explains why the observation does not apply is unlikely to.

Training and staff retention matter more in HDI production than in conventional board work, because the process depends on operator judgement at several stages. A factory that can describe how it trains operators on laser drilling, plating and inspection, and how it retains them, is describing something that affects yield directly and does not appear in any equipment list.

The Decision That Follows

Supplier selection for HDI boards is a risk decision rather than a price decision. The quotation indicates what the factory believes it can do; the audit indicates what it can demonstrate. When those two diverge, the audit is the more reliable input.

That does not mean choosing the most expensive or most elaborate supplier. It means matching the capability required by the design to a factory that can demonstrate it, and paying for the level of control the product actually needs. A design that does not require the tightest impedance tolerance should not be placed with a supplier chosen for that tolerance alone.

Used properly, the audit produces a decision a team can defend, and a working relationship with a manufacturer whose production process is understood rather than assumed. That understanding is what makes the difference when a programme hits a technical problem in the middle of a build.