IPC-6018: RF PCB Qualification & Acceptance Planning

High-frequency projects often fail to specify what a requested test should demonstrate. “RF tested” could refer to a controlled-impedance coupon, a transmission-line measurement, a functional assembly test or something else entirely. These records are related, but one does not automatically establish the others. Name the question before choosing the test vehicle.

The reference set should match the construction and end use. Do not automatically impose a space or military addendum merely because the circuit operates at microwave frequencies. If an addendum is contractually required, verify compatibility with the base document revision and identify the resulting project selections. A standard title is not a claim that RAYPCB or a particular production lot holds a certification.

Specify the exact laminate grade and the construction in which it is used. A trade name or resin family alone may leave thickness, copper option and bonding system unresolved. Hybrid boards need particular care because different regions or layers can have different material roles. The evidence plan should explain which part of the construction a material report or test vehicle represents.

Keep data-sheet values connected to their measurement conditions. A process dielectric constant and a design value used in a field solver may serve different purposes. Do not overwrite one with the other merely to obtain the same number in every document. Record the value, source, method and design use so that a later stackup review can reproduce the reasoning.

Copper and surface finish also belong in the construction description. If an RF path depends on a specified copper option or finish, changing it should trigger a review of the affected assumptions. A board can remain electrically connected while its high-frequency behavior changes. The material and finish selections therefore need the same revision discipline as a dimension on the drawing.

For an established material-family route, use the Rogers PCB guide. It supports material selection and grade-specific follow-up. This IPC-6018 page does not repeat every laminate family or turn a material supplier's typical values into acceptance guarantees for the finished board.

For frequency-dependent measurements, identify the frequency range and what is being measured between the chosen reference planes. State how launches, adapters and fixtures will be handled. If the result is extracted from more than one structure or corrected by a model, record that method as part of the evidence. Otherwise, two reasonable laboratories may produce numbers that cannot be compared directly.

Define the specimen and its relationship to the supplied board. A conveniently placed coupon is useful only to the extent that it represents the feature and construction of interest. The layer, copper, dielectric and processing history can matter to that relationship. Do not describe a test vehicle as representative without explaining what it represents and what remains outside its scope.

State the acceptance window and its source. A nominal impedance target without a tolerance does not define a receiving decision. A loss value without a length, frequency and reference-plane convention is similarly incomplete. When values are not yet fixed, ask for engineering review instead of choosing a familiar number simply to complete a quotation form.

The controlled-impedance design page provides the dedicated geometry and calculation tools. Its preliminary calculations can help frame the design question, but they are not measured evidence for the fabricated board. Retain that distinction in the release notes and customer communications.

An illustrative RF module uses a low-loss signal region together with other circuitry. The delivered board passes the agreed connectivity test, but the assembled module has a response that differs from the prototype. The electrical pass is useful evidence for connectivity; it does not identify the source of the RF difference or establish that every relevant construction parameter is unchanged.

First, compare controlled revisions: material selections, processed stackup, copper and finish, circuit geometry and any approved substitutions. Then separate board-level transmission behavior from assembly variables such as launches, components and the test fixture. This avoids replacing evidence with an assumption that either fabrication or assembly must be responsible.

If a dedicated transmission measurement is needed, agree the test structure and conditions before drawing a conclusion. A new measurement performed under a different reference-plane convention may not be directly comparable with the prototype result. Preserve the original results and the measurement configuration rather than keeping only a single favorable value in a summary sheet.

The eventual disposition should identify the requirement, the evidence and the affected scope. A solution accepted for one prototype may need a separate review before production. This workflow is an original engineering example, not a claim about a RAYPCB customer project or a substitute for the contracted specification.

Provide the fabrication revision, stackup, exact material requirements, quantity and intended assembly exposure. Identify the required qualification or acceptance records and when they must be delivered. If a special coupon, retained sample or external measurement is requested, raise it during quotation; it may affect panel planning, scheduling and the cost scope.

Do not assume that a standard reference automatically includes every optional test you might want. Equally, do not omit requirements just because they are not part of a supplier's default quotation form. Make the intended scope explicit and ask for unresolved selections to be identified. For general manufacturing verification, the quality assurance page provides the wider context.