Ask ten people in electronics what “PCB manufacturing” means and you will likely get several different answers. Some picture bare laminate panels moving through drilling and plating lines. Others picture pick-and-place machines dropping resistors and chips onto a finished board. Both are correct—but they describe two separate stages of the same project. PCB fabrication builds the empty interconnect structure. PCB assembly populates that structure with components. Treating these as one continuous, interchangeable step is one of the most common sources of quoting confusion, design revisions, and missed schedules. The clearer the boundary between the two, the smoother the handoff from a bare board to a working assembly.
The Bare Board Boundary: What PCB Fabrication Actually Delivers
Fabrication begins with a laminate—rigid, flexible, or a rigid-flex combination—and ends with a finished bare board that has no components attached. Everything in between is about creating copper paths, insulating layers, holes, and a protective surface finish that will later accept solder. A fabrication shop drills vias, patterns the copper, laminates the stackup, applies solder mask and legend, and finishes exposed pads. When the board leaves this stage, it should pass bare-board electrical testing and dimensional checks, but it cannot yet perform any circuit function.
That boundary is easier to see when you look at the capabilities that belong strictly to fabrication. Layer counts from 2 to 58 layers, laser-drilled microvias, blind and buried vias, via-in-pad with resin or copper fill, cavity boards, and heavy copper up to 30 oz. are all fabrication decisions. So are back drilling, impedance control held to ±5%, and the choice of surface finish—ENIG, ENEPIG, hard gold, immersion silver, immersion tin, OSP, or wire-bondable gold. None of these steps place a component. They define how well the board will support the components that arrive later.
Material selection also sits on the fabrication side. High-frequency and high-speed designs may call for Isola, Rogers, Arlon, Taconic, or Panasonic MEGTRON laminates; flexible designs may use polyimide, PET, PEN, or LCP substrates. These choices affect dielectric constant, loss tangent, and thermal behavior, and they are locked in before assembly begins. When you request a quote from a pcb fab, you are pricing this bare-board work: the stackup, the drilling, the plating, the finish, and the test. Components, stencils, and reflow profiles are not part of that quote.
Why the Split Matters for Design, Materials, and Tolerances
Design teams feel the separation most sharply at the documentation stage. Fabrication needs Gerbers, drill files, stackup notes, impedance requirements, and surface-finish callouts. Assembly needs a bill of materials, a pick-and-place file, polarity references, and any special instructions for fine-pitch or bottom-side parts. Mixing the two into a single ambiguous package is how boards get built to the wrong via fill, or assemblies get quoted against a bare-board drawing. Keeping the packages distinct is not bureaucracy; it is how each supplier knows exactly what it is responsible for.
The technical overlap is real, though, and that is why the handoff deserves attention. A fabrication choice such as via-in-pad fill can directly affect assembly yield: an unfilled via in a thermal pad can wick solder away from the joint, while an improperly capped via can create a uneven surface for a fine-pitch component. Likewise, a surface finish selected for wire bonding may behave differently under a standard reflow process than a finish selected for solderability. Impedance tolerances held during fabrication determine whether high-speed signals arrive cleanly, but the assembly stage determines whether the components that drive those signals are placed within the right footprint and orientation.
This is also where material families cross the boundary. A low-loss laminate chosen for an RF or microwave board is a fabrication input, but the assembly house must know its thermal limits so the reflow profile does not degrade the dielectric. Rigid-flex and flex constructions add another layer: the fabrication shop controls feature sizes down to 25 µm and flexible dielectric cores down to 25 µm, while the assembly team must handle the same boards without stressing bend areas or delaminating covers. The split is not a wall. It is a controlled interface, and the documentation should make that interface explicit.
Where the Handoff Gets Tricky: HDI, Rigid-Flex, and Cavity Boards
Advanced constructions expose the fabrication/assembly boundary most clearly because the tolerances on one side constrain what the other side can do. HDI boards with 0.15 mm minimum BGA pitch and 0.002″ minimum trace width and spacing demand fabrication precision that leaves very little room for assembly error. Stacked, staggered, and skipped microvias, including any-layer HDI constructions, change how heat moves through the board during reflow. A cavity board is another example: fabrication creates the cavity, but assembly may need to place a die or component inside it, which means the two processes must be planned together even though they remain separate operations.
Rigid-flex and flex assemblies sharpen the point. The fabrication side produces the flexible core, the plated vias, and the surface finish; the assembly side handles stiffeners, connectors, and components on areas that may later be folded. A flex board that passes bare-board inspection can still fail after assembly if the handling, support, or reflow profile is wrong. That is why responsible one-stop shops keep the two workflows visible as separate line items and separate quality gates, even when both happen under the same roof. Bare-board testing, X-ray inspection of BGAs and QFNs, IC programming, and functional testing all sit at different points along that chain, and each one answers a different question.
For buyers, the practical takeaway is to ask which stage a quote covers. A fabrication quote should describe the bare board—stackup, layer count, finish, via structure, and test. An assembly quote should describe the components, the placement technology (SMT, through-hole, or mixed), the board size limits, and the inspection steps. When both are quoted together, the clarity of that split is what protects the schedule. A board that is fabricated to the wrong via fill or finish may pass the fab house’s own checks and still create costly rework at assembly. Keeping fabrication and assembly as distinct, well-documented stages is not a formality; it is the difference between a board that fits its components and one that only looks finished.
