What Drives PCB Design Cost: Scope, Time and Revisions

Two quotations for the same board can differ by a factor of two, and the difference is almost never in the hourly rate. PCB design cost follows from the scope of the work, the engineering time the scope actually consumes and the number of times the design has to be revised before it is released. Those three are related, and the third is the one that customers control most directly and think about least.

Scope: What Is Inside the Engagement

The first variable is what the engagement includes. A schematic and a layout are two different pieces of work, and a layout on a defined schematic is a different piece of work from a layout that requires the circuit to be designed from a functional description. A design that requires component selection, a stack-up decision, an impedance plan and a compliance review carries more engineering than one that arrives with all of those settled.

The deliverables are part of the design scope as well. A Gerber set with a fabrication drawing is one thing; the native design files, the library, the assembly drawings, the test requirements and the documentation package are another. Both are legitimate, and a quotation that does not state which is included cannot be compared with one that does.

The firmware boundary is the third element. Where a product needs a driver, a protocol stack or a bootloader, the boundary between the hardware and the embedded firmware work has to be drawn explicitly, because a design that is electrically complete but has no way to be configured is not a finished product.

Engineering Time and What Consumes It

The time a layout consumes is dominated by the density of the placement and the number of constraints on the routing, not by the number of connections. A board with a moderate component count and a tight mechanical envelope can take longer than a larger board with an open layout, because the placement has to be solved before the routing can begin.

Critical nets are the second consumer of time. A design with differential pairs, controlled impedance, length matched groups or a sensitive analog section requires the routing to be planned rather than merely performed, and the planning is engineering work that produces no visible drawing. A layout that appears simple in the schematic can be demanding on the board for reasons that only become apparent when the stack-up is fixed.

The third consumer is the review and the correction that follows it. A design that is reviewed against the circuit, the component data and the manufacturing capability will produce findings, and resolving the findings properly takes time. The alternative is a design that goes to the fabricator with the findings unresolved, and the cost of that appears later as a revision.

The Cost of a Revision

A revision of the artwork is not only the engineering time to change it. It is a new stencil, a new panel, a new set of boards, a new assembly order and a new delivery schedule, and it delays everything downstream by the duration of a production cycle. The revision cost is therefore several times the cost of the change itself, and it is the reason why the review before release is worth more than the time it appears to consume.

The revisions that cost the most are the ones that are discovered late. A footprint error found at the design review costs an hour. The same error found when the first prototype fails costs a stencil, a panel, a batch of boards, a schedule and a second round of assembly. The arithmetic is not close, and it is the same error in each case.

What the Customer Controls

The completeness of the input is the largest single lever. A mechanical envelope, a connector list, a supply arrangement, a cost target and a compliance requirement given at the start remove the work of establishing them later, and each of them has the ability to invalidate work that has already been done if it arrives after the layout has begun.

The second lever is the decision discipline. Projects that stall do so because a decision has been left open while the work around it continues, and the work is then done twice. The cost of a decision made imperfectly today is usually lower than the cost of a decision deferred until the layout is finished.

The third is the willingness to accept the review findings before the artwork is released. The findings are the cheapest defects in the whole programme, and treating them as suggestions rather than as requirements moves each of them into the revision column at a multiple of its original cost. Our PCB design and layout service works with these factors stated up front, the capability limits that shape the layout are published under PCB capabilities, the design continues into PCB manufacturing and SMT assembly, and the firmware work is handled by the embedded firmware team.

Comparing Quotations That Do Not Match

The most common cause of a misleading comparison is a quotation that has been written against a description rather than against a design. If the customer has not supplied a mechanical envelope, a connector list or a supply arrangement, each supplier will assume something different, and the assumptions are invisible in the price. The one that assumed less work will appear cheaper and will be revised more.

The second cause is a difference in the review depth. A supplier who reviews the design against the circuit, the components and the manufacturing capability spends time that another does not, and the difference appears in the price and later in the number of revisions. The second supplier is not cheaper; the second supplier is selling less.

The third cause is the treatment of the changes that follow. Every design changes, and whether the change is included, priced as an increment or treated as a new engagement determines the total cost of the project rather than the cost of the quotation.

Making the Engagement Predictable

The way to reduce the cost of a design is to reduce the amount of work that has to be done twice. That means supplying the constraints before the layout begins, making the decisions that the layout depends on rather than deferring them, and treating the review findings as requirements rather than as suggestions.

It also means keeping the factory in the conversation. A designer who knows the process the board will be built on and the assembly it will go through can make cost decisions during the layout, and those decisions are structural rather than negotiable. The difference between a design that is expensive to build and one that is not is taken in the first two weeks of a project, not in the quotation.

FAQ

Is the price per pin a fair basis? It is a convention, not a measure of effort. Density, constraints and critical nets move the work more than the connection count does.

Why is a revision so expensive? Because it carries a new stencil, a new panel, a new assembly run and a delay, not only the engineering time to make the change.

What reduces the cost most? Complete input data and decisions made promptly, because both replace work that would otherwise be done twice.