Blog
From Perfboard to Production: What an Assembly-Cost Guide Implies for Your First Prototype
Perfboard has launched more working circuits than any CAD package ever will. A grid of plated holes, a few through-hole parts, a spool of wire and an hour with a soldering iron can prove a concept long before anyone opens a schematic editor. What that build cannot do is tell you what the same circuit will cost once it reaches a fabrication house and an assembly line. Cost guides — including the one published in November 2024 — are written for people who already have artwork, a bill of materials and a quantity in mind. Read from the bench where a hand-soldered prototype is still sitting, they become a checklist of everything the perfboard quietly left out.
Why a Perfboard Build Says Almost Nothing About Assembly Cost
A perfboard — also sold as perf board, proto board or stripboard — is a matrix of drilled holes with small copper pads, and every connection on it is made by hand. There is no copper artwork, no solder mask pattern to align, no stencil, no paste deposition and no reflow profile. That is precisely why it is so useful for validation, and precisely why it maps so poorly onto the cost structure described in a printed circuit board assembly pricing guide.
Take board production, the first cost block in any such guide. On a perfboard, that block is essentially the price of a blank substrate. On a fabricated board, it splits into materials and layers — FR-4 versus CEM-3, single-sided versus double-sided versus multi-layer — plus process choices such as gold plating versus tin spraying, and the difficulty added by fine line widths or small hole diameters. None of those decisions exist in a hand-wired prototype, because the prototype never had to be manufactured; it was assembled from a catalog part.
The same gap appears in component costing. A perfboard builder buys ones and twos from a distributor cart and substitutes whatever is in the drawer when a value is missing. A production BOM is sourced line by line, sometimes from existing inventory and sometimes from external suppliers, and the efficiency of that sourcing — plus how well inventory is managed for rare or expensive parts — becomes a visible line in the quote. Retail buying habits do not scale into that model.
Then there is processing. Surface mount technology pricing is typically built from patch points, the connection points assigned to each component, multiplied by a unit price usually in the range of $0.0015 to $0.002 per point. A perfboard has no patch points at all: every joint is a through-hole joint made by a human. Dual in-line package post-welding, which the guide notes is more labor-intensive than SMT, is the closest analogue — and on a perfboard that labor was yours, unpaid and unmeasured. Testing tells a similar story. Assembly quotes carry inspection and test, from $0.30 per board upward, using techniques such as automated optical inspection and in-circuit testing. A perfboard is tested by eye, a meter and hope.
Where Perfboard Habits Quietly Raise the Cost of the Real Board
Most of the money a first-time production run loses is decided long before the quote is requested, in habits carried over from the prototype bench. Knowing which ones translate into cost line items is the practical value of reading an assembly pricing guide early.
The first is part sprawl. Perfboard builds accumulate through-hole components chosen for leg length and ease of hand soldering rather than for availability. When those parts reach a BOM, every unusual package or long-lead item forces sourcing work, and rare components are exactly where procurement costs climb. Consolidating to parts with healthy distribution stock, and accepting a modern surface-mount equivalent where one exists, usually does more for the final number than any negotiation.
The second is layout generosity. Hand soldering demands space around each joint, so prototypes sprawl. Fabricated boards are priced against area and layer count, and generous spacing buys nothing at assembly. Redrawing the working circuit at realistic density — while keeping enough clearance for the process — removes cost that the perfboard version never revealed.
The third, and the most commonly overlooked, is test access. In-circuit testing needs defined nodes it can probe, and optical inspection needs a board that can actually be seen. A perfboard has neither test points nor a netlist, so nobody notices the omission until a test engineer asks where to put the probes. Adding test pads and programming headers at schematic stage costs almost nothing; retrofitting them costs a board spin.
The fourth is mixed technology. Not every through-hole part should be converted. Connectors, terminals and anything carrying mechanical load usually stay in holes, and the guide is explicit that DIP post-welding is more labor-intensive and therefore more expensive. The goal is a deliberate mix, not a blanket conversion. Anyone who has moved a perfboard circuit into production recognizes the pattern: the quote arrives as line items, and each line traces back to a decision that was never made consciously during prototyping.
From Working Prototype to a Quote You Can Read Line by Line
The transition starts with documentation. Freeze the circuit that works, draw it as a proper schematic, and generate a netlist and a BOM with real manufacturer part numbers and verified footprints. Assembly pricing is derived from those files, so their accuracy directly determines whether the first quote resembles the final invoice.
Next, decide the board honestly. Many perfboard circuits are functionally single-sided, but forcing single-sided routing with jumper wires can be worse than accepting a double-sided board with one signal layer, especially when the design also needs ground integrity. Where the circuit is more demanding, capability lists matter more than price: impedance-controlled constructions are typically quoted at a tolerance around ±5%, high-frequency laminates enter the picture for RF work, and layer counts can climb well past double-sided. A fabricator and assembler that handles rigid boards from single- and double-sided constructions upward, along with flex, rigid-flex and HDI, can usually quote the whole path in one place rather than handing files between vendors — worth knowing before you split fabrication and assembly across two suppliers for a design that may need either.
Then prepare the request properly. Supply Gerbers, the BOM and a clear quantity, and ask for pricing at two or three volumes, since testing and processing costs typically improve with batch size. Ask specifically how patch points, DIP post-welding and testing are counted, so the line items can be compared with the guide’s categories rather than treated as a single opaque figure. If the calculator or online quoting path is available, run the prototype quantity first and treat the result as a baseline, not a commitment.
Finally, compare the quote against the real alternative. A perfboard build costs parts, a blank board and hours of skilled hand labor that never appear on an invoice. Put those hours into the comparison and the decision usually becomes obvious — not because hand assembly was wasteful, but because it was doing a job that was never meant to scale. The prototype proved the circuit; the cost guide shows what it takes to repeat it, predictably, a few hundred times.
Raised in São Paulo’s graffiti alleys and currently stationed in Tokyo as an indie game translator, Yara writes about street art, bossa nova, anime economics, and zero-waste kitchens. She collects retro consoles and makes a mean feijoada.