Prototype Assembly Quotes: Setup Dominates Cost
Prototype assembly is mostly setup: stencil, program, feeder loading, first-article inspection. With five boards, setup is nearly the whole cost. That is why consigning your own components and accepting a manual or semi-manual build can be cheaper than a full turnkey prototype run.
Lead times are industry ranges, not commitments. Any factory's actual date depends on its queue that week, and on what else is already scheduled.
ChinaPCB.org does not quote boards and takes no commission. This page exists so that the enquiry you send to a factory is complete the first time. Send it yourself — see how to send an RFQ.
What this quote is priced on
Five to seven variables move the number on a prototype assembly quote. They are not equally important, and only the first two are usually discussed.
- Setup. Stencil, machine program, feeder loading and verification happen once. At five boards they are almost the entire price, and at fifty they are a small share.
- Consignment versus turnkey. Consigning parts removes sourcing and margin from the quote but puts the shortage risk on you. At prototype volume, consignment is usually cheaper.
- Manual versus line build. A hand-built board with a stencil and a reflow oven is cheaper for a few units and less repeatable. It is often the right answer for a first article.
- Rework after power-up. Prototypes get modified: cuts, jumpers, part changes. Quoting only the clean build leaves the modification work unbilled and unscheduled.
- Debug allowance. Some boards will be sacrificed. Not ordering spares means the second build waits on a second order.
The parameters a factory needs
This is the set that has to be in the enquiry before a prototype assembly price means anything. The third column is why each one changes the answer, which is also the reason a quote that skips it is not really a quote.
| Parameter | What to state | Why it changes the answer |
|---|---|---|
| Consigned or turnkey | Whether you ship parts, the factory buys them, or it is mixed by line | At prototype volume this choice moves the price more than any other. It also moves the schedule risk. |
| Board count and debug allowance | Working units plus the number you expect to sacrifice | Prototype builds consume boards. Without spares, one failure stops the build and the review date is lost. |
| Build method expectation | Manual, semi-manual with stencil and oven, or full line | The three methods have different price and different repeatability. Specifying the method is specifying the evidence you will get. |
| Date the build must inform | The design review or test event, and how hard that date is | Prototype value expires. A build that arrives after the review has cost the full price and returned nothing. |
| Expected rework | Whether cuts, jumpers or component changes are anticipated after first power-up | Rework is hand labour at engineering rates. It is routinely left out of prototype quotes and then billed. |
| Component packaging | Cut tape, partial reels, trays or loose parts, and who kitted them | Machine feeders need reels. Loose parts force manual placement, which changes both method and price. |
| Stencil requirement | Whether a stencil is included, and whether you keep it afterwards | The stencil is reusable for the next iteration. Buying it once instead of per run is a real saving on a prototype programme. |
| Documentation for hand build | Assembly drawing with reference designators and polarity, readable at a bench | A manual build is only as good as its drawing. Machine data does not help a person placing parts by hand. |
Send with the enquiry
- Gerber, centroid and BOM, plus the assembly drawing for hand work.
- Board count and the spares you want for debug.
- Whether you are consigning parts, and in what packaging.
- The event the boards must be ready for, as a date.
- Whether rework after power-up is expected, and roughly how much.
Where it goes wrong
- Production-line schedule applied to a prototype. The build arrives after the review it was meant to inform. The money is spent and the decision has already been taken without data.
- Loose parts sent for a machine build. Feeders need a carrier. The factory either hand-places at a higher rate or stops to ask, and the date moves.
- Rework not in scope. The first power-up finds a wrong value, and the fix is billed at engineering rates with no schedule slot.
- No spares ordered. One board is damaged during debug and the build cannot proceed. A second order costs more than the spares would have.
Comparing two prototype assembly quotes
Two prices are only comparable when the parameters above are held constant. Most of the spread buyers see comes from a difference in one of these, not from a difference in factory efficiency.
- Compare consignment terms. A turnkey prototype quote and a consigned one are not the same product.
- Compare the build method. Manual and line builds produce different evidence about manufacturability.
- Compare whether rework is included, and at what rate.
- Compare the date, and whether the stencil is delivered to you or retained by the factory.
What this page does not tell you
This page cannot tell you whether a manual prototype build predicts your production yield. It does not: a hand-built board passes tests that a line would fail, and the difference only appears when the design goes to volume.
Related
- Prototype assembly: what a first build can and cannot prove
- Prototype fabrication: the bare board half
- SMT assembly: the production version of the same process
- RFQ checklist: what to attach for a prototype build
Fee structure and editorial independence are described on the about page. What we can and cannot confirm about a factory is on how we verify.