Who This Checklist Is For

If you're sourcing or setting up an SMT production line — whether from scratch, moving an existing line, or adding a second one — this is for you. I'm writing mainly for two groups: operations and procurement leads in electronics manufacturing who own the line delivery, and smaller factories without a dedicated SMT engineer who still have to get a line running.

I've run point on around 37 urgent projects, several of which turned into "we need this running in two weeks" calls with vendors. Here's the process I use. Nothing here is complicated, but skip a step and you'll feel it later.

Seven steps, one checkpoint each. That's it.

The 7-Step SMT Production Line Sourcing & Setup Checklist

Step 1: Lock Down Product Mix and Volume Range

Do not open a single equipment spec sheet yet. Spend half a day writing down the product range for the next 6–12 months: min and max board size, component density, typical joint count, peak daily throughput, peak weekly throughput.

This decision drives every other one. A 400mm × 500mm LED driver board and an 80mm × 60mm controller don't just need different pick-and-place accuracy — they need different conveyor widths, different reflow profiles, and in some cases different solder paste handling entirely.

Most people skip this, or wave it away with "we can do a bit of everything." An SMT production line cannot do everything well. Pick the profile that covers 80% of your real demand, and plan for the rest as exceptions.

Checkpoint: One page, three configurations (min / typical / max). Every vendor you talk to gets this page before you ask for a quote.

Step 2: Design the Layout Around Real Constraints, Not a Catalog

An SMT production line layout is not a drawing exercise. It's a constraint-satisfaction problem. Floor area, load-bearing capacity, ceiling height, power capacity, compressed-air pressure, fume-extraction routes — these hard limits decide what's actually possible.

We had one project where every piece of equipment was selected before anyone noticed the ceiling was 3.8m and the wave solder machine plus its extraction hood needed 4.2m. We redrew the whole line. Cost us 11 days, and it was our own fault, not the vendor's.

Measure the site first. Then talk equipment. Leave at least one station's worth of expansion room. Keep the material flow direction consistent — no crossing paths, no backtracking.

Checkpoint: A dimensioned floor plan marking power drops, compressed air, fume extraction, structural columns, and personnel walkways. Ask vendors to sketch on your plan — not a blank page.

Step 3: Source Vendors by Capability, Not by Price List

When I'm sourcing an SMT production line, I rank vendors on four things in this order: service response time, install experience with our board profile, parts availability for the next 5 years, then price. Price is last, and I mean that.

It's tempting to compare machines head-to-head on spec sheets. But two machines with identical placement rates can behave completely differently once you're running your actual PCB assembly on an SMT production line at 90% utilization.

Ask each vendor for two reference customers running similar boards. Call them. Ask what happens when a feeder jams at 2 AM. That answer tells you more than any brochure.

Checkpoint: A one-page comparison of your top three vendors — response time, reference feedback, spare-part commitment — signed off before you negotiate anything.

Step 4: Confirm Utilities and Facility Readiness Before Signing

This is the step people miss. Power capacity, three-phase availability, air quality and pressure at the nozzle, humidity control for paste storage, floor flatness for printing accuracy — all of it needs to be verified, not assumed.

I wish I had tracked how many projects stalled on utility gaps rather than equipment delays. What I can say anecdotally is that it's more than half. The machine is rarely the bottleneck in week one. The building is.

Checkpoint: A signed facilities checklist from your own engineering team, dated within 30 days of the equipment order.

Step 5: Plan Installation and Commissioning Weeks Together

SMT production line setup has three phases that overlap if you're not careful: rigging and placement, power-and-air connection, and process commissioning. Each has a different crew, and they don't all need to be on-site at once.

Rigging takes 1–3 days for a typical line. Power and air, another 2–4. Process commissioning — first article, stencil tuning, reflow profiling — 3–7 days depending on board complexity. If you sequence these back-to-back with no buffer, one delay cascades.

Build in a two-day buffer between each phase. Seriously. It costs nothing on paper and saves everything in reality.

Checkpoint: A shared Gantt chart, with the vendor's name on the commissioning dates, not just yours.

Step 6: Run a First-Article Trial Before Full Handover

Do not accept handover without a first-article inspection on your real product. Run 20–50 boards through the full PCB assembly SMT production line — printing, placement, reflow, AOI, and any post-processing. Inspect to IPC-A-610 Class 2 standards, or Class 3 if your product demands it.

I said "run the trial," and one vendor heard "run a sample." We ended up discovering, two weeks after handover, that the paste printer was inconsistent on the outer edges of a 350mm board. Not a spec problem. A calibration problem that a proper trial would have caught.

Checkpoint: A written first-article report signed by both sides. No signature, no final payment.

Step 7: Plan the 30-Day Ramp, Not Just the Go-Live

Go-live is day one. Ramp is days 2 through 30. Most SMT production line sourcing plans stop at handover and then wonder why yield drops in week three.

Set weekly yield targets. Track defect Pareto each Friday. Keep the vendor's commissioning engineer on call for the first two weeks. Schedule a formal review at day 14 and day 30.

Checkpoint: A one-page ramp schedule with named owners per week. Nobody owns "quality" — somebody owns "week 2 print defects." Be specific.

Common Mistakes to Avoid

Three things that show up again and again:

  • Buying capacity you don't need yet. A line running at 40% utilization with a machine rated for 3× your peak is a cash-flow problem, not a growth plan.
  • Treating layout as a vendor responsibility. Vendors design around their equipment. Somebody on your side has to design around your building, your people, and your product.
  • Skipping the buffer days. Every schedule I've seen that "saved" two days by cutting buffer ended up losing five somewhere else.

Bottom line: efficient line sourcing is not about finding the cheapest quote. It's about compressing the time between "we need a line" and "the line makes good boards." Every step above is aimed at that compression. Where you can standardize, standardize. Where you can't, build buffer. That's the whole discipline.