If you're sourcing an SMT production line and the deadline is already tight, the single highest-ROI move is this: spend the first 4 hours nailing the layout spec and the feeder/changeover assumptions, before anyone quotes you equipment. Everything downstream — line balancing, PCB assembly integration, setup time, ramp-up yield — gets decided in that window. Skip it, and you'll pay for it in weeks, not dollars.
Sounds dramatic? Fine. But I've watched it play out often enough that I now refuse to move a project forward until that spec is signed off. Not because I'm a control freak. Because I've been the guy on the phone at 11 p.m., four days before a client's pilot run, trying to explain why the conveyor pitch doesn't match the reflow oven infeed.
Why I Say This Out Loud
I coordinate rush SMT line builds and integration projects — mostly for EMS shops and product teams who suddenly have a customer order they can't push. Over the past 6 years I've handled 140+ expedited line projects, ranging from a single-oven retrofit to a full 4-line setup for a consumer electronics client. The ones that go smoothly almost always share one trait: the spec was locked before sourcing started.
In March 2024, a client called at 9 a.m. on a Friday needing a reflow + AOI + conveyor chain configured for a 6,000-unit PCB assembly run — test units due Monday, production the following week. Normal turnaround on that scope is 5–7 business days of engineering plus shipping. We had 72 hours. We got it done because the spec had actually been drafted two weeks earlier and just needed sign-off. We paid about $2,400 extra in rush freight (on a ~$47K base order). Client's alternative was losing a launch slot they'd been holding for 8 months.
Same month, different client: no pre-spec, called us with a "I just need a line, any line" ask on a Tuesday. We made it work, but the changeover times ran 30% over what they'd projected, and they burned two days of engineering just reconciling feeder counts. Nothing catastrophic. Just expensive-in-hours.
The 12-Point Layout Check I Actually Use
I don't have hard data on how many SMT line issues trace back to layout assumptions — I've never tracked it that carefully, which I regret. But anecdotally, across my last 60 or so projects, I'd estimate 7 in 10 of the "surprise" costs (extra conveyors, re-fixturing, feeder re-buys) tie to something that was knowable before the P.O. went out.
Here's what I run through. It's not glamorous. It's just cheaper than rework.
- Board dimensions including panel rails — not just the PCB, the panel. This is the #1 source of "oh, the conveyor's too narrow" calls.
- Max and min component heights — top and bottom side. Affects nozzle selection, clearance, and whether you can run bottom-side reflow in-line.
- Feeder count and reel diameters — 7" vs 13" changes magazine layout, sometimes changes the machine class.
- Changeover frequency and target time — this dictates whether you spec a quick-swap platform or accept 40-min setups.
- Reflow profile targets and oven zone count — solder type + board mass + throughput = zone math. Don't guess.
- AOI/SPI placement and board flow direction — inline vs offline changes floor footprint more than people expect.
- Conveyor pitch and SMEMA compatibility — boring, until it isn't. Mixed-brand lines live or die here.
- Loader/unloader buffer capacity — if it's less than 30 minutes, you've built yourself a shift-start bottleneck.
- Utility drops: power, air, exhaust — a reflow oven that needs a 60A drop is a different floor plan than one that runs on 30A.
- Operator interface language and MES handshake — I've seen a line stalled 3 weeks waiting on a software localization.
- Spare parts and nozzle kit included in quote — if it's change-order territory, negotiate it now.
- Training and acceptance criteria in writing — "line running" means different things to a sales rep and a process engineer.
Look, half of these feel like common sense. That's the point. Common sense is what gets skipped when a project goes into rush mode.
The Counterintuitive Part
Here's where I disagree with a lot of people in this space. When a deadline is tight, the instinct is to move faster — call more vendors, request more quotes in parallel, compress the decision. I've done that. It usually backfires.
When I compared my Q3 2023 projects (fast-sourcing mode, 4+ vendors in parallel) against Q1 2024 (spec-first mode, 2 vendors, slower first week), the second group consistently shipped faster end-to-end. Not by a small margin — the median was about 9 days shorter from kickoff to line-acceptance. Same vendors, mostly. Different sequencing.
Seeing those two quarters side by side made me realize something I'd been dancing around for years: the sourcing timeline isn't the bottleneck. The rework timeline is. Speeding up sourcing just accelerates the mistakes.
Where This Advice Doesn't Apply
I'll be honest — this "spec first, then source" approach assumes a few things that aren't always true:
- You have someone internally who can write or validate the spec. If you don't, spend the money on a sourcing partner's pre-engineering instead. It's usually cheaper than buying the wrong oven.
- The product is reasonably stable. If the design is still changing weekly, no spec survives contact anyway — and you're better off leasing or renting a line for the pilot.
- You're not buying straight-up used equipment from a marketplace sight-unseen. That's a different game with different rules, and honestly, my approach doesn't translate well there.
I've never fully understood why some used-equipment deals in this space close in 48 hours and others drag for 6 weeks. My best guess is it comes down to whether the buyer has done this before — but I'd love to hear from anyone who's cracked that code.
One more thing, and I say this knowing it sounds preachy: the buyers who push back hardest on the "slow down and spec" advice are almost always the ones who've never been burned. The ones who nod along have all paid for it once. Five minutes of verification beats five days of correction. Every time. Not because it's clever — because it's boring, and boring is cheap.
Prices and timelines in this piece reflect personal project experience as of early 2024. Your vendor quotes will differ; verify current lead times and rates directly.