-
The questions, in order
-
1. What does an SMT production line actually cost?
-
2. What's the minimum equipment to run a line?
-
3. Should we design the layout in-house or let a vendor do it?
-
4. How long does this take, realistically?
-
5. New equipment or used?
-
6. How do I spot a sourcing partner who actually knows what they're doing?
-
7. What if we've never done PCB assembly before?
I'm the office administrator at a 55-person contract manufacturer. Last year the company added an SMT line, and the whole thing—quotes, contracts, vendor coordination, and a surprising amount of time chasing customs paperwork—landed on my desk. I'm not an engineer. I don't know trace widths or nozzle specs, and I have no business pretending to. What I do know is the budget side, the timeline side, and what it looks like when a sourcing conversation is going sideways.
Here are the questions I actually had when we were sourcing our line, answered the way I wish someone had answered them for me.
The questions, in order
- What does an SMT production line actually cost?
- What's the minimum equipment to run a line?
- Should we design the layout in-house or let a vendor do it?
- How long does this take, realistically?
- New equipment or used?
- How do I spot a sourcing partner who actually knows what they're doing?
- What if we've never done PCB assembly before?
1. What does an SMT production line actually cost?
The honest answer is "a lot more than the equipment quotes suggest." Here's the range I collected from vendor conversations in late 2024 and early 2025, and these are rough numbers—verify them against current quotes, because the market moves:
- Entry-level line (solder paste printer, one placement machine, reflow oven, basic AOI): roughly $150,000–$250,000 for the equipment.
- Mid-range (faster placement, better inspection, some inline automation): $300,000–$600,000.
- Higher-end with dual-lane placement and full inline AOI: $800,000 and up, new.
What nobody tells you upfront: those numbers usually exclude facility work. Power upgrades, compressed air, ESD flooring, and floor loading requirements can add 15–25% to the effective project cost. We budgeted for the equipment and then had to go back to finance for the infrastructure piece, which was not a fun conversation.
I'm an administrator, not a cost engineer—so treat those figures as a starting point for your own quote round, not as a benchmark.
2. What's the minimum equipment to run a line?
In process order, you need four things:
- A solder paste printer
- At least one placement machine
- A reflow oven
- Some form of inspection—AOI, or manual inspection plus X-ray if you're building anything with hidden joints
Loaders and unloaders are optional; you'll want them for volume, not for a first line. Cleaning and rework stations are separate costs.
Here's the sourcing advice I'd give anyone: build the equipment list from your board mix, not from what a vendor is motivated to sell you. When I first started this process, I assumed the right move was to take a turnkey package and let the supplier define the scope. That was wrong. We ended up re-specifying twice because the original package was tuned for simple single-sided boards, and our actual mix included double-sided assemblies with fine-pitch components.
3. Should we design the layout in-house or let a vendor do it?
Vendors will offer a layout. Accept it as a starting point, not a final answer. Their layout is typically optimized for their equipment placement, not for your material flow, operator movement, or utility access.
Ours looked fine on paper. In practice, the stencil storage cabinet ended up in the main aisle, which meant an operator had to step around it every time. Small thing. Adds up over a year.
The pieces worth deciding yourself, or with an independent integrator: product mix and expected volume, material flow from incoming inspection to finished goods, operator walk paths, power and air drop locations, and ESD zoning. Get those nailed before equipment lands, because moving a reflow oven after the fact is not a small project.
4. How long does this take, realistically?
From signed order to first good board, ours took about 14 weeks. Breakdown: roughly 6 weeks of lead time, 2 weeks of installation, and 6 weeks of commissioning, calibration, and training.
That was on the fast end, mostly because we'd already finished facility prep before the equipment shipped. Lead times reported by vendors in 2024 ranged from 4 to 10 weeks depending on equipment type and configuration. Two things that reliably pushed timelines out: ordering near Chinese New Year, and configurations that needed factory-level customization.
Add at least a month of buffer to whatever the vendor quotes. Not because they're lying—they're quoting their part of the work, not the facility delays, permit waits, and finance reviews that happen on your side.
5. New equipment or used?
I went into this assuming used equipment was a bad idea. Three used-equipment dealers later, my view hadn't changed much—until we put a refurbishment quote next to a new-equipment quote for a comparable specification.
Seeing them side by side, the pattern was clear: used was roughly 40% cheaper, but it came without software updates, without a spare-parts commitment, and with service through a third party rather than the manufacturer. For a company with an experienced maintenance team and a stable product mix, that trade may be worth it. For us—no internal SMT expertise at the time—new was buying operational support as much as it was buying hardware.
I'd say used is worth serious consideration if you're piloting a product or already have process engineers on staff. Otherwise, the total cost of ownership gap closes faster than the sticker price suggests.
6. How do I spot a sourcing partner who actually knows what they're doing?
My initial approach to SMT production line sourcing was completely wrong. I thought a sourcing partner's main job was negotiating the price down. It's not—the price negotiation is the easy part. The value is in catching configuration mistakes before they turn into purchase orders, and in knowing which combinations of equipment actually run together without integration headaches.
What I now ask every sourcing partner for, without exception:
- A written configuration list with specific equipment models, not categories
- The service and support terms in writing—response time, spare parts availability, training included or not
- Two reference customers I can actually call, roughly my size
- A clear statement of what's excluded from the scope
Any partner who can't put equipment models on paper is quoting a concept, not a line. Ask what standards they hold the output to—most competent integrators will reference IPC-A-610 for acceptability criteria, and that's a reasonable starting point for a shared vocabulary between you, the vendor, and your future customer.
7. What if we've never done PCB assembly before?
You can still do this. Just budget for the parts nobody quotes.
Training is one: expect one to three weeks per shift for operators, more if your product mix is complex. Process engineering time is another: you'll either hire for it or contract it, and either way it costs money. Then there's the ramp-up curve—plan for three to six months before output stabilizes at target yield.
There's also a lower-risk path worth considering: outsource PCB assembly first, buy the line when volume justifies it. That's the route we took. We outsourced for about 18 months, learned what our real requirements were from working with an assembly house, and then wrote a much better equipment spec because of it.
One thing I'd flag from experience: don't let the handover happen without a formal acceptance test. We didn't have a documented acceptance process on our first line—cost us when a calibration issue surfaced two weeks after sign-off and the vendor argued it was operator error. The second time we bought equipment, I wrote the acceptance criteria into the purchase order before the deposit went out. Should have done that from the beginning.
That's the summary from the procurement side of the table. The technical decisions still need technical people—but if you're the one holding the budget and the contract, at least these seven questions will keep the conversation on track.