SMT Machines: The Equipment Behind Modern PCB Assembly

Last updated: May 2026 · How surface-mount assembly really works, machine by machine

When people say “SMT machine” they usually picture a pick-and-place robot — but that machine is only one station in a chain. A working SMT line is a sequence of specialized machines that print solder paste, place components, melt the joints, and inspect the result. This guide walks through what “SMT machine” really means, the line station by station, the types of pick-and-place equipment, the speed and accuracy numbers, how the reflow oven works, where SMT meets through-hole, and exactly which files a line needs from your design.

What “SMT machine” actually means

SMT stands for surface-mount technology — components that sit on pads on the board surface rather than passing through holes. “SMT machine” most often refers to the pick-and-place machine, a robotic system that picks tiny surface-mount devices (SMDs) from reels, trays, or sticks and places them onto the board with micron-level accuracy.

Why it’s really a team of machines

The pick-and-place is the most visible and most photographed station, which is why it gets the “SMT machine” label. But SMT assembly is a coordinated team of machines, not one. Solder paste must be printed before placement, the joints must be melted after it, and inspection happens between stages. Understanding the full line — not just the robot arm — is what clarifies how boards are actually built and where defects come from.

The SMT line, station by station

A complete SMT line runs in sequence, each machine handing the board to the next. The three core machines are the stencil printer, the pick-and-place, and the reflow oven, with inspection stations between them.

Why the order matters

Each stage depends on the one before it. Paste must be printed accurately or placement lands components on the wrong amount of solder; SPI catches paste defects before parts are placed, when they’re still cheap to fix; placement must be correct before reflow locks everything in permanently. Inspection between stages catches problems while they can still be corrected rather than after the joints are frozen.

Types of pick-and-place machines

Not all placement machines are alike, and a well-designed line often combines two types for efficiency.

High-speed chip shooters

Chip shooters are optimized for rapid placement of small, simple passives — resistors, capacitors, and the like. These are the speed kings of the line, placing tens of thousands of parts per hour. A single modern high-speed head can exceed 250,000 components per hour (CPH). They handle the bulk of a board’s component count, since most boards have far more passives than complex parts.

Flexible / fine-pitch placement machines

Flexible placers are slower but more precise, built for larger and fine-pitch parts — ICs, connectors, BGAs, and odd-form components. They use advanced vision systems to align delicate parts accurately, trading raw speed for the precision those parts demand. A chip shooter would damage or misplace these; the flexible placer handles them carefully.

Pairing the two

A well-designed line pairs a chip shooter for the bulk of passives with a flexible placer for the complex parts, so each machine does what it’s best at. Optimized multi-machine lines can exceed 400,000 CPH total throughput by splitting the work this way.

How a placement head actually works

Inside any pick-and-place machine, a few mechanisms repeat thousands of times a minute. Component feeders — reels, trays, or tube sticks — present parts at fixed pick positions; a nozzle on the head uses vacuum to lift each part, a vision system images it to check it picked correctly and to measure its exact rotation and offset, and the head then places it on the pasted pad with a tiny corrective adjustment so it lands precisely on target. Fiducial marks on the board let the machine calibrate the board’s true position before it places anything, compensating for slight variations in how each board sits. Heads carry interchangeable nozzles sized for different parts and swap them automatically, which is how one machine can place a tiny passive and a large connector in the same program. Understanding this is useful for designers: clear fiducials, adequate part spacing, and standard component packaging are exactly what let the machine run fast and accurately.

Speed, accuracy, and component range

Modern placement machines are remarkable in both the range of parts they handle and the precision they achieve.

The size range

Today’s SMD machines handle a wide size range — from tiny 0201 (and even 01005) passives, smaller than a grain of sand, up to large BGAs and connectors. They switch nozzle types dynamically to grip different parts, and align each one with vision systems for placement accuracy measured in microns. This range is why a single line can build a board mixing tiny passives with large processors.

Matching the machine to the job

High-mix machines optimize for frequent product changeovers, suiting shops that run many different boards in small batches. Dual-lane machines double throughput in the same floor space by running two boards side by side. The right machine depends on your volume, your component mix, and the smallest part you need to place reliably — there is no single “best” machine, only the best fit for a given production profile.

The reflow oven and thermal profile

The reflow oven is the unsung hero of the line — the station that actually creates the solder joints.

How reflow works

The pasted, populated board travels through a multi-zone oven following a precise thermal profile with four phases: preheat (a gentle ramp up), soak (flux activates and temperatures across the board equalize), reflow (the peak, above the solder’s melting point — about 217–220 °C for lead-free SAC305 — where the paste melts and forms joints), and cooling (the joints solidify). The board never stops; it moves continuously through the zones, each held at a controlled temperature.

Why the profile is critical

Get the thermal profile right and every joint forms cleanly and simultaneously. Get it wrong and you see tombstoning (small parts standing on end), cold joints, or thermal damage to components. The profile must suit the specific board, paste, and component mix — which is why setting it is a skilled task, not a fixed setting. This controlled, all-at-once soldering is exactly what makes SMT so consistent compared with hand work, where each joint is heated individually.

The stencil and paste deposit you don’t see

Long before the oven, the quality of every joint is largely decided at the stencil printer. The stencil is a thin metal sheet, laser-cut from your paste layer, with an aperture over each pad; the printer squeegees solder paste through those apertures so a precise volume lands on each pad. Too much paste bridges fine-pitch pins, too little starves a joint, and misalignment smears paste across the board — which is exactly why the SPI station immediately measures paste volume, height, and position in 3D before any component is placed. The lesson for designers is that the paste layer in your files is not an afterthought: aperture sizing (often slightly reduced from the pad for fine-pitch or large thermal pads) directly controls how well the line can build your board.

SMT vs. through-hole and mixed assembly

SMT dominates modern electronics, but it has not entirely replaced through-hole, and many boards use both.

Why SMT dominates

SMT is faster, allows much smaller components, and is built for automation — the whole line above runs with minimal human intervention. The vast majority of components on a modern board are surface-mount for these reasons.

Where through-hole survives

Through-hole (THT) persists for parts that need mechanical strength — large connectors, heavy capacitors, anything that takes physical stress — or where hand assembly is genuinely preferred. The leads anchored through the board give a mechanical robustness that surface pads can’t match for high-stress parts.

Mixed-technology boards

Many real boards are mixed-technology: the SMT parts are placed and reflowed first, then the through-hole parts are added by wave soldering or by hand. A full PCBA line may extend beyond the core SMT chain to add DIP insertion, selective wave soldering, conformal coating, and final assembly — the SMT line is the heart, but not the whole, of complex board production.

What an SMT line needs from your design

An SMT line can only build what your files describe. Missing or mismatched data is a leading cause of assembly delays.

• Gerber/ODB++ including the paste layer — the paste layer is used to cut the stencil that prints solder paste, so it must be present and correct.
• BOM with exact part numbers and values, so the right components are sourced and loaded.
• Centroid / pick-and-place file — the X-Y position, rotation, and side for each part, which tells the machine where everything goes.
• Assembly drawing with polarity, pin-one, fiducials, and DNP/DNI notes, to resolve any ambiguity.
• Fiducial marks on the board so the machines can align accurately to the actual board position.

The most common file problem

Mismatched or missing centroid data is a leading cause of assembly delays — if the rotations or positions are wrong, parts go down crooked or backward. Verify your centroid against the board before sending, and make sure every file comes from the same design revision so they agree with one another.

Letting a full SMT line build your board

Highleap Electronics runs complete SMT lines — paste printing, SPI, automated placement, multi-zone reflow, and AOI/X-ray inspection — plus through-hole and mixed-technology assembly, from prototypes to volume production. We review your files first with a free DFM check that catches paste-layer, centroid, and footprint problems before a board ever reaches the line. See our fabrication services too.

Get an SMT assembly quote →

Frequently asked questions

Is a pick-and-place machine the same as an SMT machine?

It’s the most prominent one, but a full SMT line also includes a solder-paste printer, a reflow oven, and inspection equipment (SPI, AOI, X-ray). “SMT machine” usually means the pick-and-place specifically.

How many components can an SMT machine place per hour?

A single modern high-speed chip shooter can exceed 250,000 CPH; an optimized multi-machine line can exceed 400,000 CPH total throughput.

What’s the difference between a chip shooter and a flexible placer?

Chip shooters place small passives very fast; flexible placers handle larger and fine-pitch parts (ICs, BGAs) more precisely but slower. Lines often use both together.

Do I need a stencil for SMT?

For machine paste printing, yes — the stencil is cut from your paste layer, which is why that layer must be included in your files.

Can SMT and through-hole be combined on one board?

Yes. Mixed-technology boards reflow the SMT parts first, then add through-hole parts by wave soldering or by hand.

What is the smallest component an SMT machine can place?

Modern machines handle down to 0201 and even 01005 passives, plus fine-pitch ICs and BGAs, using vision-guided alignment.

Why does the thermal profile matter so much?

It controls how the paste melts and the joints form. A wrong profile causes tombstoning, cold joints, or thermal damage; a correct one forms every joint cleanly in one pass.