Pick and Place Machine: Types, Specs & Selection Guide

A pick and place machine is the robot at the heart of an SMT line: it picks surface-mount components from feeders and places them onto a PCB at anywhere from a few thousand to over 100,000 parts per hour, with placement accuracy down to ±25 µm or tighter. The two machine families do different jobs — a chip shooter blasts down small passives at high speed, while a flexible placer handles fine-pitch ICs, BGAs, and odd-form parts with precision — and most real lines run both. This guide breaks down the types, the specs that actually matter (and the ones that mislead), real CPH and accuracy numbers from current machines, and a selection framework so you buy for your board mix, not a datasheet headline.

Key Takeaways

• A pick and place machine picks SMT components from feeders and places them on a PCB; the two families are high-speed chip shooters and high-accuracy flexible placers.
• Chip shooters run 20,000–100,000+ CPH at roughly 70 µm accuracy for passives; flexible placers run 6,000–40,000 CPH down to ±25 µm for ICs, BGAs, and odd-form parts.
• Headline CPH is measured under optimum conditions — real-world throughput is roughly 70% of the rated figure once feeding, vision, and changeovers are counted.
• IPC-9850 is the standard that makes machine specs comparable; it ties accuracy to speed and uses Cpk (target ≥ 1.33) to score placement capability.
• Match the machine to your board mix and volume, not a spec sheet: component range, feeder count, accuracy, and changeover time matter more than peak CPH.

What Is a Pick and Place Machine?

A pick and place machine is the automated placement system that mounts surface-mount devices onto a board. It’s the core of SMT assembly and typically the single most expensive machine on the line — often around half the line’s capital cost — because everything downstream depends on parts landing in the right spot at the right angle. It sits between the solder paste printer and the reflow oven: paste goes down first, the machine places components into the wet paste, and reflow fixes them in place.

Underneath the covers, every machine is the same five subsystems working together. Board handling (a conveyor and rails) brings the PCB in and holds it flat. Feeders present the components — on tape reels, in tubes, or on trays. A placement head carries one or more vacuum nozzles that pick and place. A vision system reads board fiducials and inspects each picked part. And servo-driven axes move the head, the board, or both. The differences between a $15,000 desktop unit and a seven-figure production line come down to how fast, how accurately, and how repeatably those five subsystems run.

How a Pick and Place Machine Works

The placement cycle is the same whether the machine costs five figures or seven. Here’s what happens to every component.

1. Load the board and read fiducials. The conveyor positions the PCB and a camera reads the fiducial marks to calculate the board’s exact position and any skew, so placement coordinates map to the real board, not the nominal one.
2. Feed the components. Tape, tube, and tray feeders present parts in order. Electric servo feeders index reels precisely and report part data back to the machine.
3. Pick with the nozzle. The head moves to the feeder and a vacuum nozzle lifts the part. Nozzle size is matched to the component — a 0.4 mm nozzle suits an 0402, while a connector needs a wide nozzle — because a mismatch causes pickup failures and tombstones.
4. Center and inspect on the fly. A vision system images the picked part to measure its offset and rotation, verify it’s the correct part and polarity, and correct the placement coordinates before the part ever touches the board.
5. Place with Z control. The head moves to the XY target, rotates the nozzle to the right angle, and sets the part down with controlled Z force. On a chip shooter the board table moves under the head; on a gantry placer the head moves over the board.
6. Verify and continue. The machine repeats until the board is populated, then passes it to reflow. Many lines add post-placement AOI to catch missing, shifted, or wrong parts before the oven.

Chip Shooter vs Flexible Placer: Pick and Place Machine Types

Placement machines split into two families built around opposite priorities, plus a modular class that blends them. A chip shooter is built for raw speed on simple parts — resistors, capacitors, small actives — using a stationary turret, revolver, or gantry head that can place 20,000 to over 100,000 components per hour. The trade-off is accuracy (around 70 µm) and a component-size ceiling near 20 mm, which is why chip shooters don’t place fine-pitch ICs.

A flexible placer inverts that: slower at 6,000–40,000 CPH, but accurate down to ±25 µm and able to handle everything from 01005 passives to 150 mm connectors, including BGAs and package-on-package. Modular or all-in-one machines combine both behaviors and let you scale by adding heads or gantries. Here’s how they compare.

Attribute	Chip shooter	Flexible placer	Modular / all-in-one
Placement speed	20,000–100,000+ CPH	6,000–40,000 CPH	Configurable (blends both)
Placement accuracy	~70 µm	Down to ±25 µm	±25 µm and tighter
Component range	Small passives, ≤ ~20 mm	01005 to 150 mm, BGA/PoP	Wide, mixed
Architecture	Turret / revolver / gantry	Gantry / split-axis	Modular heads & gantries
Best for	High-volume passives	Fine-pitch ICs, odd-form	High-mix or scalable lines
Typical boards	LED, consumer, power	Complex / active boards	EMS, growing factories

Here’s the part that surprises people buying their first line: you usually want both, in series. A chip shooter places the bulk of the passives first, then a flexible placer follows with the ICs and connectors, because neither machine alone is optimal for a mixed board. And that pairing has a catch — a faster chip shooter does nothing for your output if the flexible placer is the bottleneck. Line throughput is set by the slowest station, so balancing the two matters more than the headline speed of either.

Pick and Place Machine Specs: CPH, Accuracy & How to Read Them

Placement datasheets are written to impress, not to inform, so learn to read past the headline. Two numbers dominate, and both are conditional. Components per hour (CPH) is almost always quoted under optimum conditions — a single small component, gang-picked, on a short travel path — so the real-world figure on a mixed board runs closer to 70% of the rated number once feeding, vision, and changeovers are counted. Placement accuracy in microns is meaningless on its own: it only means something paired with the component size it was measured on and the Cpk that describes how repeatable it is.

This is exactly the problem IPC-9850, titled Surface Mount Placement Equipment Characterization, was written to solve. It standardizes how placement speed and accuracy are measured and reported, ties accuracy to speed across a range of component sizes, and uses the process-capability index Cpk to score consistency — a machine at Cpk ≥ 1.33 places within spec with minimal defects. The test is deliberately boring and repeatable: components placed on 200 mm glass panels with sticky media (to remove board variables) against NIST-traceable fiducials, with 88,000 placements across 20 panels to get a statistically meaningful attribute-defect rate. The numbers below are vendor-rated, optimum-condition figures — useful for comparison, but derate them for planning.

Machine (illustrative)	Class	Rated CPH	Accuracy	Notes
Yamaha YRM20	High-speed modular	Up to 115,000	±0.025 mm (Cpk ≥ 1.0)	Rotary high-speed head
Panasonic NPM-WX	Modular	47,000 / 35,000	±25 µm / ±15 µm	Figures are head-dependent
JUKI RS-1R	Flexible high-speed	~47,000	Laser-centered	6-position adaptive head
NeoDen N10P	Desktop / economy	20,000	±0.01 mm	0201 to 40 × 40 mm

Notice that no single machine hits its best speed and its best accuracy at the same time — the high-accuracy mode and the high-speed mode are different configurations on the same datasheet. Read the footnotes: ‘under optimum conditions,’ ‘chip component,’ ‘single-lane,’ ‘Cpk ≥ 1.0’ are not decoration, they’re the conditions that make the headline true. The honest way to compare two machines is to ask each vendor for an IPC-9850 characterization on your actual component mix.

How to Choose a Pick and Place Machine: Selection Checklist

A wearables startup we worked with bought a high-CPH chip shooter expecting throughput to jump, then watched the line crawl. Their boards were IC-heavy — lots of QFN and BGA, very few passives — so the chip shooter sat half-idle while a single flexible placer bottlenecked everything at roughly a third of the expected output. The fix wasn’t a faster chip shooter; it was more flexible-placement capacity to balance the line. Buy for your board, not the brochure. Run this checklist before you sign.

1. Start with your board mix and volume. Count components per board, board sizes, and monthly volume. That sets the CPH class and how many machines you need — not the other way around.
2. Map your component range end to end. Identify the smallest part (0201 or 01005) and the largest (connectors, shields, odd-form). The machine has to cover both ends; 01005 placement wants ±15 µm or better and a high-resolution vision system.
3. Right-size accuracy to your hardest part. ±25 µm covers most consumer work. Fine-pitch BGA, QFN, and RF want tighter — and verify it with a Cpk ≥ 1.33 at that component size, not a best-case microns figure.
4. Count feeder slots and types. Enough slots to hold the entire BOM avoids mid-run changeovers. Check tape, tube, and tray support, and prefer electric servo feeders over pneumatic for accuracy and data feedback.
5. Weigh changeover time, not just speed. High-mix shops live or die on quick-change feeder carts and offline setup. Peak CPH is irrelevant if you change jobs every hour.
6. Plan the whole line, not one machine. Balance a chip shooter against the flexible placer that follows it, and budget for the rest of your SMT equipment — printer, reflow oven, and AOI — so no single station starves the others.
7. Budget for total cost of ownership. Feeders, nozzles, spares, and service add up, and maintenance often runs around 15% of ownership cost. Feeder lock-in between brands is real — mixing a Yamaha and a JUKI line means two feeder inventories.
8. Buy for real throughput. Derate the rated CPH to roughly 70% for capacity planning, and ask the vendor for an IPC-9850 report on your component mix before you commit.

Frequently Asked Questions About Pick and Place Machines

What is a chip shooter?

A chip shooter is a high-speed pick and place machine built to place small, simple components — resistors, capacitors, small actives — at very high rates, often 20,000 to 100,000+ CPH. It uses a turret or revolver head and trades accuracy (around 70 µm) for speed, so it doesn’t place fine-pitch ICs.

How fast is a pick and place machine?

It depends on the type. Chip shooters run 20,000 to over 100,000 components per hour; flexible placers run 6,000 to 40,000. Those are optimum-condition ratings, though — real-world throughput on a mixed board is typically around 70% of the headline number once feeding and changeovers are counted.

What is the difference between a chip shooter and a flexible placer?

A chip shooter is fast but low-accuracy, built for small passives at high volume. A flexible placer is slower but accurate to ±25 µm and handles a wide range — 01005 parts up to 150 mm connectors, plus BGAs and odd-form components. Most production lines run a chip shooter and a flexible placer together.

How accurate are pick and place machines?

Chip shooters place to roughly 70 µm; flexible and modular machines reach ±25 µm, and high-end units claim ±15 µm or tighter. Accuracy only means something with the component size and Cpk it was measured at — IPC-9850 ties accuracy to speed and uses Cpk ≥ 1.33 as the capability benchmark.

How much does a pick and place machine cost?

Desktop and entry-level machines for prototyping run roughly $10,000–30,000. Prosumer and mid-range machines for small to medium EMS run about $50,000–150,000. Industrial high-speed and modular machines run $200,000 to over $500,000, before feeders, nozzles, and service contracts.

What is CPH in SMT placement?

CPH means components per hour — the placement-rate spec for a pick and place machine. It’s almost always quoted under optimum conditions with a single small part, so treat it as a ceiling. For capacity planning, derate to roughly 70% of the rated CPH to reflect real mixed-board output.

Do pick and place machines place through-hole components?

They’re designed for surface-mount devices, not standard through-hole parts that insert into drilled holes. Some machines handle odd-form or pin-in-paste components, and many lines pair SMT placement with automated insertion or hand soldering for the through-hole parts that remain on the board.

What is IPC-9850?

IPC-9850, Surface Mount Placement Equipment Characterization, is the industry standard for measuring and reporting pick and place machine speed and accuracy on equal footing. It ties accuracy to speed, defines a glass-panel test vehicle with NIST-traceable fiducials, and uses Cpk to quantify placement capability across component sizes.

Match the Pick and Place Machine to Your Board

The right pick and place machine is the one that fits your board mix, your volume, and your changeover rhythm — not the one with the biggest CPH on the cover. Map your component range, right-size accuracy with Cpk, count your feeder slots, balance the line, and derate the headline speed before you plan capacity. If you’d rather skip the capital and run, send your Gerber and BOM and we’ll place your boards on a profiled, IPC-9850-characterized line as part of a DFM review.