Home

Company

Project

SMT Line-Up

Smart Production Line

Reflow Oven

SMT Stencil Printing Machine

Pick & Place Machine

DIP Machine

PCB Handling Machine

Vision Inspection Equipment

PCB Depaneling Machine

SMT Cleaning Machine

PCB Protector

I.C.T Curing Oven

Traceability Equipment

Benchtop Robot

SMT Peripheral Equipments

Consumables

SMT Software Solution

SMT Marketing

Applications

Services & Support

Contact Us

English
Bahasa indonesia
Сербия
Česky
Dansk
Deutsch
Español
Français
Hrvatski
Italiano
magyar
Nederlands
Polski
Português
Pусский
românesc
Slovenščina
Suomalainen
Türk dili
Tiếng Việt
العربية
فارسی
עִברִית
한국어
日本語
News & Events
As a global intelligent equipment provider, I.C.T has continued to provide intelligent electronic equipment for global customers since 2012. 
You are here: Home » Our Company » Industry Insights » Top 7 SMT Line Configurations for LED Lighting Manufacturing

Top 7 SMT Line Configurations for LED Lighting Manufacturing

Publish Time: 2026-11-21     Origin: Site

The Seven Configurations at a Glance

Rank

Configuration

Board signature

What decides the line

Inspection floor

TOP 1

Bulb and LED module line

Small round or square MCPCB, heavily panelised

Changeover time and feeder uptime, not placement accuracy

2D paste check + AOI

TOP 2

Tube and linear line

Boards up to 1.2 m, thin, warp-prone

Printer board length and conveyor support

SPI before placement

TOP 3

Panel light line

300 × 600 mm to 600 × 600 mm, large area

Thermal evenness across the panel and clean stencil release

SPI + 3D AOI

TOP 4

LED strip and flex line

Reel-fed, very high component count, low mix

Feeder pitch, nozzle choice, placement force

AOI with lens-damage rules

TOP 5

Street light and high-bay line

Thick MCPCB, high thermal mass, few large parts

Reflow zone count and profile control

SPI + X-ray on thermal pads

TOP 6

Driver and power supply line

Mixed SMT and through-hole on one assembly

The SMT-to-wave or selective-solder handoff

AOI + functional test

TOP 7

Automotive LED lighting line

High-reliability, traceability mandated

Closed-loop SPI and 3D AOI, plus data retention

Full SPI + 3D AOI + traceability

Two things worth saying before the detail. First, the ranking is by how broadly the configuration applies in the LED lighting industry, from the highest-volume commodity product to the most tightly specified. It is not a quality ranking — TOP 7 is the most demanding line on this list, not the worst. Second, no single configuration is "best". A factory that builds bulbs and street lights should not buy one line and hope; it should buy the TOP 1 configuration and subcontract or add a second line for the TOP 5 work.

---

TOP 1 — The Bulb and LED Module Line: Throughput Wins, Precision Follows

The bulb is the product that made LED lighting a mass-market category, and the line that builds it is the most misunderstood. Buyers assume precision is the constraint because the components are small. It is not. The constraint is how many boards per hour you can keep flowing without stopping for a reel.

What the boards actually look like

A typical bulb or module board is a small metal-core rectangle or disc, panelised many-up so the line can handle several boards per transfer. Component count per panel is high but the component mix is narrow: the same LED package repeated dozens of times, plus passives and one or two ICs. That narrowness is the whole economic story. Repetition rewards speed; variety punishes it.

Where this line loses time

Three places, in order of impact:

  1. Reel changes. A panel with a hundred identical LEDs empties feeders fast. If the line stops every forty minutes, your placement speed is irrelevant.

  2. Paste volume drift on small apertures. Small LED pads are less forgiving of paste variation than the pad size suggests. Excess paste floats the part; starved paste leaves an open joint that only shows up after burn-in.

  3. Polarity escapes. A reversed LED on a bulb board is a brightness defect, not a functional failure — it passes electrical test and fails the customer.

Building the bulb line

  • Loader and unloader sized for panel throughput, not for single boards.

  • Automatic stencil printer with a board envelope that comfortably exceeds your panel; a machine that handles 50 × 50 mm to 510 × 510 mm covers the great majority of bulb and module panels, and cycle time matters more here than micron-level positioning bragging rights. The I.C.T-5151 sits in this envelope, with positioning at ±0.01 mm and printing at ±0.025 mm, a cycle under 7 s, and a board thickness range of 0.4 mm to 6 mm — the last figure is what lets the same printer handle a thin bulb panel and a stiff driver board without a mechanical change.

  • Inline 2D paste inspection rather than a full 3D SPI, if the budget has to go somewhere. On a repeated small-pad pattern, area and offset are enough to catch the drift that matters.

  • One high-speed mounter plus one flexible mounter. The high-speed head eats the LED repetition; the flexible head handles the connector, the IC, and the odd-shaped part that would otherwise force a second pass. Current-generation high-speed platforms in this class run in the 115,000 cph range single-lane and up to 120,000 cph dual-lane — Yamaha quotes exactly those figures for its YRM20. Treat vendor cph as a ceiling measured under the vendor's own conditions, never as your line rate.

  • Reflow with enough zones to hold a stable soak, not the cheapest oven that reaches temperature.

  • AOI configured for polarity and lens condition, with the polarity library built before the first production run, not after the first complaint.

What to insist on at acceptance

Run a full panel, not a test coupon, and measure three things: paste volume on the smallest aperture across the whole panel (not the centre), placement offset at the four corners, and a deliberate reversed-LED sample that the AOI must reject. If the AOI passes the reversed sample, the inspection configuration is not finished — no matter what the machine's brochure says about defect coverage.

---

TOP 2 — The Tube and Linear Line: 1.2 m Boards Change Everything

This is the configuration that separates people who have actually built an LED line from people who have read about one. A T8 tube board is long. Not "large panel" long — 1.2 m long. Every machine in the line inherits that dimension, and most standard SMT equipment was never designed around it.

Why long boards break the usual assumptions

  • The printer is the first bottleneck. A machine with a 510 mm board envelope cannot print a 1.2 m panel at all, regardless of accuracy. LED-specific printers exist precisely for this: I.C.T lists a 1.2 m LED stencil printer and a long-board model as separate products from its standard range, because the requirement is mechanical, not optical.

  • Support is the second. A thin 1.2 m laminate sags between support points. Sag means uneven squeegee contact at one end of the print stroke and uneven reflow later. The fix is support pins and a conveyor that holds the board along its length, not just at the edges.

  • Warpage is the third. Long boards accumulate warp from reflow heat and from the panel's own residual stress. A line that prints flat and reflows warped will produce opens at the board ends, where nobody looks until functional test.

Configuring for length

  • Long-board printer rated for your actual panel length, with a support system matched to it. If you also build panels and drivers, a mid-size machine that reaches 900 × 600 mm — the envelope of the I.C.T-9060 — covers large panels without buying a dedicated long-board machine, but it will not print a 1.2 m tube.

  • SPI, not 2D. On a long print stroke, paste height drifts front-to-back. A 2D area check will not see it; a 3D SPI will, and the drift pattern tells you whether the squeegee or the support is at fault.

  • Two mounters in series, or one mounter with a long-board conveyor. A single head cannot keep up with a long panel's point count, and the transfer time between two machines is often cheaper than the cycle time lost on one.

  • Reflow with a conveyor that supports the full length and a cooling section that does not bow the board on exit.

The accessory that undoes the printer

Buying the long-board printer and keeping the standard conveyor. The printer will print correctly and the board will still fail downstream, because the sag moved from the print stroke to the transfer. Budget the handling as part of the configuration, not as an accessory.

---

TOP 3 — The Panel Light Line: Large Area, Even Heat, Clean Release

Panel lights sit between the tube and the street light in difficulty. The board is large and square — commonly 300 × 600 mm up to 600 × 600 mm — with LEDs distributed across the whole area rather than clustered. That distribution turns two ordinary process steps into the dominant yield drivers.

The two drivers

Stencil release on a large aperture field. A 600 × 600 mm panel has a lot of paste area. If the stencil does not release cleanly across the full field, you get paste smearing at the trailing edge — visible as brightness variation across the finished panel, which is exactly the defect a customer notices first in a lit room.

Thermal evenness through reflow. A large board has a large thermal mass and a large temperature gradient from edge to centre. If the centre runs cooler than the edges, the joints in the middle are formed at a different point in the profile than the joints at the perimeter. On a panel light, that shows up as a subtle colour or brightness shift across the lit surface — the kind of defect that passes every electrical test and still gets rejected.

The panel line configuration

  • Printer with the board envelope to match, plus a stencil tension and separation setup that suits large apertures. The 5151's envelope reaches 510 × 510 mm; larger panels need the 900 × 600 mm class machine or a dedicated large-format printer.

  • 3D SPI across the full field, sampled at the corners and centre, not just at the middle. The centre-only measurement is the most common SPI mistake on large panels.

  • Mounter with a large-board conveyor and stable support, because a 600 mm panel flexes during placement and flex changes the placement height.

  • Reflow with a profile developed on the actual panel, measured with thermocouples at edge and centre. If the two curves do not converge before the reflow zone, the oven's zone balance is wrong for this board.

  • 3D AOI, because a 2D system cannot distinguish a lifted LED from a correctly seated one under a diffuser.

---

TOP 4 — The LED Strip and Flex Line: Reel Thinking on a Panel Line

Flexible strip is the highest-volume, lowest-mix LED product, and its economics look more like packaging than like PCB assembly. The board is a long flexible circuit, often processed reel-to-reel or on carriers, and the same LED appears thousands of times per metre. Everything about the line should be tuned to continuity, not to flexibility.

What makes it different

The feeder, not the head, sets the rate. Strip production consumes LEDs at a rate that exposes every weakness in tape and reel: pocket tolerance, cover-tape peel force, splice quality, and reel-to-reel orientation consistency. A single mis-oriented reel does not produce one defect — it produces a run of them before anyone notices.

Placement force matters more than placement accuracy. A domed lens LED can survive a small placement offset and be destroyed by excess Z-force. High-power ceramic emitters need firm vacuum but a force limit. The nozzle choice and the force setting are product-specific, and they belong in the setup sheet, not in the operator's memory.

Lens damage is invisible to electrical test. A scratched or compressed lens passes every test and fails in the field as a light-distribution defect. Inspection has to be configured for it.

The strip line configuration

  • Printer sized to the carrier, with paste volume control tuned for a repeated small aperture across a very long pattern.

  • SPI if the strip carries dense fine-pitch drivers, or 2D paste check if the pattern is simple and repetitive. Be honest about which case you are in — the wrong choice here is expensive in both directions.

  • High-speed mounter with feeder capacity as the primary specification. Feeder slots and non-stop feeder exchange matter more than the headline cph. A dual-lane platform running in the 120,000 cph class is the right shape for this work; a single-lane machine with fewer slots will spend its life waiting.

  • Nozzle library built per LED package, with a written force limit for each. This is the single highest-value document on a strip line.

  • AOI with lens and orientation rules enabled, and a repair station that does not touch the lens with metal tools.

  • Handling that supports the strip along its full length, because a flex circuit will not hold itself flat.

For factories running strip alongside panel work, the practical answer is often a flexible mounter that can do both — the JUKI RS-1R class of machine exists for exactly this compromise — rather than a dedicated strip line that sits idle when the strip order ends.

Reel discipline: the document nobody writes

On a strip line, the highest-return piece of paperwork is a one-page reel acceptance rule. It states, per LED part number: pocket tolerance, cover-tape peel force range, splice method, orientation convention, and the quarantine action when a reel fails. It is not a quality manual; it is the thing that stops a mis-oriented reel from running for an hour.

Most factories have this knowledge in one supervisor's head. When that supervisor is on leave, output drops and nobody can say why. Write it down, put it at the feeder bank, and make the first article of every new reel a visual check against the orientation photo — not against the drawings, against a photo of a good reel.

Nozzle and force: the setup sheet that saves the lens

Every LED package on the line needs three recorded values: nozzle part number, pickup height, and Z-force limit. If the setup sheet does not carry them, operators will inherit whatever the last product left behind, and a domed lens will eventually meet a flat-chip nozzle at too much force. The failure is silent — the part still lights up.

---

TOP 5 — The Street Light and High-Bay Line: Thick MCPCB, Real Thermal Mass

Street lights and high-bay fixtures are where LED assembly stops being a speed problem and becomes a heat problem. The board is a thick metal-core PCB carrying high-power emitters, sometimes with a large thermal pad under each one. The thermal mass is an order of magnitude beyond a bulb panel, and the reflow oven has to deliver a profile that actually forms the joint on that mass without overheating the emitter.

Why the usual profile fails here

SAC305 melts at 217°C, and a conventional lead-free profile peaks between 230°C and 245°C with 45 s above liquidus — those are the figures in AIM's own SAC305 paste datasheet, not an industry rumour. On a thick aluminium-core board, the board reaches that peak later than the thermocouple on a thin test coupon suggests. The result is a profile that looks correct on paper and produces cold joints under the thermal pad, because the pad never spent long enough above liquidus.

The high-bay configuration

  • Printer with strong support and a paste deposit tuned for a large thermal pad. A 1:1 aperture on a big pad is how you get solder balling around the emitter; aperture design and stencil thickness have to be evaluated together.

  • 3D SPI, without exception. Paste volume on a thermal pad is the leading indicator of voiding later.

  • Mounter with force control and a vision system that can handle a tall, domed emitter. This is not a chip-shooter job.

  • Reflow with enough heating zones to build a real soak, and nitrogen as an option where the emitter or the finish justifies it. I.C.T builds a customised LED nitrogen reflow oven for this class of work; the point is not the brand but the zone count and the ability to hold a profile on a heavy board.

  • X-ray on the thermal pad, sampled per batch. AOI cannot see under the emitter, and voiding is the failure mode that kills a street light two summers later.

Air or nitrogen: choose by failure mode, not by brochure

Nitrogen improves wetting and reduces oxidation, and on a high-power emitter with a large thermal pad it can meaningfully reduce voiding. It also costs gas, consumables and a tighter oven enclosure, and it does nothing for a profile that is already wrong. The decision rule is simple: if your failures are oxidation-driven — poor wetting on an old finish, discolouration at the joint, voiding that persists after the profile is corrected — nitrogen is worth testing. If your failures are thermal-mass-driven, fix the profile and the zone balance first. Buying nitrogen to compensate for a shallow oven is an expensive way to keep the same defects.

The acceptance test that matters

Profile the actual production board with thermocouples on the thermal pad, not on the board surface, and record the time above liquidus at that point. If the pad's curve does not reach the wetting window the paste datasheet specifies, no amount of downstream inspection will fix the joint — you are inspecting a defect you built in.

---

TOP 6 — The Driver and Power Supply Line: SMT Meets Through-Hole

Every LED fixture needs a driver, and the driver board is a different animal from the LED board: fine-pitch SMT on one side, and real through-hole power components — transformers, electrolytics, connectors — that cannot be reflowed. This is the configuration where the line architecture, not the equipment list, decides the cost.

The architectural decision

You have three options, and they are not interchangeable:

  1. SMT line plus wave soldering. Highest throughput for a stable, high-volume driver. The cost is a second thermal process, a pallet design, and the floor space between them.

  2. SMT line plus selective soldering. Better for mixed or medium-volume work, and far kinder to a board that also carries fine-pitch SMT near the through-hole joints. A selective machine such as the I.C.T-SS550 exists for the case where you cannot flood the whole board with solder.

  3. SMT only, with through-hole parts hand-soldered or offloaded. Only defensible at low volume, and it quietly becomes the bottleneck as soon as demand rises.

What to specify

  • Printer and mounter that can handle the tallest component on the board — a transformer height is a conveyor and nozzle clearance question, and it is usually discovered late.

  • Reflow profile that respects the electrolytics. A capacitor rated for a certain reflow exposure has a limit, and the driver board is where that limit gets tested.

  • Flux management for the second soldering step. Wave and selective both leave residue that has to be cleaned or justified; the choice affects coating decisions downstream.

  • Functional test at the end of the line, not in a separate room. A driver that fails functional test three days later has already been packed, shipped internally, and counted as good output.

Pallet design decides yield

A driver board that goes over a wave or through a selective machine rides on a pallet, and the pallet is where most of the avoidable defects are born. Three specifics: the mask must protect fine-pitch SMT joints from solder splash, the hold-down must stop the board from lifting at the wave, and the pallet must not trap flux in a way that defeats your cleaning step. A pallet that was designed for an older board and reused by trimming will work well enough to pass the first batch and badly enough to fail the audit.

Where driver lines quietly lose money

Designing the SMT line first and adding the through-hole step afterwards. The handoff height, the pallet return path, and the buffer between the two processes determine the real line rate. Get them into the layout before the purchase order, not after the first week of production.

---

TOP 7 — The Automotive LED Lighting Line: Traceability or Nothing

Automotive lighting is the most tightly specified configuration on this list, and it is the one where "we inspect the board" is not an answer. A car headlight or tail light assembly is a safety-relevant part; the customer's audit will ask not only whether the joint is good, but how you know, on which board, at what time, with which parameters.

What changes compared with the other six

The inspection chain becomes closed-loop. 3D SPI does not just reject panels — it feeds back to the printer. 3D AOI does not just flag defects — it links to the SPI record for the same board so a reviewer can see paste volume and placement outcome together. That linkage is the actual product you are buying at this tier.

Data retention is a deliverable. Board-level traceability means the SPI and AOI records are retrievable by board ID, which means the marking step has to be in the line, not bolted on later.

The process window tightens on purpose. You will be asked for CPK on placement, for a documented profile per product, and for evidence that a change in paste lot did not shift the process. A line that cannot produce that evidence is not automotive-capable, regardless of its equipment list.

What an audit actually asks for

Automotive customers rarely open with a machine list. They open with a scenario: "show me one board from six months ago and its process record." If that request takes an afternoon of searching, the line is not audit-ready. The practical preparation is unglamorous — decide the board identifier format before the marking station is specified, define how long records are kept, and test the retrieval path once, with a real board, before the customer arrives.

The configuration for an automotive line

  • Automatic printer with closed-loop paste control, plus 3D SPI inline and networked to the printer.

  • Placement with force control and documented accuracy, verified with a glass board or a CPK study rather than a brochure figure.

  • Reflow with recorded profile per product, and a rule that the profile is re-verified after any heater or conveyor change.

  • 3D AOI inline, with the defect library under version control.

  • Laser marking or labelling inline so every board has an identity before it reaches inspection.

  • A data layer that survives a server rebuild. This is the part buyers forget: traceability that lives on one PC in the quality office is not traceability.

I.C.T has delivered this shape of project for LED manufacturing in Europe, including on-site line support for a German LED manufacturer covering installation, integration, training and trial production — the unglamorous work that decides whether a line is audit-ready.

---

Inside a Real LED Lighting Line: What 160 to 180 PCS/H Actually Took

Configuration advice is cheap until you watch a line come up. This is a real project, anonymised at the customer's request.

The situation. An automotive-lighting electronics manufacturer in Portugal was replacing an existing line with a new one built around a stencil printer, 3D SPI and two placement machines. The first product was a controller board, and the material and programs had to migrate from the previous platform. The plant's product range then widened from automotive lamps and controllers into sensors and radar modules — which is typical of LED lighting suppliers moving up the value chain.

What went wrong first. Almost nothing failed for the reason the team expected. The machines were mechanically sound; the losses were in the handover:

  • Programs migrated from the old platform needed PCB origin, MARK points, feeder pickup positions and component parameters re-taught — the old coordinates were close enough to start a run and wrong enough to lose yield.

  • One placement machine threw intermittent vacuum errors traced to a vacuum pump whose sensitivity was set too high, with the energy-saving function cancelling the vacuum between picks.

  • The printer and the SPI disagreed about signal state until the interface was re-commissioned.

  • The incoming power cabling was flagged for upgrade before the line ran continuously — the existing conductor size carried a safety recommendation.

The configuration that was installed. An I.C.T-5151 automatic printer, a 3D SPI, and two placement machines rather than one — the arrangement the plant chose for this product mix. Printing accuracy on this platform is ±0.025 mm, which matters on an automotive board where paste volume on a small pad is the difference between a pass and a rework.

The result. Line output moved from about 160 PCS/H to 180 PCS/H, with the final capacity and quality accepted against the customer's requirement, according to I.C.T project records. That is a meaningful gain — unglamorous, and exactly what a line replacement is supposed to deliver. The interesting part is what produced it: not a faster machine, but a line that stopped losing time to recalibration, vacuum faults and interface disagreements.

Why this matters for an LED lighting line. The same three failure modes appear on every LED configuration in this article. A strip line loses output to feeder orientation, a panel line to paste drift, a driver line to the through-hole handoff. If you are specifying a line, put the commissioning scope in the contract — the migration, the re-teaching, the interface test, the acceptance run. The equipment is the easy part to buy.

---

Why LED Boards Punish the Wrong Line

Four physical facts drive everything above. None of them are about the machines.

1. Thermal history becomes a product claim

An LED luminaire's lifetime number is not a marketing estimate; it is the output of a measurement chain. IES LM-80 requires lumen output and chromaticity to be measured "at least every 1,000 hours, for a minimum of 6,000 hours" at defined case temperatures including 55°C and 85°C, per the U.S. Department of Energy's solid-state lighting research. The extrapolation is governed by IES TM-21, which caps the projection at six times the tested duration — a 6,000-hour test supports an L70 claim of at most 36,000 hours, a 10,000-hour test at most 60,000 hours, as driver maker Bright Star summarises the rule.

Here is the consequence for a line builder: your reflow profile and your joint quality sit upstream of a claim measured in tens of thousands of hours. A cold joint under a thermal pad raises junction temperature. Higher junction temperature accelerates lumen depreciation. The customer's 50,000-hour number quietly stops being true, and nobody finds out for three years.

2. Efficacy keeps moving the component count

LEDs sold today average close to 100 lm/W, with premium and professional parts exceeding 200 lm/W, according to the International Energy Agency's 2026 lighting analysis. Every efficacy step means fewer emitters for the same lumens — so the placement count per board falls while the requirement for colour and brightness consistency rises. A line specified for yesterday's emitter count is over-equipped on speed and under-equipped on consistency. This is the single most common way an LED line ages badly.

3. The board is a heat sink, and it fights the process

Metal-core laminates conduct heat away from the joint during reflow — which is good for the LED and awkward for the solder. Thick boards need longer above liquidus; thin long boards warp instead. The same equipment cannot be optimal for both, which is precisely why this article separates the tube line from the street light line.

4. Polarity, lens and orientation are not chip-component problems

Standard AOI logic was built for chip components on FR-4. An LED has a polarity mark that may be under a diffuser, a lens that can be damaged without any electrical symptom, and a package height that changes the inspection angle. If the inspection library was configured for passives and reused for LEDs, it will report high coverage and miss the defects that matter.

Five defects that pass electrical test

This list is the practical reason LED lines need different inspection from general PCBA lines. Every item below can leave the factory as "good output":

  1. Reversed LED. Lights up, wrong colour point. Only a polarity rule in AOI, or a lit test, catches it.

  2. Lifted emitter on one pad. Contact is marginal; the joint reads fine at room temperature and opens when the fixture heats up.

  3. Voided thermal pad. Electrically continuous, thermally resistive. Invisible to AOI, obvious to X-ray.

  4. Compressed or scratched lens. No electrical symptom at all; shows up as a light-distribution complaint.

  5. Cold joint at the board end. Forms where the board ran cool through reflow; passes a quick continuity check and fails after thermal cycling.

If your inspection plan does not name which of these five it catches and how, it is a coverage claim rather than a plan.

Where the capacity is being built

Lighting consumed around 8% of global electricity demand in 2024 — about 2,200 TWh — in buildings and outdoor applications, and LED adoption is uneven: China and India lead with market shares exceeding 85% and 75% respectively, while parts of Africa, Southeast Asia and Latin America are catching up fast (IEA). Two practical implications for anyone buying a line: the next wave of capacity is being installed in places where local service and spare parts decide uptime, and it is being installed by factories that will widen their product range within two years. Specify a line that can grow, and a supplier who can reach you.

---

How to Choose: Match the Board, Then the Volume

Work down this table until one row fits. If two rows fit, you have a product range, not a product — plan two configurations.

If your board is…

And your volume is…

Build this

Why not the alternative

Small panelised MCPCB, narrow component mix

High, repetitive

TOP 1 bulb/module line

A precision line here buys accuracy you cannot sell

Up to 1.2 m, thin, warp-prone

Medium to high

TOP 2 tube/linear line

A standard printer cannot accept the board at all

300 × 600 mm to 600 × 600 mm, LEDs spread across the area

Medium

TOP 3 panel line

A bulb line's inspection density misses centre-to-edge drift

Flexible strip, thousands of identical emitters

Very high

TOP 4 strip line

A panel line's feeder count becomes the bottleneck

Thick MCPCB, high-power emitters, large thermal pads

Low to medium

TOP 5 street/high-bay line

A fast line with a shallow oven builds voids

Mixed SMT and through-hole power components

Medium

TOP 6 driver line

SMT-only forces hand soldering that caps output

Any of the above, with an audit and traceability requirement

Medium

TOP 7 automotive line

A non-networked SPI/AOI chain cannot produce the evidence

Floor space: the constraint nobody budgets

A long-board configuration is not just a longer printer. It is a longer printer, a longer conveyor, a longer reflow oven and a longer AOI, and the aisle space to service them. Factories that expand inside existing buildings discover this late, when the machines are already ordered.

The honest way to plan it is to draw the line at the length your tube product needs, then check whether a U-shaped layout fits the floor you actually have. A U-shape compresses the footprint at the cost of tighter conveyor routing, and for a line that shares a building with assembly and packing it is often the only workable answer. Decide the layout before the equipment list, because the layout changes the number of conveyors and buffers you need to buy.

Three questions to settle before the equipment list:

  1. What is the longest board you will run in the next three years? Not today — the answer usually doubles once a tube or linear product is added.

  2. What is the smallest aperture on your densest board? That, not the largest component, sets the printer and stencil specification.

  3. Who signs the acceptance run? If the answer is "the supplier's engineer", the configuration is not finished.

---

How We Ranked These Configurations

Five criteria, applied in this order:

  1. How widely the configuration applies across LED lighting manufacturing — the reason the bulb line ranks first and the automotive line ranks seventh.

  2. Whether the board's physical limits, not its component count, decide the equipment. Configurations where physics forces a specific machine ranking (long boards, thick MCPCB) rank above ones where a generic line would also work.

  3. Whether inspection is a closed loop or a checkpoint. Configurations that need SPI-to-printer or SPI-to-AOI linkage rank as more demanding, because they change what you are buying.

  4. Changeover elasticity — how badly the configuration punishes a product change. A line that cannot change over is a line that only works for one customer.

  5. Verifiability of the specification. We only cite machine envelopes and process windows that a buyer can check on a manufacturer's page or in a published standard.

Two deliberate exclusions. No pricing. Configurations of the same nominal size differ by several times in cost once printer board length, mounter head count, oven zone count and inspection depth are set, so any single dollar figure would be fiction — and the widely circulated LED line price ranges carry no stated basis. No discontinued models as recommendations. During research, one frequently recommended high-speed platform turned out to be marked "Discontinued Product" on the manufacturer's own specification page — the Yamaha YSM20R, still quoted at 95,000 CPH and ±0.035 mm (±0.025 mm), with the current equivalent being the YRM20 generation. If you are offered a discontinued model as new, that is a negotiation fact, not a defect.

---

Sources

  • International Energy Agency, The next wave of LED lighting: smarter, circular and more efficient (2026) — lighting electricity share, LED adoption by market, efficacy levels. https://www.iea.org/commentaries/the-next-wave-of-led-lighting-smarter-circular-and-more-efficient

  • U.S. Department of Energy / LSRC, LED Luminaire Reliability: Impact of Color Shift — IES LM-80 test requirements. https://www.energy.gov/sites/prod/files/2017/04/f34/lsrc_colorshift_apr2017.pdf

  • Bright Star Lighting, LM80 and TM-21: How LED Lifetime Data Relates to Driver Selection — the six-times projection rule. https://www.brightstar-power.com/info/lm80-and-tm-21-how-led-lifetime-data-relates-103699553.html

  • AIM Solder, WS488 SAC305 solder paste datasheet — alloy melting point, peak temperature and time-above-liquidus window. https://www.digchip.com/datasheets/download_datasheet.php?id=1595022&part-number=22456&type=prod

  • Yamaha Motor Europe, EFX 2026 SMT announcement — YRM20 placement speed. https://www.yamaha-motor.eu/gb/en/b2b/robotics-business-unit/surface-mount-technology/smt-news-and-events/smt-news/Yamaha-Robotics-to-focus-on-assembly-and-inspection-innovations-at-EFX-2026-in-Stuttgart

  • Yamaha Motor, YSM20R specification page (discontinued marking and rated accuracy). https://global.yamaha-motor.com/business/smt/mounter/ysm20r/spec/

  • I.C.T product specifications — 5151 printer envelope and accuracy, 9060 printer envelope, LED long-board printer range, LED nitrogen reflow oven. Linked at point of use.

---

FAQ

Which SMT line configuration is best for LED lighting?

There is no single best configuration, because LED lighting spans products with incompatible board physics. A bulb line and a street light line cannot be the same line without compromising both. Choose by board length, board thickness and inspection requirement first, then by volume.

Do I need SPI for LED production, or is AOI enough?

SPI and AOI answer different questions. SPI catches paste volume and offset before placement, which is the cheapest moment to fix a defect. AOI catches placement and solder defects after reflow, which is the most expensive moment. On tube, panel, street light and automotive configurations, SPI pays for itself; on a simple high-volume bulb panel, a 2D paste check can be a defensible compromise.

Can one line build both LED boards and driver boards?

Yes, with caveats. The driver board brings tall through-hole components and a second soldering process. If drivers are a small share of output, add selective soldering to an SMT line rather than building a dedicated driver line. If they are a large share, the handoff logistics will force a dedicated architecture.

Why is a 1.2 m tube board so much harder than a large panel?

Because the constraint is mechanical, not optical. Most printers are specified by board envelope, and a 510 mm envelope cannot accept a 1.2 m board regardless of accuracy. Beyond the printer, a thin long board sags between supports, which changes print contact at one end of the stroke and creates warpage downstream.

How do I know whether a quoted model is current?

Check the manufacturer's own specification page for the model before signing, and look for a discontinued marker. Model generations turn over quietly; a platform that was best-in-class three years ago may now be listed as discontinued while remaining widely quoted in catalogues and second-hand listings.

Is nitrogen reflow necessary for LED boards?

It helps where the emitter, the finish or the thermal pad justifies it, and it is not a substitute for a correct profile. On a thick metal-core board, a well-developed air profile beats a nitrogen profile that never gets the thermal pad into the wetting window.

What should the acceptance run include for an LED line?

A full production panel, paste volume measured on the smallest aperture across the whole panel, placement offset at the corners, a profile measured on the thermal pad rather than the board surface, and a deliberately reversed LED that the AOI must reject. If any of those five is skipped, the line is not accepted — it is merely installed.

How long does it take to get an LED line into stable production?

It depends on how much of the work is commissioning rather than installation. In the Portuguese automotive-lighting project described above, the measurable gain came from re-teaching programs, fixing a vacuum setting and re-commissioning an interface — none of which are equipment purchases. Budget commissioning time and a training scope into the project, and treat the first production week as part of the delivery.

---

About I.C.T

I.C.T supplies turnkey SMT, DIP and conformal coating production lines for electronics manufacturers, covering factory planning, equipment integration, installation and technical service. The company's LED-related work spans complete line delivery, long-board printing, inspection integration and on-site commissioning — including projects in Germany, Portugal and Central Asia. If you are configuring a line for a specific LED product family, the useful starting point is your board drawing and your target output, not a machine list.

Get a Quick Quote

Copyright © Dongguan ICT  Technology Co.,Ltd.