
New or expanding plant
You're selecting a whole line. Start at Can You Match My Cell and Will My Floor Take It.
See how it matchesWe build the middle of the crystalline-silicon module line — cell in-feed to pre-lamination: laser scribing, tabber-stringing, robotic layup, bussing and test, configured to your cell route and busbar count.
We do one thing and configure it three ways. BC, TOPCon and PERC each get a dedicated machine — not a general stringer padded out to fake BC.
At a glance
Four numbers tell you the class of line you're dealing with — no reading required. Each is proven against a technical agreement.
pcs/h on the SML-S60 flagship stringer.
On A-grade cells at the scribing station.
A floor number, held over 24 hours.
Across the range, 25BB-ready.
Utilization here means 1 minus unplanned downtime over 24 hours — a floor number you can hold, not a peak reading you'll see once and never again.
Our segment
Buying the wrong scope wastes weeks. So here's exactly what we cover: the cell after it enters your plant, up to just before lamination — six stations in one chain. Everything before and after, we don't touch.
Wafer and cell manufacturing (ingot pulling, wafering, texturing, diffusion, PECVD, printing), laminators, framing machines, cleaning machines, junction-box mounting, plant EPC and module installation. If a supplier claims all of it, that's how inquiries get burned.
Find your path
Three situations, three paths — so you can place yourself in three seconds and skip what isn't yours.

You're selecting a whole line. Start at Can You Match My Cell and Will My Floor Take It.
See how it matches
You want a single-station swap without stopping the whole line. Start at Lossless Scribing and What Makes BC Hard.
See the swap path
You're short one machine and it has to drop into a running line. Start at Core Line Equipment and Changeover Boundary.
See drop-in optionsMatch your cell
Each route, size, thickness and throughput has a machine that fits — and size is banded by model, not a blanket claim. 156–220 mm is the site-wide union, but stringers (S60 / S40 / S15) take 166–210 mm and only the scriber SML-C20 reaches down to 156 mm. No machine here "does it all."
| Dimension | BC Stringer · SML-S60 | Multi-Busbar · SML-S40 | Entry Stringer · SML-S15 | Entry Scriber · SML-C20 |
|---|---|---|---|---|
| Process stage | Stringing | Stringing | Stringing | Scribing |
| Route / cell | BC | PERC / TOPCon | TOPCon / PERC | Mono / PERC (no back-Al cells) |
| Cell size | 166–210 mm | 166–210 mm | 166–210 mm | 156–220 mm |
| Throughput | ≥4800 pcs/h | ≥3400 · ≥3200 pcs/h | 1100 pcs/h | ≥1600 pcs/h |
| Cell thickness | 110–200 μm | 110–200 μm | 120–200 μm | 120–220 μm |
The multi-busbar SML-S40 runs ≥3400 pcs/h at 3–12BB and ≥3200 pcs/h at 13–20BB — one machine, two throughput bands as the busbar count climbs. Thickness is banded too: the stringers hold 110–200 μm (the entry S15 from 120), while the SML-C20 scriber runs 120–220 μm — worth checking against your route if you cut on the thin edge.
The gap you cut at scribing travels all the way to the module.
Lossless cut
One cut decides whether defects ride through to the finished module. Laser grooving plus thermal cleaving is lossless — the microcracks, chips, dust and mechanical loss that mechanical cleaving leaves behind, it simply doesn't create. Chipping rate holds at ≤0.05% on A-grade cells.
| Dimension | Traditional Mechanical Cleaving | Laser Non-Destructive · SML-C72 |
|---|---|---|
| Cutting method | Mechanical snapping | Laser grooving + thermal cleaving |
| Microcracks | Left along the break edge → hidden-crack source | None after cutting |
| Chipping rate | Inherently higher; worse on thin cells | ≤0.05% (A-grade cells) |
| Processing dust | Heavy dust, needs fire precaution | Minimal dust, no fire concern |
| Surface / mechanical | Surface damage, reduced strength | Minimal damage, strength retained |

One clean cut removes four downstream problems at once: no hidden-crack source at the break edge, no chip-driven stops at later stations, no shop-floor fire precaution around cutting dust, and no mechanical stress carried into stringing. The result comes from the method — water-spray cooling at the cut shrinks the heat-affected zone, so the silicon is never thermally overstressed.
Why trust us
For a BC string, the consistency of seam exposure and cell pitch decides the module's EL appearance and power uniformity. It's the one chokepoint of the whole BC line — so here is exactly how we hold it.
Seam exposure ≤±0.3 mm (ribbon centerline to pad centerline) directly sets EL appearance and power consistency.
Out-of-spec exposure means cold solder and hidden cracks, which means the whole string is reworked. Let rework rate climb and your rated capacity is fiction.
A four-axis industrial robot with 0.01 mm repeatability, plus CCD correcting cell by cell so each busbar meets its ribbon.
Positioning ≤±0.05 mm, angle ≤±0.04°, pitch accuracy ±0.3 mm, string straightness ±0.5 mm.
Inline string EL inspection catches cold solder and hidden cracks. Optional equipment, not standard
Defect catch
A chipped or cracked cell welded into a string scraps the whole string. So the cheapest place to catch it is the earliest — at scribing in-feed, by optical inspection with automatic NG rejection.
The earlier a bad cell is rejected, the less work is wasted on it.
One bad cell reaching the stringer takes the whole string down with it.
In-feed CCD checks position, chipping, missing corners, cracks, 90° flip, dirt and scratches, with automatic NG rejection.
Out-feed re-checks for cracks, chipped corners and off-size cells before the cell moves on.


A bad cell caught at in-feed costs one cell. Caught at the module, it costs the string.
Why BC is hard
Negative pitch is an extra gear a regular stringer can't shift into — and it's where SML-S60 and SML-S40 draw their capability line. SML-S60 gives both gears (negative −1 to −0.5 mm, positive 1–5 mm); SML-S40 gives positive only. That's a boundary, not a verdict that S40 is worse.
Cells sit nearly edge-to-edge with almost no gap, enabling layouts a regular positive pitch can't produce.
Negative pitch means cell-on-cell placement with far less geometric tolerance; a regular stringer can't hold that gear.
SML-S60 supports negative −1 to −0.5 mm and positive 1–5 mm; SML-S40 stays at positive 1–5 mm.
Busbars run 3BB–20BB with 25BB-ready headroom; stepping up busbars is a tooling change, not a new machine.
Core equipment
Three of the machines you'll configure a line around — one representative figure each, full parameters on the product page behind the click.

It works the stringing station of a BC line, turning laser-cut cells into finished strings; that length range spans short module strings up to full-length ones, so one machine covers your product mix without a hardware swap.
See the SML-S60
It opens the line at scribing, cutting each cell in two stages — laser grooving, then thermal cleaving — with no blade; the long source life keeps that first, defect-setting cut consistent over years of running.
See the SML-C72
It hands strings from stringing onto the glass in module geometry; its 1230 mm docking height meets the stringer's string outfeed, so the two machines line up directly with no transfer table between them.
The full core set is SML-S60, SML-S40, SML-S15, SML-C72, SML-C20 and SML-A6.
See all equipmentFull range
Beyond the core machines, the range covers the combined, high-throughput and finishing stations — so a full BC or TOPCon module line can come from one source instead of being stitched together from four.

Scribe-and-string combo that folds cutting and stringing into one machine, so a compact line loses the transfer and buffering between those two stations.

Fully automatic bus bar welding station that joins finished strings into the module circuit — the bussing step between layup and final electrical test.

Layup-and-weld combo that places strings on the glass and welds them in a single station, compressing two steps of the line into one footprint.

Inline frame marking machine that marks each finished frame in the flow, so traceability is added without pulling modules off the line to label them.

Module IV tester that measures finished-module electrical output at the test station, offered across multiple power grades to match your module range.

EL inspection unit that images the laid-up module for hidden cracks and cold solder before it leaves the line for lamination.
Your floor
Before anything ships, check these against your building — so a mismatch shows up now, not on the delivery dock. Read the floor load first: ≥600 kg/m². The flagship SML-S60 weighs 7000 kg on its own; an under-rated floor sinks and vibrates, positioning drifts, and the seam consistency built up earlier is lost. It's the condition most often overlooked until it's too late.
The second trap is measuring the wrong dimension. The footprints below are the machine's own outline; what you actually have to clear is the installation space with room to operate around it. For the SML-S60 that's 9000×3500×2800 mm, not the 8050 mm body — the SML-S40 needs 8000×1800×2650 mm and the SML-S15 7000×2000×2500 mm. Lay the line out to the operating envelope, or the machines fit on the drawing and jam on the floor.
One more the drop planners miss: the SML-S60 needs both φ16 air hoses feeding at once, not a single line teed off — size the compressed-air drop for simultaneous supply. Two machines also break the shared 380 V pattern: the SML-C20 scriber runs on 220 V single-phase, and the SML-A6 layup robot on single-phase three-wire. The A6 ships behind a safety door or fence, so leave its guarding clearance in the layout as well.
The full shared utility spec — power feeds, compressed-air quality classes, environment band and floor rating, machine by machine — is laid out on the line utilities and layout page →
| Model | Footprint (mm) | Weight | Power inlet | Air / Exhaust |
|---|---|---|---|---|
| SML-S60 | 8050×2800×2500 | 7000 kg | 4×16 + 1×10 mm² | ≥1200 L/min (2×φ16) / >600 m³/h · φ110 ×3 |
| SML-S40 | 7850×1600×2500 | 4500 kg | — | ≥1200 L/min (2×φ16) / >600 m³/h · φ102 |
| SML-S15 | 6900×1800×2300 | 3000 kg | 4×10 + 1×6 mm² | ≥600 L/min (1×φ16) / >300 m³/h · φ102 |
| SML-C20 | 1000×650×1500 | 400 kg | — | 0.5–0.8 MPa (φ12 inlet) |
Know the floor load and the operating envelope before the crate arrives, not after.
Changeover
Can one machine handle every spec I run? Within a compatible range, yes; past it, the answer is tooling. Knowing where that line sits up front is the difference between a spec you can plan around and one you only discover at acceptance.
Equipment is built and accepted against a single cell blueprint you provide and sign; that signed drawing is the reference the acceptance spec table is read against.
Within the compatible range, other specs need matching tooling, billed separately.
Within 3BB–20BB, stepping up busbars is a tooling change; 25BB is ready.
If a spec falls outside a machine's compatible range, it needs checking first, not a one-machine-fits-all assumption.
How to start
Three steps to a configuration proposal — we propose against your cell first, you decide second, and the uncertainty of the buying process comes off the table early.
Size, route, busbar count.
NowMatched to your cell and floor.
Reply within 24 hoursAcceptance is judged by the agreement's spec table.
After confirmationDelivery
Acceptance is measured against the technical agreement's detailed spec table, in black and white. Our team installs and commissions on site. A good-looking capacity slide means nothing — if it doesn't hit the agreed spec table, it hasn't passed.
Our engineers carry out installation and commissioning on your floor, not a remote handover.
The run proves the line only on real inputs: qualified, sufficient raw material and qualified, sufficient operators on your side during acceptance.
If acceptance is held up by conditions on the customer's side, that time isn't charged against the line's measured performance.
Acceptance is judged against the technical agreement's parameter table, not a promise or a slide.
Training
Free training before acceptance — at least once, at most twice — sized so your people can run and maintain the line without calling us back. The hours that decide independence aren't the operation basics; they're the troubleshooting-and-upkeep block, where operators learn to read an abnormal stop and change a wear part themselves.
| Sessions / cost | Free before acceptance, at least 1, at most 2. |
|---|---|
| On-site operation, 10 h | Principle 1 h + operation 4 h + process parameters 4 h + system parameters & maintenance 1 h. |
| Commissioning, 3 h | Abnormal-cause diagnosis & handling 2 h + parts replacement & daily upkeep 1 h. |
| Safety | Safety-knowledge training on in-process protection. |
After-sales
Once the warranty year ends, you're not cut loose. The first year is a free warranty; after it, the machine gets lifetime repair service, and the parts that wear — suction cups, springs, filters — stay in long-term supply. A machine well past its warranty still has a defined path back to running, not a dead end the first time a wear part fails.

Get a proposal
Low-effort next step: send your cell spec, get a configuration proposal within 24 hours.
Or keep looking — see the equipment · see the process.
Reply within 24 hours · free 1-year warranty + lifetime repair · 48h response / 72h solution.