First decide how long this stencil has to live
The most common gap in a stencil enquiry is not foil thickness. It is that nobody says how many builds the stencil has to survive. Without that, material, frame and polishing cannot be settled.
The reason is simple: a stencil has two completely different identities.
One is a single-use tool — print two prototype boards, validate one revision of a pad design, hand-print a few samples, then archive it. Cost should be as low as possible, storage space matters more than service life, and an unframed foil is the standard answer.
The other is a long-term production asset — one product runs for a year, going on and off the printer every week with cleaning in between. Every mount is a mechanical disturbance. Foil tension, fiducial clarity and aperture wall condition all degrade with cycle count. The selection logic inverts: maintainable consistency first, cost second.
So the order is: answer "how long does it live and how many mounts" → framed or unframed → material → then foil thickness, polishing and coating. Run it backwards and you get the classic rework case, an unframed foil quoted for prototyping being asked to carry volume production.
Where each of the three materials stops
The deployed stencil service page lists three materials. They are not price tiers; they are three different process routes.
Stainless-steel foil (laser cut, standard) — the default for the overwhelming majority of jobs. The laser cuts each aperture from the paste-layer data, so aperture position follows the data, and the foil is stiff enough to take repeated squeegee shear. This is what laser surface mount stencils refers to. Unless there is a specific reason, a stencil should be laser-cut stainless.
Nickel (electroformed) — not "better stainless", a different way of forming the aperture. Electroforming grows metal along the aperture wall, producing a wall condition different from a laser cut one and friendlier to paste release. Its use case is small apertures with a tight area ratio — when the calculated area ratio sits right on the 0.66 rule-of-thumb threshold and the foil is already as thin as it can go. Whether that route is needed must be confirmed by engineering during review of the Gerber data, never from a board just "looking dense".
Kapton / PET — hand printing. Film materials lack the stiffness of metal foil, do not go into an automatic printer's tensioning fixture, and do not take production-level squeegee cycles. They solve a different problem: getting paste onto a handful of boards at almost zero tooling cost. Using them on equipment as a cheap stainless substitute is a selection error.
One line: stainless is the default; evaluate electroformed nickel only when the area ratio is tight; Kapton/PET only for hand-printed samples.
Framed or unframed changes more than price
On an order form this looks like a cheap/expensive choice. What it actually changes is four things: how the foil is tensioned, how repeatable each mount is, storage cost, and the volume it suits.
Framed (glue-in) stencils: the foil is bonded and tensioned into a permanent frame through a mesh border, so it leaves the factory already at its target tension. The deployed page describes this build as smooth, rigid apertures giving precise prints, double-bonding for high-wear use, non-fading fiducials, and suitability for high-volume and long-term production. The cost is a physical frame of storage space per product, plus the frame itself.
Unframed foils: only the foil is delivered; it is mounted into a reusable frame in the shop and removed afterwards. The deployed page describes minimal storage space, compatibility with screen and hand printing, fewer materials and lower environmental impact, lower cost than framed stencils, and superior print performance. What it really does is move tensioning from the tool to the shop's reusable frame — the consequence most often missed at selection time.
So the dividing line is not batch size itself but how many times this stencil will be mounted, and whether each tensioning is reproducible. Forty mounts a year on an unframed foil means forty tensioning events that depend on the shop's frame and operator. A framed stencil fixes that variable at the factory.
One-off validation, frequent revisions, prototyping → unframed foil. One product in long-term repeat production with batch-to-batch consistency requirements → framed. In between, check whether the shop's reusable frame sizes cover your board and whether the line has a stable tension verification routine.
Foil thickness and frame size: two numbers set by outside constraints
Foil thickness sets paste volume per unit area, available as 0.08 / 0.10 / 0.12 / 0.15 / 0.20 mm. It cannot be chosen alone; it goes with the smallest aperture on the board into the area ratio — aperture area ÷ aperture wall area, kept at 0.66 or above per the IPC rule of thumb. Below that, paste tends to stay in the aperture and you see insufficient deposits or opens. With 0.2 mm pitch BGAs or 0.3 mm pitch QFN/QFP on the board (the finest pitches our assembly supports), thickness moves toward the thin end. Only boards with large pads, connectors and high paste demand move toward 0.15–0.20 mm.
The conflict shows up on boards carrying both fine-pitch devices and large power pads: thin foil releases cleanly at fine pitch but starves the big pads; thick foil does the reverse. The way out is a step (multilayer) stencil or separate top/bottom stencils, not a compromise thickness that satisfies neither. That trade-off belongs to the process-level decision and is covered in the separate process-selection article rather than repeated here.
Frame size (framed stencils) comes in 370×470, 420×520, 450×550, 550×650 and 584×584 mm. It aligns not with the bare board but with the panelised outline plus the clamping margin and squeegee stroke the printer needs. Compare against the panel drawing, not the single board. Without an outline layer the frame match cannot be confirmed, which is one of the most common rework loops on stencil orders. Custom frames and oversized foils outside the usual range can be reviewed case by case.
Layer choice is settled here too: top, bottom, or top+bottom on one stencil versus separate ones. One stencil for both sides saves cost but both sides share one foil thickness — when the two sides need clearly different paste volumes, split them.
What polishing and nano-coating each solve
These get listed together as paid options, but they act at different points.
Polishing addresses aperture wall condition. Grinding polish is standard; electrolytic polish is an option that improves paste release. When the area ratio sits near the threshold, or one class of aperture keeps showing insufficient transfer, electrolytic polish is one of the measures to evaluate — its place is after thinning the foil and before redesigning pads.
Nano-coating addresses stability over the service life. The deployed page describes it as an option that cuts wipe frequency and extends service life. It does not change aperture geometry; it changes how the paste tends to adhere to aperture walls and the stencil underside. Its value therefore scales with print count: on a single-use prototype stencil it buys little, while on a long-running product where wiping eats cycle time it buys back uptime.
No wipe intervals or life-in-prints are given here — those depend on paste, board design and line parameters and require engineering review. What is certain is the direction: polishing serves "can this print at all", coating serves "will print 200 look like print 1".
Tension and cleaning: who owns them in service
Selection only covers the moment of delivery. Most consistency problems happen afterwards, so the boundary needs stating.
As-delivered condition is on the stencil supplier: foil tensioned to its target state (framed stencils arrive finished; unframed foils are delivered to suit the reusable frame), sample inspection of critical aperture dimensions and positional deviation, fiducials clean and free of burrs.
In-service condition is on the line, and three things are controllable:
- Mount reproducibility. Whether every mount reaches the same tension and alignment by the same method. Unframed foils depend on this heavily, because tensioning happens shop-side.
- Cleaning and wiping. Residue in apertures directly changes the effective aperture area — that is, it changes the area ratio you calculated. Wipe frequency is a line parameter, not a stencil parameter; nano-coating can relax it, not remove it.
- When to judge condition. Stencil condition shows up in paste deposits before it shows up in the stencil's appearance.
Point 3 drives the inspection strategy: do not judge a stencil by looking at it, judge it by post-print evidence. The evidence chain already exists on the assembly side: SPI → inline 2D AOI → post-reflow 3D AOI → X-Ray → ICT / FCT, accepted to IPC-A-610 Class 3, with voiding on hidden joints such as BGAs checked against <10% under nitrogen reflow, plus 100% electrical test on finished boards. SPI is the earliest sentry: same product, same aperture data, and when the SPI volume distribution starts drifting, suspect stencil condition and cleaning before touching pad design.
Specific tension values, deflection, cleaning agents and ratios, and life-in-prints are not stated here — all of them have to be confirmed by engineering against your equipment, paste and board design.
What to submit, and what engineering returns
The inputs for a stencil order are explicit:
- Files: Gerber RS-274X including the paste layer, solder mask layer and outline layer; or ODB++ or DXF. Without the solder mask layer, aperture-to-pad correspondence cannot be checked; without the outline layer, frame matching cannot be confirmed. Those two cause most rework loops.
- Specs to state on the order: material (stainless foil / electroformed nickel / Kapton-PET), framed or unframed, foil thickness (from 0.08–0.20 mm), frame size for framed builds, layers (top / bottom / both on one stencil or separate), quantity.
- Options: polishing (grinding standard, electrolytic optional and helps release), nano-coating.
- Context worth sending: expected mount count and production span, printer and reusable frame sizes, finest pitch on the board and whether large pads are mixed in. Those three drive every trade-off above.
Quantity and quotation are confirmed after review of the Gerber data, quantity and process requirements. That step cannot be skipped, because the area ratio check, the need for a step or split stencil, and frame matching all have to be read from the data.
Full specifications and the ordering route are on the SMT stencil manufacturing service page; assembly capability and inspection methods are under PCBA assembly; aperture design has to line up with what the board itself can hold, see manufacturing tolerances and capabilities overview; acceptance criteria and inspection methods are in quality assurance and PCB quality inspection methods.
FAQ
Q: How do I actually choose between framed and unframed?
A: By how many times the stencil will be mounted and whether tensioning is reproducible. One-off validation, frequent revisions and small prototype runs suit unframed foils — minimal storage, lower cost than framed, usable for screen and hand printing. A product in long-term repeat production with batch-to-batch consistency requirements suits framed: the foil is bonded and tensioned into a permanent frame via a mesh border, arrives at target tension, uses double-bonding for high-wear service, and keeps non-fading fiducials. Batch size is only a proxy; mount count is the real variable.
Q: What is the difference between laser-cut stainless and electroformed nickel, and should I upgrade?
A: Not a tier difference, a forming difference. Laser-cut stainless foil is the standard option and covers the overwhelming majority of boards. Electroforming grows the aperture wall, giving a wall condition friendlier to release, and its case is an area ratio sitting on the 0.66 rule-of-thumb threshold with the foil already as thin as it can go. Do not upgrade because a board "looks dense": calculate the area ratio from the smallest aperture and target thickness first, then let engineering confirm from the Gerber data.
Q: When is nano-coating worth paying for?
A: In proportion to print count. The coating cuts wipe frequency and extends service life without changing aperture geometry. On a single-use prototype stencil it buys little; on a long-running product where wiping consumes cycle time it buys back uptime. How far it relaxes wiping depends on paste, board design and line parameters and needs engineering review, so no interval figures are given here.
Q: Is frame size chosen from the board dimensions?
A: No — from the panelised outline plus the clamping margin and squeegee stroke the printer needs. Common sizes are 370×470, 420×520, 450×550, 550×650 and 584×584 mm, and you compare against the panel drawing, not the single board. Without an outline layer the frame match cannot be confirmed, which is the most frequent rework loop. Custom frames and oversized foils can be reviewed case by case.
Q: Are tension and cleaning stencil specs or line parameters?
A: As-delivered tension, sample inspection of critical aperture dimensions and positional deviation, and clean burr-free fiducials are deliverables. In-service mount reproducibility, wiping and cleaning frequency are line parameters. Judge stencil condition from post-print evidence, not from looking at the stencil: SPI volume distribution moves first, then inline 2D AOI, post-reflow 3D AOI, X-Ray and ICT/FCT, accepted to IPC-A-610 Class 3. If SPI drifts on unchanged aperture data, check stencil condition and cleaning before changing pads.
Q: Can Kapton / PET replace stainless to save money?
A: No. Those films are for hand printing. They lack metal foil stiffness, do not fit an automatic printer's tensioning fixture, and do not take production-level squeegee cycles. They exist to get paste onto a very small number of boards at almost zero tooling cost. Using them on equipment is a selection error and shows up directly as inconsistent deposits.
Q: My board has 0.3 mm pitch QFNs and large pads. What thickness?
A: Usually no single thickness works — thin foil releases cleanly at fine pitch but starves large pads, thick foil does the reverse. The way out is a step (multilayer) stencil or separate top/bottom stencils so each thickness zone meets its own demand. Our assembly supports BGA down to 0.2 mm pitch, QFN/QFP down to 0.3 mm, and 01005 as the smallest component; whether a step or split build applies is confirmed by engineering after Gerber review.