1. Get the definition straight: area ratio is not aperture width
Aperture size is the number people quote when ordering a stencil, but it is not the number that decides whether paste comes out of the hole. The SKPCB SMT stencil service page states what engineering actually looks at: area ratio = aperture area / aperture wall area.
The numerator is the face the paste presents to the pad as it releases. The denominator is the side wall area holding the paste back by friction. So the ratio describes something quite physical: the area pushing paste down versus the area holding it in place.
That is also why an aperture dimension on its own tells you nothing. Keep the aperture the same and change the foil thickness, and the wall area changes with it, so the ratio moves. Aperture size and foil thickness are not two parameters you optimise separately; they are two inputs to one ratio.
2. What the 0.66 line actually is: an empirical floor, not an acceptance criterion
The service page states this line with its qualifiers intact: it is an IPC rule of thumb that keeps the area ratio at 0.66 or above; below that, paste tends to stay in the aperture and you get insufficient deposits or open joints.
Three things have to be read at once:
- It is a rule of thumb, not an SKPCB acceptance criterion, and it carries no yield commitment.
- It describes a tendency. Below 0.66 does not mean the print is impossible, only that the process window has moved to the side where luck matters. The same stencil in that range may print today and fail after a paste lot change or a move to a different printer.
- The page states its own limit: what it gives are common reference ranges, and the released stencil data after engineering review governs the actual apertures.
Treating 0.66 as "0.67 means safe" is the most common misuse of this rule. Its correct use runs the other way: a calculated value below 0.66 tells you this aperture and this foil should not be released as-is, and something in the design has to change.
3. Foil thickness has five steps, so it is not a continuous dial
The service page lists foil thickness as 0.08 / 0.10 / 0.12 / 0.15 / 0.20 mm. That fact drives the order of operations: foil thickness moves in discrete steps, not fine increments.
When the area ratio comes out low there are two routes: open up the aperture, or go thinner on the foil. Thinning looks more direct, but one step is a whole step, and it applies to every pad on the board, including the large pads whose deposit was already adequate or already on the light side. Changing foil thickness is a global move; changing an aperture is a local one.
So the default order is: find the margin in the aperture first, and only consider stepping the foil once the aperture has nowhere left to go. How far the aperture can move is bounded by pad geometry, neighbouring pad spacing and the solder mask opening, which section 6 collects.
This order matches how the service page ranks the deciding variables: how often the stencil goes on a printer and the finest pitch on the board — budget comes after those. Finest pitch narrows the usable foil range first; the remaining freedom lives in the apertures.
4. Once finest pitch fixes the foil, look at the other end of the board
This is where mixed-technology boards get difficult. The finest-pitch component sets an upper bound on foil thickness, and under that thin foil the large pads on the same board will receive a light deposit. Apertures cannot fully recover it, because on the large-pad side the area ratio was generous to begin with. The problem there is not release, it is volume.
That conflict — two classes of pad on one stencil wanting different deposits — is a local-thickness problem. The route is a step or multilevel stencil, and the decision conditions and costs are covered separately in Step and multilevel stencils: when local thickness is worth changing. This article only marks the boundary: area ratio answers "does it come out", a step answers "is there enough". Mixing the two questions usually produces self-cancelling proposals like "thicken the foil and shrink the aperture".
5. Three non-geometric variables acting alongside the ratio
Area ratio is geometry, but whether paste releases cleanly from the wall also depends on three non-geometric things. The service page gives a specification for each; what follows is only which process owns them, with no added parameters:
Wall condition. The page lists two polishing routes: grinding polish as standard, electrolytic polish as an option, described on the page as one that improves paste release. So when the ratio sits tight and the design has no margin left, wall treatment is a process variable you can discuss instead of only redrawing.
Coating. Nano-coating is optional, and the page describes it as a way to cut wipe frequency and extend service life. Note that this is a gain in continuous-print stability and life, not a rescue for a ratio that does not stand up on its own.
Foil material. The page lists stainless-steel foil (standard), nickel (electroformed) and Kapton / PET (hand printing). Material and process route, together with frame tension, already have a dedicated article in SMT stencil material and frame tension selection. The point here is that material is an independent variable next to the ratio, not a substitute for geometry that falls short.
6. Aperture design has to line up with the board
One sentence on the service page should be read as a hard constraint: aperture design has to line up with what the board itself can hold. In numbers, that constraint comes from two pages:
- Manufacturing capabilities: minimum trace/space 2 mil/2 mil, minimum hole 0.075 mm. How fine the board can go sets how dense the pads get, and therefore how much room is left to enlarge apertures. The page also states that the listed windows are not universal commitments for every product, and that materials, stack-up, dimensions, processes and acceptance are confirmed from project data.
- Manufacturing tolerance (per IPC-6012, imperial governing): solder mask registration ±0.003", the mask opening needs to be 0.003" larger than the pad (mask swell), and pads spaced under 0.010" may need special handling. Together these set the physical ceiling on enlarging an aperture: push it far enough to collide with the mask dam and it stops being a stencil problem and becomes a solder mask problem.
So "the ratio is low, enlarge the aperture" holds on the drawing and may not hold on the board. This step means going back to those two pages for the boundary, not just editing numbers in CAM.
7. Four things to submit before ordering
The service page states that order quantity and quotation are confirmed after review of Gerber data, quantity and process requirements. To get a conclusion out of the first review, submit these four together:
- Files: Gerber RS-274X including the paste, solder mask and outline layers, or ODB++ or DXF. A paste layer alone leaves the section 6 mask and outline constraints uncheckable.
- Finest pitch on the board: the first input that fixes the usable foil range, and per the service page it ranks ahead of budget.
- How often this stencil goes on a printer: the page ranks this alongside finest pitch, and it drives framed versus unframed (framed for long-term volume, unframed for short runs and prototypes) and whether nano-coating is worth adding.
- How Top and Bottom are handled: the page supports top, bottom, or top+bottom on one stencil, or on separate stencils. When the two sides differ in pad construction, one stencil couples the foil decision across both.
If the stencil should travel with assembly, raise it together with PCBA / SMT assembly, where the BOM, placement data and revision review terms are stated.
FAQ
Q1: The area ratio works out to exactly 0.66 — can it be released?
Not as a pass mark. The service page frames it as a rule of thumb keeping the ratio at 0.66 or above, with paste tending to stay in the aperture below that. Sitting on the boundary means a narrow process window, so it should be judged in engineering review together with the polishing route, finest pitch and the board-side constraints, with the released stencil data governing.
Q2: The ratio is low — change the aperture or the foil?
Change the aperture first by default. The service page lists five discrete foil thicknesses (0.08/0.10/0.12/0.15/0.20 mm), and stepping one changes the deposit on every pad on the board, whereas an aperture change is local. Go back to foil thickness only once the aperture is blocked by the mask opening (which needs to be 0.003" larger than the pad) or by neighbouring pad spacing.
Q3: Can you quote from aperture dimensions alone?
Specifications yes, apertures no. The denominator of the area ratio is wall area, which requires foil thickness; and whether an aperture is buildable also depends on the mask opening and the outline. That is why the service page asks for Gerber RS-274X with paste, solder mask and outline layers, or ODB++ or DXF.
Q4: Can electrolytic polish or nano-coating rescue a low area ratio?
They address different problems. The service page describes electrolytic polish as improving paste release, a wall-condition variable; nano-coating is described as cutting wipe frequency and extending service life, which is continuous-print stability and life. Both can enter the discussion when the ratio is tight and the design has no margin, but neither replaces a geometry that does not stand up.
Q5: Large pads are starved and the fine-pitch parts will not allow a thicker foil — now what?
That is a paste volume conflict, not a release conflict, and area ratio will not solve it. The route is a step or multilevel stencil, with the decision conditions and the squeegee-sealing and cleaning costs covered in "Step and multilevel stencils: when local thickness is worth changing". Establish which class the conflict belongs to before touching local thickness.
Q6: What does putting Top and Bottom on one stencil cost?
The service page supports both one stencil and separate stencils. When the two sides differ substantially in pad construction, one stencil means the foil thickness has to satisfy the stricter side, coupling two decisions that could have stayed independent. Print frequency and whether the two sides have comparable finest pitch are the main inputs to that call.