Ask a supplier for a stencil and the answer is a thickness, usually 0.12 mm. That one number sets the paste volume on every pad of the board, and on a mixed technology panel it cannot be right for all of them. A 0.4 mm pitch BGA and a press-fit connector a few centimetres away want paste volumes that differ by a factor of five. Most print defects do not start at the printer. They start with a stencil that was ordered from a default rather than from a calculation.
The area ratio decides whether paste leaves the aperture
Paste transfers because it adheres to the pad more strongly than to the aperture wall. The ratio of opening area to wall area predicts that. For a rectangular aperture, area ratio is width times length, divided by two times width plus length, times thickness. A laser cut stencil needs about 0.66 or better. Electroformed foil releases more cleanly and works down to roughly 0.50. Aspect ratio, width divided by thickness, is the older check and should stay above 1.5. Both numbers fall as apertures shrink and foil gets thicker, which is the entire argument for thin foil on fine pitch work.
| Pitch | Starting aperture | Foil for area ratio 0.66 | What it prints |
|---|---|---|---|
| 1.27 mm BGA | 0.60 mm round | 0.15 to 0.20 mm | Comfortable, thick foil carries volume for coarser parts |
| 0.80 mm BGA | 0.40 mm round | 0.12 to 0.15 mm | Stable at normal speeds |
| 0.50 mm QFN | 0.23 to 0.25 mm | 0.10 to 0.12 mm | Works, watch the paste wipe interval |
| 0.40 mm BGA | 0.20 to 0.25 mm round | 0.08 to 0.10 mm | Needs nano coated foil and a clean wipe |
| 0.30 to 0.35 mm WLCSP | 0.18 to 0.20 mm | 0.06 to 0.08 mm | At the edge of laser cut foil, electroformed preferred |
These are starting points. The decision is made by measurement: a 3D SPI volume window of 50 to 150 percent of nominal on the fine pitch apertures, with the height distribution narrow. A print that hits volume but has a wide height spread will still produce voids on a ground pad.
One board, several paste volumes
Mixed technology boards are where a single foil thickness fails. A step stencil keeps a thin body for the fine pitch area and steps up locally over connectors, press-fit regions and power pads. The step is laser welded, the position normally holds to plus or minus 0.10 mm, and the transition has a slope that the squeegee can climb without leaving a ridge. It is a tooling cost and a few days of lead time, and it is cheaper than a second print pass or a hand dispense step at the end of the line.
| Board content | Stencil decision | Reason |
|---|---|---|
| 0.4 mm BGA plus 1206 passives | 0.10 mm body, no step needed | Large passives reach a sound joint on less paste |
| Fine pitch plus press-fit or THT connectors | Step up to 0.18 to 0.20 mm locally | Pin in paste needs volume to fill the hole |
| QFN thermal pad | Window pane apertures, 60 to 80 percent coverage | Controls float and trapped air under the pad |
| RF shield frame | Segmented bead, not one long slot | A continuous slot bridges and smears |
| Heavy copper or cavity build | More volume, aperture at 100 percent of pad | The pad pulls heat away and the joint starves |
Where the print actually goes wrong
| Symptom | Usual stencil cause | What fixes it |
|---|---|---|
| Solder balls beside a fine pitch part | Poor release, paste smeared under the foil | Thinner foil, nano coating, wipe every few prints |
| Starved joints on 0.4 mm pads | Area ratio under 0.6 | Thinner foil, rounded aperture corners |
| Bridging between 0.5 mm leads | Aperture too wide, web too thin | Width at 90 percent of pad, web 0.15 mm |
| Tombstoning on 0402 | Too much paste per pad | Aperture at 80 to 90 percent of pad area |
| Voids under a QFN thermal pad | One large opening traps gas | Window pane, plus a reflow soak long enough to vent |
Paste, squeegee and profile belong to the same decision
Type 4 paste, 20 to 38 micron, suits 0.4 mm pitch. Type 3 covers 0.5 mm and coarser, Type 5 goes below 0.35 mm. Squeegee angle sits between 45 and 60 degrees, speed between 20 and 80 mm per second, pressure around 1 kg per 50 mm of blade length. On fine pitch work the underside wipe matters more than any of those: a dry wipe every five to ten prints keeps the foil clean enough that release stays predictable across a shift.
How SUPERB Assembly runs the print step
SUPERB Assembly runs seven SMT lines with automatic printers rated to 400 by 340 mm, 3D SPI on the line, and placement down to 01005 and 0.30 mm pitch on HDI builds. Reflow is a twelve zone profile under nitrogen below 500 ppm with SAC305, and the SPI record is kept with the lot, so a drifting print is caught on the first panel rather than after AOI. The stencil specification is agreed at NPI from the volume each pad group needs, and where a board carries both fine pitch and a press-fit connector, the step stencil is quoted with the tooling instead of being discovered at first article.
- SMT assembly workshop: lines, printers, SPI and reflow capability in one place.
- AOI and X-ray inspection: what each method sees after the paste has become a joint.
- PCB testing guide: the inspection and test steps a build passes through.
Frequently asked questions
Can one stencil cover 0.4 mm BGA and 0.65 mm QFN?
Yes, at 0.10 to 0.12 mm with nano coated foil, provided the area ratio is calculated for the smallest aperture and SPI confirms volume on both. If the QFN thermal pad needs a heavier deposit, add a step rather than thickening the whole foil.
Does a thinner stencil always print better?
No. Thinner foil releases better but deposits less, which weakens joints on connectors, power pads and heavy components. The right thickness is the one that satisfies the smallest aperture and still gives enough volume everywhere else, and that is a per board answer.
How is paste volume verified in production?
With 3D SPI on every panel: volume as a percentage of nominal, height, area and offset. The window is agreed at NPI, typically 50 to 150 percent, and the process is monitored on volume trend rather than on single readings, because one outlier is noise and a drift is a worn stencil or a dirty aperture.
Next step
Send the Gerber data, the BOM and the assembly drawing with the mixed technology list. The reply states the stencil thickness, the step map if one is needed, the paste type for each pitch group and the SPI volume window that will be applied, so the first article is a confirmation rather than a discovery.