Stencil cleaning is the task that gets done when the line stops for something else. A wipe at break, a wash at the end of the shift, a full clean when the paste is changed. The schedule looks reasonable and it hides the real question: at which print does the aperture stop delivering the paste volume the SPI expects, on this board, with this paste.
Stencil cleaning frequency is a process setting, in the same way reflow profile and print pressure are. Set by habit, it is either too loose, and the defects arrive quietly in the middle of a run, or too tight, and the line spends time cleaning a stencil that was still working.
Why the interval is a decision, not a schedule
Two boards can run on the same line with the same paste and need different intervals. A stencil with 0.5 mm pitch apertures and a 0.12 mm foil clogs differently from one printing 0402 pads at 0.15 mm, because the paste has further to travel and less area to release from. A board with a large ground plane aperture spreads paste sideways and pulls it back on the next print, while a board of small apertures only loses volume.
The interval therefore belongs to the stencil and the paste, and the line has to be able to measure whether it held. That measurement exists: the SPI reads deposit height, area and volume on every pad, and it is the only honest way to know whether the last clean was needed.
What actually blocks an aperture
Paste is not a liquid that drains. It is a suspension of solder powder in flux, and the flux is what stays behind. Each print leaves a film on the aperture wall, the film thickens, and at some point the wall is rough enough that paste sticks instead of releasing. The aperture does not close suddenly. It loses volume first, then area, then it skips.
Aperture size sets the pace. Below an area ratio of about 0.66, which is the working limit in IPC-7525, paste release is already marginal on a clean stencil, so the same film that is harmless on a large aperture is enough to fail a small one. Nano-coated foils slow the film down, and they delay the interval rather than removing the need for one.
The under-wipe, the dry wipe and the solvent
An automated under-wipe passes a roll of fabric under the stencil and pulls paste out of the apertures from below. A dry wipe does most of the work on no-clean paste, and a solvent wipe is needed when the flux film has started to cure. Running a dry wipe on a paste that needs solvent polishes the film instead of removing it, which is worse than not wiping, because the aperture looks clean and releases less.
Vacuum assist changes the result again: it pulls paste through the aperture rather than pushing it back onto the board side. The combination is a machine setting, and it has to match the paste, not the shift pattern.
Signals on the SPI that say clean now
Deposit volume drifts down before a defect appears. A slow fall across a run, with the same stencil and the same paste, is the wall film building. A single aperture that drops out and comes back after a wipe is a clog. Paste smeared across the aperture edge, visible as extra area on one side, is the wipe itself leaving residue.
Bridging between fine pitch pads and missing deposits on the smallest apertures are late signals. By the time they appear, the interval has been too long for several panels, and the boards in between were built with less paste than the process assumes.
Intervals by paste and by print count
| Paste or condition | Working interval | Why |
|---|---|---|
| No-clean, large apertures, 0.15 mm foil | Dry under-wipe each print, solvent wash at paste change | Flux film builds slowly and the aperture has margin |
| No-clean, fine pitch below 0.5 mm | Under-wipe each print plus a solvent wash every few panels | Small aperture loses volume long before it skips |
| Water-soluble paste | Solvent wash far more often, and at every stop | The flux is made to be washed out and cures quickly in air |
| Paste left on the stencil during a long stop | Clean before restart, not after | Paste dries in the apertures while the line is idle |
| Change of paste type or alloy | Full wash between the two | Mixed residues behave unlike either paste |
Where the interval is stretched too far
The common pattern is a line that cleans at every break and not at every print. A break is a clock, and the number of prints between breaks changes with the board. A panel with 400 placements and a panel with 40 placements are both cleaned at the same time of day, and one of them ran ten times the prints.
The second pattern is a stencil kept on the machine between orders. Paste sits in the apertures overnight, the flux cures, and the next run starts on a stencil that needs a full wash rather than a wipe. The third is a paste change without a wash, where two residues sit in the same aperture.
How SUPERB PCBA sets the interval
The interval is set on the first article run, with the paste and the stencil that the order will use, and confirmed against 3D SPI data rather than against the clock. Where deposit volume starts to fall inside a run, the interval is shortened before the next panel, not after a defect. Stencil condition is tracked as a tooling asset, alongside the fixtures it runs with, so a foil that has been cleaned past its life is replaced rather than patched.
The cleaning equipment side is covered in stencil cleaning equipment, the asset records in tooling and stencil control, and the aperture geometry that sets the limit in stencil thickness and area ratio.
Questions buyers ask
Can a stencil be cleaned too often?
Yes. Every solvent wash and every wipe puts a small amount of wear on the aperture walls and the coating. A stencil washed far more than it needs loses its release properties earlier than one washed correctly.
Does cleaning frequency change the paste volume on the board?
It changes it if the interval is wrong. With the interval right, volume stays inside the same band from the first panel to the last.
Should the customer specify the interval?
The customer sets the deposit specification and the defect limit. The interval is a line setting that has to meet them, and it is confirmed on the first article.
Next step
Send the BOM, the Gerber set and the paste specification if one is fixed. The process review returns the cleaning method and interval the order will run with, and the SPI limits that would trigger a shorter one.