Solder Paste and PCB Stencil Design: What You Need to Know

Detailed Explanation

The solder paste stencil is the first manufacturing tool in the SMT assembly process — it controls exactly how much paste lands on each pad. Get the stencil design right and assembly runs cleanly; get it wrong and you're debugging tombstoned 0402 capacitors and bridged QFP pins after every reflow.

Unlike other assembly parameters (reflow profile, machine placement offsets) that can be adjusted by the assembly house, the stencil is a fixed physical tool that the designer largely determines through the paste layer in the PCB layout. Understanding the key rules saves one or more respin iterations.

What a Stencil Is

An SMT stencil is a thin flat sheet of metal — usually 430 stainless steel — with apertures laser-cut at each pad location. The aperture positions replicate the paste layer (F.Paste / B.Paste) from the PCB Gerber data.

During paste printing, the stencil is clamped flat against the PCB surface, and a steel squeegee blade is drawn across it at controlled pressure and speed. Paste fills the apertures and — when the stencil is lifted — transfers to the PCB pads. The quality of paste transfer depends on: aperture geometry, stencil thickness, stencil material surface roughness, and paste rheology (viscosity and particle size).

Stencil Fabrication Methods

Laser-cut stainless steel: Standard for most designs. The laser cuts through the stencil from one face, leaving walls that are slightly tapered (laser kerf angle). Good for pitches down to ~0.4 mm.

Electroformed (electropolished nickel): The stencil is grown from nickel around a mandrel, producing perfectly vertical walls with lower surface roughness than laser-cut. Releases paste more reliably at small apertures (useful below 0.3 mm pitch). More expensive and longer lead time than laser-cut.

Stencil Thickness

Stencil thickness is the single parameter with the most direct effect on paste volume. Thicker stencils deposit more paste; thinner stencils deposit less.

When an assembly contains both large passives needing 0.15 mm paste volume and fine-pitch ICs needing 0.10 mm paste, a single-thickness stencil is a compromise. A step stencil resolves this (see FAQ).

The IPC-7525 Area Ratio Rule

The area ratio predicts whether paste will reliably release from an aperture:

Area ratio = Aperture area ÷ (Stencil thickness × Aperture perimeter)

For a rectangular aperture of dimensions L × W in a stencil of thickness T:

Area ratio = (L × W) / (2 × (L + W) × T)

IPC-7525 specifies a minimum area ratio of 0.66. Below this, the surface tension of paste on the stencil walls exceeds the gravitational and squeegee-pressure force trying to push paste out, leading to inconsistent release.

Worked example:

An 0402 passive has a pad of roughly 0.5 mm × 0.65 mm. In a 0.12 mm stencil:

Area ratio = (0.5 × 0.65) / (2 × (0.5 + 0.65) × 0.12)
= 0.325 / (2 × 1.15 × 0.12)
= 0.325 / 0.276
= 1.18 ✓ (well above 0.66)

For a 0201 passive with a pad of ~0.25 mm × 0.35 mm in the same stencil:

Area ratio = (0.25 × 0.35) / (2 × (0.25 + 0.35) × 0.12)
= 0.0875 / (2 × 0.6 × 0.12)
= 0.0875 / 0.144
= 0.61 ✗ (below 0.66 — move to 0.1 mm stencil or electroformed)

Paste Volume Calculation

Paste volume deposited per pad (approximately) = Aperture area × Stencil thickness.

At ~50% solder volume (roughly the metal-to-flux ratio in paste), the final solder joint volume is approximately half the paste volume. This relationship is approximate — paste density, reflow conditions, and surface tension all affect the final joint geometry — but it gives a useful sanity check when comparing paste volume to pad size and expected fillet height.

For a given pad, the paste volume should be sufficient to form a solid fillet after reflow without bridging adjacent pads. Guidelines:

• For 0402 and larger: full-pad apertures (100% of pad area) are standard
• For fine-pitch QFPs (0.5 mm pitch): 80–90% of pad area is common to prevent bridging
• For BGAs: aperture design is closely linked to ball pitch and ball diameter — follow the component manufacturer's stencil recommendations

Common Solder Paste Defects and Their Causes

Tombstoning (drawbridging):
A 0402, 0201, or chip capacitor stands on one end after reflow. Cause: one pad reaches liquidus before the other, creating an unbalanced surface tension force. Contributing factors:

• Unequal paste volumes on the two pads (stencil aperture misalignment or asymmetric pad design)
• Thermal gradient across the component (one pad over a copper pour, the other over FR4)
• Pad sizes that are not symmetric (layout error or padstack difference between footprint pads)

Solder bridging:
Adjacent pads are shorted by merged solder. Cause: too much paste for the pad pitch, or paste deposited outside the pad area.

• Aperture too large relative to fine-pitch pad spacing
• Paste slump (paste spreading after deposition, before reflow) — related to paste rheology and storage conditions
• Stencil aperture reduction for fine-pitch pads (80–90% of pad area) is the standard fix

Insufficient solder / open joint:
Solder joint too small or absent. Causes:

• Area ratio below 0.66 — paste doesn't release from aperture reliably
• Stencil not flat against PCB (gap caused by a tall component on the other side of the board, or board warp)
• Paste dried on the stencil (paste pot too long since last use, low humidity)
• Pad not solderable (oxidised copper, contaminated OSP surface — see PCB surface finishes)

Solder balls:
Tiny solder spheres scattered across the board after reflow. Causes:

• Paste deposited onto solder mask rather than onto the pad (stencil misalignment or over-large aperture)
• Flux outgassing too quickly during preheat zone (preheat ramp rate too fast) — see reflow profile design for preheat zone ramp rate guidelines

Paste Layer in the PCB Layout

Most EDA tools generate the stencil automatically from the pad copper layer, adding a slight reduction (soldermask expansion in reverse) to produce the paste layer. This auto-generated paste layer is the starting point — the assembly house's DFM review may modify aperture sizes for fine-pitch areas.

Check the paste layer (F.Paste / B.Paste in KiCad) before generating output files:

• Every SMT pad should have a paste aperture (no paste on through-hole pads or test points that will be touched by the squeegee)
• No paste aperture on exposed thermal pads (package thermal pads for QFN/DFN) without intentional reduction — thermal pad paste should be ~50–70% of the pad area to prevent the IC from floating on excessive paste during reflow
• No paste on gold edge connector pads

Design Considerations

• Thermal pad stencil design for QFN/DFN packages: A solid paste aperture covering the full exposed thermal pad produces a very large paste volume — enough to float the component during reflow, causing bridging on signal pads. Standard practice is to use a segmented aperture (a 3×3 or 4×4 grid of smaller apertures within the thermal pad, covering ~50–70% of the total pad area) to reduce paste volume and allow flux outgassing paths.
• Stencil life and cleaning: A laser-cut stainless steel stencil can print thousands of panels if cleaned regularly. Most assembly houses clean the stencil every 10–20 panels; dried paste in an aperture reduces effective aperture size and degrades paste release. For prototype runs where the stencil is stored between uses, store it clean and dry.

Common Mistakes

• Not checking the paste layer in the EDA tool before generating output files, then discovering that thermal pads or test points have full-size paste apertures.
• Using the default EDA-generated paste layer for fine-pitch components without verifying the area ratio — auto-generated apertures for 0.4 mm pitch QFPs often need manual reduction.
• Designing a pad with a paste aperture covering solder mask web (the tiny gap between fine-pitch pads) — this deposits paste on the mask, which produces solder balls after reflow.
• Specifying a 0.15 mm stencil for an assembly that includes 0.4 mm pitch ICs, then troubleshooting persistent bridging that could have been avoided with a 0.1 mm stencil.