SMT Tombstoning: Causes, Diagnosis and Prevention

Quick answer: SMT tombstoning happens when one end of a small two-terminal component wets and pulls before the other end can hold it down. The first checks are the two pad geometries, solder-paste deposits, placement, copper connections and a measured reflow profile. There is no universal stencil ratio, thermal-relief width or oven setting that prevents it on every board.

What is SMT tombstoning?

A chip resistor or capacitor “tombstones” when it rises onto one end during reflow, leaving the other termination open. Surface tension can lift the component when the solder at one pad wets earlier or more strongly than at the other. The visible defect is the result of an imbalance; the cause may be in the board, paste print, placement, component or thermal process. A photo of the finished joint alone cannot prove which one failed.

Start with the observed pattern, not a preset fix

Record the affected reference designators, lot, board position, component orientation and which end lifted. Compare the defect map with solder-paste inspection (SPI), placement inspection and the board layout. If the same end lifts across many boards, investigate asymmetry at that location. If defects move with a paste lot, component lot or oven recipe, test that variable before changing the footprint.

Keep a baseline defect count and change one factor at a time. A trial should compare equivalent boards and inspection criteria; otherwise a lower defect count may simply reflect a different sample or operating condition. Rework restores individual boards but does not identify a repeatable production control.

Check pad and copper symmetry

Compare the two pad dimensions, solder-mask openings and copper connections. A pad tied directly to a broad copper plane may heat differently from a pad on a narrow trace. Thermal relief, trace balancing or a footprint correction can be considered when the layout and defect map support that explanation. Do not apply one fixed spoke width or via distance to every design: clearance, current, heat dissipation and fabricator capability also matter.

Review nearby vias as part of the complete copper path, including whether an open via could draw solder away. A via is not automatically the root cause. For a new revision, ask design and assembly engineers to approve the revised land pattern and manufacturability check before changing released artwork. See QueenEMS’s PCB DFM review page for the wider handoff.

Compare both solder-paste deposits

SPI can show whether one pad received a different paste volume, shape or placement. Check stencil aperture design, stencil alignment, underside contamination, print release and paste handling. The correct aperture design depends on component size, pad geometry, stencil thickness, paste and printer capability. A claimed “1:1 ratio” or a fixed percentage reduction is not a universal rule.

If a print change appears promising, validate it on the actual board and examine other joints. Reducing paste to solve one defect can create insufficient solder elsewhere. Placement offset should be checked separately: even matched deposits may pull unevenly if the component lands off-center.

Measure the reflow profile on the assembly

Use thermocouples on a populated board or representative test vehicle to compare the affected region with other locations. Evaluate the profile against the solder-paste and component suppliers’ limits, including time above liquidus, peak temperature and maximum component temperature. A slower soak may help a board with a large thermal difference; a ramp-to-spike profile can also be appropriate. The profile is chosen by measured results, not by imposing one soak temperature and duration on all products.

AIM Solder’s lead-free reflow guide calls its profile values reference guidelines and recommends collecting a profile for each assembly at the beginning of production. It identifies both ramp-soak-spike and ramp-to-spike as possible approaches. That is why a universal claim that one profile “guarantees” no tombstoning would be misleading.

When should you change the design instead of the process?

If the defect repeatedly follows one footprint or copper connection after print and placement are stable, a board revision may be justified. If it follows a batch, printer or oven zone, inspect that process first. A robust corrective action names the suspected mechanism, the measured evidence, the approved change and a verification run. Avoid promising a yield improvement without a real before-and-after production record.

FAQ

Does tombstoning only affect 0402 components?

No. Small, light two-terminal parts are susceptible, but the risk depends on pad, paste and process conditions rather than one package label.

Will thermal relief always fix tombstoning?

No. It is worth evaluating when copper asymmetry is supported by the layout and defect evidence. It will not fix a blocked stencil aperture, poor placement or a material-wetting issue.

Should every board use a soak profile?

No. Compare measured temperature differences and supplier limits. A soak may help in some high-thermal-mass designs, while another profile may be better for a different assembly.

For a focused review, send QueenEMS the Gerbers, assembly drawing, BOM, stencil data if available, defect map, SPI images and measured reflow profile through the contact page. Those records make a proposed fix testable before the next build.