Skip to the stories
Electronics & Semiconductors

PCB Tombstoning: Causes, Inspection and Process Controls

  • li
  • 923 reads

Uneven wetting can lift a chip component during reflow. Paired checks of pads, paste, placement and temperature help pinpoint the cause.

PCB Tombstoning: Causes, Inspection and Process Controls

A chip component that stands on one end has lost a connection, but its appearance alone won’t reveal why. PCB tombstoning occurs when unequal soldering forces rotate a two-terminal component during reflow. The practical task is to find what made the two joints behave differently: pad and copper geometry, paste deposits, placement, component condition or heating history.

Tombstoning PCB defect with one capacitor termination lifted away from its pad

Start the investigation at the affected reference designator. Compare its two joints and the conditions immediately before reflow, then check the same component at a known-good location. A profile adjustment on its own may hide a symptom while leaving the cause in place.

What happens when a chip tombstones?

During reflow, molten solder wets the component terminations and the board pads. If one joint wets first, or if the forces acting at the two ends differ, the component can rotate and lift away from one pad. The other end may remain attached, leaving an open connection at the lifted termination.

A component can stand nearly upright or rise only partway. Both conditions matter: partial lift can also interrupt electrical contact. “Manhattan effect” and “drawbridge effect” describe similar appearances. A chip shifted sideways is a different defect; inspect the orientation and both terminations before classifying it.

Balanced wetting compared with one end lifting while the other remains soldered

The circuit determines the functional consequence. An open series resistor can break a signal path. A disconnected bypass capacitor may still allow the product to power on while reducing the intended decoupling. A functional check should therefore reflect what the affected component does.

Small packages need closer process control

Small chip resistors and ceramic capacitors are common candidates because their low mass gives an imbalance less resistance to rotation. Packages such as 0201 and 0402 warrant close attention to paste printing and placement overlap. Package names need their sizing convention: an imperial 0402 is not the same size as a metric 0402.

Imperial / EIA designationMetric designationNominal body length × widthAssembly consideration
040210051.0 × 0.5 mmCheck termination overlap and copper connections at both ends.
020106030.6 × 0.3 mmSmall printing or placement differences can become significant.
0100504020.4 × 0.2 mmConfirm package-specific printing, placement and inspection capability.

These are nominal body dimensions, not pad dimensions or acceptance tolerances. Use the exact component drawing for body height, termination length and recommended footprint. A resistor and capacitor with the same nominal package can behave differently because their termination construction and solderability may differ. Other two-terminal parts can also lift, so diagnose the actual component rather than assuming every case involves a capacitor.

Nominal EIA 0402, 0201 and 01005 chip body dimensions

Check pads and copper before changing the oven

Compare each pad’s length, width, spacing and exposed solderable area with the component manufacturer’s recommended land pattern. Oversized or asymmetric pads can change solder shape and the forces on the terminations. Reducing pad dimensions without checking attachment requirements may replace a tombstone problem with an inadequate joint.

Then trace the copper connected to each pad. One end tied to a broad plane and the other connected through a narrow trace may heat differently. Internal copper, nearby vias and surrounding components can also affect the local thermal response. A recurring lifted end beside a copper plane makes that area worth investigating; it does not by itself establish the plane as the cause.

Matched copper connections or thermal reliefs may reduce a difference, but must still meet current capacity, grounding and signal needs. An unfilled via near a solderable pad can also affect solder volume. Review the pad, solder-mask opening, copper and via treatment together before changing the layout.

Comparison of unequal copper connections and matched connections at a chip footprint

Compare the two printed deposits

Unequal solder paste can leave one termination higher or in contact with less paste before melting. Differences in solder volume may also change the forces at the ends of the component. Check the two deposits’ volume, height and position individually; a board-wide average can conceal a local mismatch.

Potential contributors include blocked stencil apertures, registration error, contamination on the stencil underside and inconsistent paste release. A symmetric aperture design doesn’t guarantee that the printed deposits will be symmetric. An SMT stencil needs to suit the package and board, and the resulting print should be checked on the assembly.

If measurements identify excess paste at the outer ends of a chip as a contributor, a controlled aperture change may be useful. A universal reduction isn’t appropriate for every package. Verify the resulting joints, attachment and print repeatability, and watch for opens or poor wetting elsewhere.

Verify placement and thermal history

Offset or skew can give the two terminations unequal overlap with their paste deposits. A placement-height error can leave one end without proper contact even when the image appears centered. Check package definitions, nozzle condition, pickup, board support and placement settings at the affected reference designators.

Don’t rely on solder self-alignment to correct a component placed beyond its process tolerance. Increasing placement force without checking component limits can damage a ceramic body or displace paste. When available, retain the pre-reflow image; it can help distinguish an initial placement error from rotation during melting.

Measure the temperature history at the assembly before changing the oven recipe. Zone setpoints don’t show the temperatures reached at the two joints. Compare the affected area with a stable location and assess whether the joints approached melting at different times. Small chips can be difficult to measure because sensor size and attachment may disturb the reading, so use a consistent method and a suitable instrumented board.

Preheat or soak adjustments may reduce temperature differences, but excessive exposure can deplete flux activity or exceed a material’s recommended limits. Keep the measured profile within the solder paste window and component limits. Raising the peak can’t correct a poor footprint or unequal paste deposits.

Changes to atmosphere or alloy also need controlled trials. Nitrogen can accelerate wetting and may worsen an existing imbalance; it isn’t an automatic fix. Switching to air can affect other joints. Judge the full assembly result, not only the lifted component.

Record paired measurements

Use measurements from the affected component alongside a known-good position or build. Numerical limits must come from the selected component, paste and qualified assembly process; a generic checklist cannot set production acceptance criteria.

ParameterUnit or calculationWhat to recordHow to use it
Body and termination dimensionsmmPart number, package convention, body and termination dimensionsConfirm the land pattern and placement definition.
Pad and mask geometrymm; mm²Length, width, gap and exposed area at both endsCompare the drawing with the exposed pads.
Stencil aperture and thicknessmm or µmAperture dimensions, shape, foil thickness and revisionReview deposit design and local release behavior.
Printed paste volumemm³V1 and V2 for the paired padsCompare individual deposits, not just a board average.
Relative volume difference100 × |V1 − V2| / [(V1 + V2) / 2]Percentage and original volumesCompare trials consistently; this is not a pass/fail limit.
Paste height and positionµm; X/Y offset in mmBoth deposits and their positions relative to the padsIdentify unequal seating or print registration.
PlacementX/Y in mm; rotation in degreesOffset, skew and approved Z-height or force settingsCheck overlap before reflow.
Temperature history°C and secondsMeasurement locations, preheat, peak and time above the alloy’s liquidusCompare the board profile with material limits.
Heating rate°C/sTemperature interval and measurement methodAssess changes without treating zone settings as board temperature.
Defect frequencyDefective components / inspected opportunitiesReference designator, panel position, lot and trial sizeCheck whether a change improved repeatability.

For example, illustrative volumes of 0.008 and 0.010 mm³ produce a relative difference of 22.2% using the formula above. That figure quantifies the mismatch; it doesn’t define an acceptable production limit. Keep the absolute volumes and measurement settings with the percentage so later trials can be compared consistently.

Trace recurring defects to a cause

Look at where the defects occur. Repeated failures at one reference designator call for a different line of inquiry from scattered failures across identical components. A pattern at one panel edge may point toward support, printing or heating conditions. A change that coincides with a new component reel, paste lot or stencil revision provides another testable lead.

  1. Preserve the evidence. Record the lifted end, board orientation, reference designator and lot before rework.
  2. Check the pre-reflow state. Compare paired solder-paste inspection measurements and placement images.
  3. Review the design. Inspect pad geometry, copper connections, vias and the component drawing.
  4. Check materials and heat. Review paste handling, termination condition and measured profile.
  5. Run a controlled trial. Change one identified factor and record the other settings, sample size and defect count.
  6. Inspect the whole board. Check for bridging, insufficient solder and other defects alongside any reduction in tombstones.

Automated optical inspection can flag component position and visible joint abnormalities. Microscope review helps confirm the physical condition. Electrical testing establishes circuit behavior, but doesn’t show whether printing, placement or heating caused the defect. The investigation needs these records together; neither one image nor a passing functional test identifies the root cause.

Tell tombstoning apart from other solder defects

ConditionPhysical distinctionLikely electrical concernInspection focus
TombstoningOne end of a chip lifts from its intended padOpen connection or loss of the component’s functionOrientation and attachment at both terminations
Solder bridgingSolder joins conductors that should remain separateUnwanted short circuitClearance and continuity between adjacent nets
Poorly formed or cold jointA joint fails to form properly; the component need not stand uprightOpen, high-resistance or unstable connectionWetting, joint integrity and process history
BillboardingA chip rolls onto its narrow sideDepends on termination contact and assembly requirementsBody orientation and specified mounting condition
Solder grapingFine solder particles remain uncoalesced around a depositJoint-quality concern distinct from chip rotationParticle coalescence and the surrounding joint

A solder bridge is not another name for a tombstone, though multiple defect types can appear on one board. A dull lead-free joint alone doesn’t establish cold soldering, and a chip standing upright doesn’t prove that the reflow peak was too low.

Rework the component and verify the repair

Use an approved rework method suited to the component and PCB. Isolate the board from power, follow ESD controls and inspect the chip, pads and nearby parts. A cracked body, damaged termination or lifted pad needs assessment; pressing the component flat doesn’t resolve that damage.

For a reusable component, rework must release it without force against solid solder, restore a suitable solder amount and position both ends correctly. Replace the part if its integrity is uncertain or assembly requirements call for it. Heating tools, preheat and exposure limits depend on the package and board.

After rework, inspect both joints and check the relevant electrical function. Record the location and any required cleaning or retest. A successful repair addresses that board; it doesn’t demonstrate that the production cause has been removed. Verify the next build with a corrective-action trial.

Use process records to close the loop

PCB assembly can be reviewed across design for assembly, stencil fabrication, paste printing, 3D solder-paste inspection, placement, reflow, optical inspection and functional testing. Each stage supplies different evidence: paste inspection checks the deposits, placement review checks the chip before melting, and post-reflow inspection checks the completed attachment.

For example, the available equipment includes a SINIC-TEK S8080 3D solder-paste inspection system and a SINIC-TEK A510DL 3D optical inspection system. Equipment readings need to be judged against the project’s component requirements and agreed acceptance criteria.

For a board review, useful records include PCB files, a bill of materials with exact manufacturer part numbers, placement data and the assembly drawing. For an existing defect, add clear photographs, affected reference designators, failure counts and available printing or profile data. Those details help focus the investigation on a measurable difference between the two ends of the component.

Common questions

Can a board with a tombstoned capacitor still power on?

Yes. A disconnected decoupling capacitor may not prevent startup, though the circuit has lost part of its intended filtering. Powering on alone doesn’t establish that the assembly meets its requirements.

Does lead-free solder always cause more tombstoning?

No. The alloy’s melting behavior, paste formulation and imbalance between the two joints all matter. Treat each alloy and process as a specific combination; changing alloy won’t necessarily resolve a footprint problem.

Should every tombstone be blamed on the assembly operator?

No. Design geometry, component condition, paste printing, placement and thermal history can all contribute. Preserve the evidence and identify the process contribution before assigning corrective action.

The useful next step is a paired comparison: measure both pads, deposits and placement, then relate those findings to the local thermal history. That turns a visible defect into a testable process question and gives each corrective change a result to verify.

Source: li · www.bestpcbs.com

From the same set