Skip to main content

1.5 Common Defects & Corrections

Bridging, icicles, skips—what causes them and how to eliminate them (with the smallest reliable change).

Most soldering defects—bridges, icicles, skips—come from a small set of causes: poor flux coverage, wrong top-side heat, unbalanced chip/main wave setup, or sloppy exits. Always check flux and preheat first, then change one setting at a time: chip wave for bridging, dwell/speed for fill, exit technique for icicles. Selective fixes are local (nozzle size, Z, path), while wave fixes are global (wave balance, angle, peel). Design tweaks—like correct hole sizes, thermal reliefs, and thief pads—remove chronic problems entirely. Measure results with photos and counts, save changes in the recipe, and avoid “big guess” adjustments. Small, targeted moves make defects vanish and keep them gone.

1.5.1 Triage first (don’t turn five knobs)

  1. Prep check (always first): flux coverage/dose and top-side temp at wave entry (1.2).
  2. Mechanics: pallet seals/clearances, finger cleanliness/angle (1.4.4), nozzle Z/height (1.3.6).
  3. Process knob: change one—speed/dwell, fountain height, chip↔main balance—then recheck on one panel.



1.5.2 Quick map (symptom → likely cause → smallest fix)

Symptom

Likely cause(s)

First fix (fastest that sticks)

Bridging (between pins/rows)

Flooded flux film; cold entry; chip wave too low/short; parallel long rows; fountain too tall

Trim flux dose; add/extend chip wave dwell +0.3 s; raise top-side 5–10 °C; lower fountain 0.2–0.5 mm; rotate layout vs wave or add thief (13.1/13.4)

Bridging at row end

Lazy exit; no tail

Add end-dwell 0.3–0.5 s then quick kick-out; program a thief tail past last pad; sharpen pallet exit edge

Icicles / spears

Over-dwell with cool exit; low pot; long lead protrusion

Faster exit; +5 °C pot; trim dwell −0.2 s; enforce 0.5–1.5 mm protrusion (12.1)

Skips / non-wet / de-wet

Oxide/aged finish; flux starved; short contact; thermals to planes

Add light extra spray pass; slow belt (longer contact); 4-spoke thermals in next rev; try N₂ on wave

Poor top-side barrel fill

Inadequate preheat/flux in holes; drag too fast; hungry copper/planes

Raise top-side temp 5–10 °C; second light pass; reduce speed 2–3 mm/s or +0.3 s spot-dwell; bump pot +5 °C if still short

Solder balls / spitting

Wet solvent at entry; fountain turbulence; dirty fingers

Dry earlier (same total heat, smoother ramp); lower fountain 0.5–1.0 mm; clean/degunk fingers & rails

Blowholes / voids in barrels

Moist PCB; flux boils in hole; mask dams gas

Bake boards; gentler, longer preheat; relieve mask annulus 0.10–0.15 mm (13.1)

Dull/grainy joints

Cold joint; over-baked flux

+5 °C pot or +0.2 s main dwell; even preheat (no burnt zones); consider N₂

SMT splash / contamination

Overspray; fountain too tall; masking gaps

Tighten spray window; lower fountain; improve pallet seal and add clip-on fences

Selective: random non-fill

Z too high (no engulf); path outruns cup

Touch-off Z-zero; slow to 5–8 mm/s on that segment; add quick-wet + targeted slow second pass

Selective: persistent bridges in 2.00 mm rows

Cup overfills trench; end dwell missing

Smaller non-wettable nozzle; add end-dwell + kick; 5–10° approach angle



1.5.3 Root causes you can design out (next spin)

  • Holes too tight / thin annular rings → undersized barrels never fill. Fix: lead Ø +0.20–0.45 mm (by thickness/copper), annular ≥0.25–0.30 mm, teardrops (1.1).
  • No thermal reliefs to planes → cold joints. Fix: 4 spokes, 0.25–0.40 mm (1.1).
  • Rows parallel to wave → long solder dams. Fix: rotate or add robber pads/pallet thieves (1.1, 1.4).
  • SMT crowding miniwave → splash/bridges. Fix: 3–4 mm keepout or pallet masking (1.1, 1.3).
  • Leads too long → icicles. Fix: plan protrusion 0.5–1.5 mm after solder (1.1).



1.5.4 Knob guide (change the right thing)

  • Looks flooded / bridges early? → Flux & preheat first, then chip wave height/dwell.
  • Looks starved / no top-fill? → Conveyor speed & dwell, then pot +5 °C.
  • Exit looks stringy? → End-dwell + quick lift, small air/N₂ knife; don’t just raise temperature.
  • Selective only on a few pins failing? → Local pass (spot dwell) and smaller nozzle, not global changes.



1.5.5 Mini playbooks (by defect)

A) Bridging (fine rows)

  1. Verify UV flux carpet and top-side in band (1.2).
  2. Raise chip 0.2–0.5 mm or +0.3 s dwell; keep main calm.
  3. Add end-dwell + kick; if still there → lower main 0.2 mm and add thief.
  4. Next rev: robber pad or pallet tail.

B) Poor top-side fill

  1. +5–10 °C top-side via preheat; slow belt or +0.3 s spot-dwell.
  2. Second light pass (drag or selective).
  3. If still low → +5 °C pot; check hole clearance/thermals for next spin.

C) Icicles on tabs/posts

  1. Faster exit; small knife.
  2. +5 °C pot; shorten dwell; confirm lead protrusion.
  3. For huge tabs: mini-dip or local spot dwell, not whole-board heat.



1.5.6 Selective vs wave: what’s different

  • Selective fixes are local: nozzle Ø, Z touch-off, path speed/dwell, fountain height. Prefer two-stage passes (quick wet → slow finish).
  • Wave fixes are global: chip↔main balance, conveyor angle, peel/knife. Don’t try to make main wave do chip’s job.



1.5.7 “Do not” list (saves days)

  • Don’t raise pot temp to fix flux/preheat mistakes—spatter and oxidation follow.
  • Don’t widen AOI limits for bridges; fix chip wave / exit.
  • Don’t drown boards in flux; dry, even films beat puddles.
  • Don’t tweak three settings at once—you lose the cause.
  • Don’t accept “random non-fill” until you’ve checked hole size vs lead.



1.5.8 Tiny DOE when you’re stuck (3× panels)

  • Factor A: Speed (−2 mm/s / nominal / +2 mm/s)
  • Factor B: Fountain (−0.3 mm / nominal / +0.3 mm)
  • Factor C: End-dwell (0 / +0.3 s)
     Measure bridges/row, top-fill %, icicles. Keep the combo that improves the Pareto with the smallest heat.



1.5.9 Inspect what proves the fix

  • Top-side barrel photos on the worst header (same pads each run).
  • Count bridges per row, icicle length, skip rate on 3 panels before/after.
  • Save settings in recipe comments (what you changed and why).



1.5.10 Pocket checklists

Before first article

  • Flux pattern even; dose logged; no overspray
  • Top-side temp at entry in band (per flux family)
  • Chip and main heights set; angle 6–8°; finger path clean
  • Selective: Z-zero taught; nozzle Ø sized to pad +1–2 mm

If bridging

  • Add/extend chip; lower main a touch; end-dwell + kick
  • Thief tail or pallet insert; re-photo same row

If non-fill / skips

  • Slow belt or +0.3 s spot dwell; raise top-side 5–10 °C
  • Second light pass; check hole/thermal design for next rev

If icicles

  • Faster exit; light knife; +5 °C pot only if needed
  • Verify lead protrusion; shorten dwell on tabs



Bottom line: most wave/selective defects are prep + first contact + exit problems. Keep flux thin and even, hit top-side activation, let the chip wave scrub and the main wave finish, then peel cleanly. Change one knob, verify on one panel, and write it down. That’s how defects disappear—and stay gone.


Back to top