Fae Intelligence Fae Intelligence
← All sample reports · About Run your own investigation — free

CAPA Record

CAPA Reference / Source

Problem / Nonconformance Description

QFN-48 packages on Oven #1913 are exhibiting cold solder joints at 4.2% defect rate (84 defects / 2,000 joints inspected) beginning June 2, 2026, compared to 0.3% (6 defects / 2,000 joints combined) on Ovens #1911 and #1912 running the same paste lot, boards, and process concurrently — a 14x elevation isolated to one oven.

Failure location: Ground-pad thermal via array under QFN-48 packages — highest thermal-mass joint location on the IoT controller PCBA.

Applicable standards (rigor model): IPC-A-610 Class 2, J-STD-001 — as stated by the user; this is the quality rigor frame in use, not a claim of regulatory applicability beyond what the user confirmed.

Product affected: Industrial IoT controller PCBA (single product line, confirmed by Maria Sanchez's board PN traceability).

Customer exposure: 127 boards produced June 2–4, 2026 on Oven #1913 under AOI + visual inspection only (X-ray not applied during that window). All 127 boards identified in staging; shipment hold placed on Order #MIC-26-0487 (Midwest Industrial Controls) at 1:47 PM June 10, 2026 — no defective product has left the facility. 100% X-ray of all 127 boards initiated by Dan Kim, results pending by end of shift June 10, 2026.

Decisive Contrasts (IS / IS-NOT)

DimensionISIS NOT
What (defect)Cold solder joints, QFN-48 thermal via array (ground pad)Other joint types — TBD, pending Dan Kim's component-type breakdown
Where (equipment)Oven #1913Ovens #1911, #1912 — same paste, same boards, 0.3% defect rate
Where (board location)QFN-48 ground-pad via arrayOther board locations — TBD, pending Dan Kim's X-ray distribution
When (onset)June 2, 2026 onwardBefore June 2, 2026; before Zone 7 drift and new paste lot coincided
How many (extent)4.2% (84/2,000 joints) on #1913≤0.3% on #1911/#1912 running identical paste and boards
Who / which productIoT controller PCBA (single product line, confirmed by Maria)Other board designs — none identified in this run
Which materialPaste lot PST-26-0339 present on all three ovens (Ag 2.8%, liquidus 219°C)Paste lot alone as sufficient cause — #1911/#1912 ran same lot at 0.3%

Decisive contrast 1: Paste lot PST-26-0339 was distributed nearly evenly across all three ovens (~500 boards each, per Maria Sanchez's traceability pull) yet only #1913 shows elevated defects. This single contrast is what shifts the investigation from a paste-lot theory to an oven-specific theory.

Decisive contrast 2: The defect is isolated to one equipment unit while board design, stencil, placement, and components are confirmed identical across all three ovens. This points to a machine-specific variable, not a process-design or material-only variable.

Decisive contrast 3: Onset (June 2) trails the Zone 7 drift onset (May 29/30) by 2–3 days but trails the paste lot introduction (June 1) by only 1 day. This timing pattern is what drives the multi-factor interaction hypothesis rather than a single standalone cause.

6M Fishbone (breadth)

Machine

Material

Method

Man (People)

Measurement

Mother Nature (Environment)

5-Why Chain

Chain A — Machine (Primary)

→ Insufficient thermal energy delivered to QFN ground-pad via array during reflow TAL window

Evidence: 84/2,000 joints defective on #1913 vs. 6/2,000 combined on #1911/#1912 — Dan Kim's X-ray data

→ Zone 7 on #1913 running -4°C to -6°C below setpoint, reducing heat delivery to the highest thermal-mass joint location on the board

Evidence: Raj Patel's thermocouple trend logs — drift onset May 29/30, 2026 at 11:47 PM; persistent through June 10, 2026

→ Hardware failure (heater element, thermocouple, or control card) — spike-and-degrade pattern is consistent with hardware degradation rather than a setpoint change

Evidence: thermocouple logs show intermittent spikes to -3°C beginning May 29, stabilizing at persistent -4°C to -6°C from May 30; specific failed component TBD pending Raj's diagnosis

→ No automated alarm triggered or escalated when drift began; KIC gold-board verification cadence (90-day) did not catch inter-verification drift; product-specific profiling not in protocol

Evidence: Raj noticed drift manually "week of June 1" rather than through an automated alert; last KIC verification April 28, 2026 passed; alarm configuration status TBD

→ Oven monitoring protocol relies on periodic gold-board KIC verification rather than continuous zone-level deviation alerting; gold-board fixture does not replicate the thermal behavior of a populated high-thermal-mass joint; product-specific profiling not required by current SOP

Evidence: SOP design-vs-execution status TBD pending control plan review; alarm configuration TBD pending Raj's diagnosis

Chain B — Material (Contributing)

→ PST-26-0339 (introduced June 1) has elevated liquidus (219°C vs. 217°C on prior lot), reducing thermal margin at the via array

Evidence: PST-26-0339 supplier COA — Ag 2.8% vs. 3.0% standard, liquidus 219°C; Maria's distribution data — all three ovens ran the lot at roughly equal volume

→ Incoming inspection either did not check COA Ag content against internal acceptance limits, or internal limits do not adequately constrain paste lots for the facility's thermal conditions

Evidence: incoming QC procedure and acceptance criteria TBD pending Maria's review

→ Either internal limit matches J-STD-006 range (2.5–4.0%) rather than a tighter ≥3.0% process requirement, or the COA was not reviewed against internal limits before release

Evidence: control plan / SOP review pending — Anna

→ Acceptance criteria may be set to supplier specification rather than process-capability-based limits accounting for worst-case oven thermal margin

Evidence: TBD pending materials SOP review

→ Incoming acceptance criteria for solder paste were not developed with explicit reference to oven thermal margin at high-thermal-mass locations — a latent vulnerability exposed when oven performance degrades

Evidence: TBD pending SOP and control plan review

Root Cause(s)

RC-A — Zone monitoring and oven verification protocol gap (Primary)

Mechanism: Zone 7 hardware degradation → sustained -4°C to -6°C undershoot → insufficient thermal energy at the QFN ground-pad via array during TAL → cold solder joints.

Verdict: PROBABLE. Advances to CONFIRMED when Raj Patel's product-specific KIC profile shows the QFN via-array on #1913 falls below 219°C (or achieves inadequate TAL) while the control oven (#1911 or #1912) holds above 219°C on the same board under the same profile.

RC-B — Paste incoming acceptance criteria gap (Contributing)

Mechanism: PST-26-0339 released without evaluation against process thermal margin → elevated liquidus (219°C) narrows the tolerance window at the via array → combined with RC-A's undershoot, crosses the cold-joint threshold.

Verdict: PROBABLE. Advances to CONFIRMED when Maria Sanchez confirms either (a) an internal Ag limit ≥3.0% existed and was not applied to PST-26-0339, or (b) no prior SAC305 lots ran below 3.0% Ag, establishing this lot as a novel uncontrolled condition.

Interaction mechanism: Neither condition alone crosses the defect threshold. #1911/#1912 ran the same paste at 0.3% defects (RC-B alone is insufficient). Zone 7 drifted 2–3 days before defects appeared (RC-A alone did not immediately produce defects). Together — degraded oven + narrowed-margin paste — the via-array temperature crossed below the effective liquidus on #1913 specifically.

Ruled-Out Hypotheses

Escape Point

EP-1 — Zone 7 deviation alerting (associated with RC-A)

What should have happened: an automated alert on sustained zone deviation beginning May 29/30, escalated to maintenance.

What actually happened: no alert was generated or escalated; Raj Patel noticed the drift manually approximately the week of June 1 — roughly 3 days after onset.

Verdict: PROBABLE. The absence of a triggered/escalated alert is established by Raj's own account of manual discovery, but whether an alarm exists at all, exists but was misconfigured, or exists and fired without escalation is still TBD pending Raj's alarm-configuration check. This determines whether the gap is a design gap (no alarm capability) or an operational gap (alarm ignored) — advances to CONFIRMED when that check returns.

EP-2 — Incoming paste inspection (associated with RC-B)

What should have happened: Ag content on the PST-26-0339 COA evaluated against internal acceptance criteria before release to the line.

What actually happened: the lot was released to the line June 1; whether Ag content was checked against internal limits at all is currently unknown.

Verdict: LEADING. This is the least-evidenced escape point in the record — it rests entirely on inference (a COA deviation exists, and the lot reached the line) rather than a direct account of the inspection event itself. Advances to PROBABLE or CONFIRMED only after Maria Sanchez's SOP and receiving-log review.

EP-3 — In-process inspection, June 2–4 (independent of root cause)

What should have happened: X-ray or equivalent subsurface inspection capable of detecting cold joints at the QFN via array.

What actually happened: AOI + visual inspection only was in use during June 2–4, per Dan Kim's direct report; this method cannot reliably detect subsurface via-array cold joints. 127 boards were produced and moved to staging under this inspection regime.

Verdict: CONFIRMED. This is directly established by Dan Kim's stated inspection-method timeline (100% X-ray began June 5, the day after the NCR opened; AOI + visual sampling only before that) — this finding does not depend on any pending item and holds regardless of how RC-A/RC-B resolve.

Cause Completeness: Occurrence = RC-A and RC-B above. Detection = EP-1, EP-2, EP-3 above.

Evidence

Corrective Action

Verification actions (not corrective) — required because neither RC-A nor RC-B currently carries a CONFIRMED verdict; these actions confirm or refute a suspected cause and are tracked separately from corrective action:

#ActionAddresses (RC / escape)Owner (named)Target DateVerification Method
CA-1Take Oven #1913 offline; diagnose and repair the Zone 7 fault to restore zone temperature within ±1°C of setpointConfirmed Zone 7 out-of-spec condition (RC-A precursor fact)Raj PatelWithin 48 hours of CAPA issue (by June 12, 2026)Post-repair thermocouple reading confirms Zone 7 within ±1°C of setpoint, sustained over one full production shift
CA-2Configure zone temperature deviation alarms on Oven #1913's control system, with escalation to a named recipientEP-1 (confirmed: no alert generated or escalated for the observed deviation)Raj PatelWithin 5 days of CAPA issue (by June 15, 2026)Alarm function test — simulate a zone deviation, confirm alert generates and escalates; threshold settings documented in maintenance record
CA-3Quarantine and disposition remaining PST-26-0339 reels — hold from production, notify supplier, determine return/destroy/conditional-use dispositionConfirmed material deviation from internal 3.0% Ag target (RC-B precursor fact)Maria SanchezImmediate (June 10, 2026)Written quarantine tag on all reels; supplier notification and disposition decision documented before any reel is released
CA-4Establish standing 100% X-ray (or equivalent subsurface inspection) coverage for QFN-48 ground-pad via array joints, replacing AOI-only coverage for this failure modeEP-3 (confirmed detection gap — independent of RC verdicts)Dan Kim (execution); Anna (procedure owner)Within 5 business days (by June 17, 2026)Updated inspection procedure documented and approved; inspection records show 100% X-ray coverage on all subsequent QFN-48 production
CA-5Update oven qualification SOP to require product-specific thermal profiling (minimum: QFN ground-pad via array on actual production PCBA) in addition to gold-board KIC, at initial qualification and post-maintenance re-qualificationRC-A systemic gap / EP-1AnnaSOP draft within 10 business days (by June 24, 2026)SOP reviewed and approved per document control; first implementation executed as part of CA-1 post-repair verification
CA-6Update incoming solder paste acceptance criteria to include a process-capability-based Ag floor (candidate: ≥3.0%) referenced explicitly to worst-case oven thermal margin at high-thermal-mass joint locations; update receiving procedure to require COA review against these criteria before releaseRC-B systemic gap / EP-2Maria Sanchez (criteria owner); Anna (process validation)SOP draft within 10 business days (by June 24, 2026); implemented and trained within 15 business days (by July 1, 2026)SOP reviewed and approved per document control; incoming inspection records for the next 3 paste lot receipts demonstrate COA review against the updated criteria

Preventive Action

Scope: the same failure mode — an undetected zone-temperature drift combined with a marginal incoming material — could recur on Ovens #1911 or #1912, or on any other reflow equipment in the facility running high-thermal-mass assemblies.

#Preventive ActionScopeOwnerTarget DateVerification Method
PA-1Assess zone-deviation alarm capability on Ovens #1911 and #1912; if same make/model/control system as #1913, configure identical alarm thresholds concurrently with CA-2; if different, conduct individual capability assessment#1911, #1912, and any other reflow ovens in the facilityRaj PatelWithin 5 days of CAPA issue (by June 15, 2026)Alarm function test on each oven; threshold settings documented per unit in maintenance records
PA-2Conduct a product-specific KIC profile baseline on Ovens #1911 and #1912 using the protocol established in CA-5, to document current via-array performance as a forward baseline#1911, #1912Raj PatelWithin 10 business days (by June 24, 2026)Profile results documented and retained; any underperformance found triggers immediate maintenance review
PA-3Apply the updated paste acceptance criteria (CA-6) to all future SAC305 receipts across all product lines and ovens, not limited to this product or Oven #1913All SAC305 paste procurement, facility-wideMaria SanchezConcurrent with CA-6 (by July 1, 2026)Incoming inspection records for all post-implementation paste receipts show COA review against the updated criteria
PA-4Review incoming acceptance criteria for other thermally sensitive materials (flux, other solder alloys) to determine whether similar process-capability-based limits are neededAll thermally sensitive incoming materialsMaria SanchezWithin 20 business days (by July 8, 2026)Review documented; any resulting SOP updates completed and approved per document control

Effectiveness Verification

Metric: Cold solder joint defect rate on QFN-48 packages from Oven #1913, measured by 100% X-ray inspection.

Baseline: 0.3% (historical #1913 rate pre-June 2, 2026; concurrent #1911/#1912 rate during the defect window).

Acceptance threshold: ≤0.5% cold-joint defect rate across the full post-repair sample.

Minimum sample (as stated by the CAPA owner): ≥2,000 QFN-48 joints inspected by 100% X-ray on #1913 post-repair AND ≥10 production days — whichever takes longer to satisfy.

Early warning trigger: any single production day exceeding 1.0% — pause production on #1913 and investigate before continuing the verification run.

Secondary gate: zero customer complaints on product shipped from #1913 during the verification window.

Check schedule:

CheckpointTimingAction
30-day check~July 10, 2026Dan Kim reports cumulative defect rate to Anna; continue if no early-warning trigger and rate tracking ≤0.5%
60-day check~August 10, 2026If minimum sample (2,000 joints, 10 days) is met and rate ≤0.5%, Anna may declare effectiveness verified
90-day check~September 10, 2026Final effectiveness confirmation if not already declared

Reporting owner: Dan Kim — daily defect-rate tracking, weekly summary to Anna.

Closure authority: Anna — declares effectiveness verified or triggers secondary investigation.

Closure Criteria

This CAPA may be closed when ALL of the following are true:

1. Zone 7 hardware fault on #1913 diagnosed, repaired, and documented (Raj Patel)

2. Product-specific KIC profile on #1913 post-repair confirms via-array achieves ≥219°C with adequate TAL, on file

3. Zone-deviation alarms configured and function-tested on #1913 (and #1911/#1912 per PA-1)

4. Oven qualification SOP updated to require product-specific profiling — approved per document control

5. Incoming paste acceptance criteria updated with a process-capability-based Ag floor — approved per document control, applied to ≥3 subsequent paste lot receipts

6. PST-26-0339 disposition complete — quarantine resolved, supplier notification documented

7. Effectiveness verification criteria met: ≥2,000 joints, ≥10 production days, ≤0.5% defect rate, zero customer complaints during the verification window

8. All items in Open Items & Gap Impact below are resolved or formally accepted as residual risk by the Quality Manager

9. Quality Manager sign-off recorded

Open Items & Gap Impact

#GapWhy OpenImpact
1Product-specific KIC profile results (Raj Patel)Pending; expected end of shift June 10 or early June 11Decisive for RC-A. Confirms or refutes the via-array thermal mechanism; if via-array temps prove equivalent across ovens, RC-A's primacy must be reexamined
2Zone 7 drift root-cause diagnosis (Raj Patel)Component not yet identifiedDetermines CA-1 specificity; if cause is a setpoint/config change rather than hardware, the Man/Method branch reopens
3Alarm configuration status on #1913 (Raj Patel)Whether alarms exist, thresholds, and whether one fired May 29/30 is unknownDetermines whether EP-1 is a design gap or an operational (escalation) gap — changes CA-2 scope and accountability
4Oven make/model/control system across #1911, #1912, #1913 (Raj Patel)Not yet confirmedDetermines PA-1 scope and timeline; does not affect RC-A verdict
5Incoming QC procedure / Ag acceptance criteria (Maria Sanchez)SOP not yet reviewedDecisive for RC-B corrective action type — execution failure vs. criteria design gap; changes CA-6's specific content
6SAC305 paste lot history, 12 months (Maria Sanchez)Not yet pulledDecisive for RC-B weight — novel condition vs. latent vulnerability exposed by the oven fault
7Prior paste lot comparative profile run (Raj Patel + Maria Sanchez)Requires identifying and locating in-date prior-lot inventoryHigh-value: would convert the RC-A/RC-B interaction from inference to direct observation if performed
8SPI height-map capability and data re: squeegee WO-26-0610 (Dan Kim / SPI owner)Not yet checkedIf height maps show a via-array deposition change post-squeegee-install, reopens the Method branch as a third contributing factor
9QFN-48 component lot / moisture / bake status (Maria Sanchez)Not yet reviewedIf a new lot with inadequate bake coincided with June 1, reframes the interaction as three-factor rather than two
10PST-26-0339 receiving/inspection date (Maria Sanchez)Not yet pulledEstablishes exact EP-2 timeline; does not change the EP-2 finding but sharpens the narrative
11Dan Kim's 100% X-ray results on the 127 held boardsExpected end of shift June 10Determines the content of customer notification to Midwest Industrial Controls — confirmed defects vs. precautionary hold cleared
12Dan Kim's component-type defect breakdown from existing X-ray dataNot yet pulledRefines whether the failure is QFN-48-specific (thermal-mass-driven) or broader (systemic undershoot); affects rework scope, not RC verdict
13Physical oven floor layout (Anna / Linda Chen)Not yet confirmedLow-probability environmental variable; needed to formally close the Mother Nature branch
14Reflow profile comparison across ovens — identical setpoints or legacy difference (Raj Patel)Not yet pulledIf #1913 was configured with less margin by design, adds a Method finding and a profile-standardization action not currently in this record