8D Problem Solving Report

Document Reference: NCR#26-0244

Organization: Beacon Circuits

Report Date: June 2026

Cycle: 1 (open — D6 verification pending)

8D Lead: Anna Novak, SMT Process Engineer

Regulatory / Quality Frame: ISO 9001, IPC J-STD-001, IPC-A-610 Class 2

D1 — Team

D2 — Problem Description

Problem Statement:

> QFN-48 thermal-pad joints on Reflow Oven #1913 show cold joints and insufficient wetting at 4.2% (84/2,000 inspected) against a <0.5% target, first observed in June 2 output — Ovens #1911 and #1912 running the same assembly with the same paste and print process show 0.3%.

IS / IS NOT Analysis:

Evidence referenced:

• Dan Kim 100% X-ray inspection results — #1913 suspect-window output (84/2,000 defective)
• Oven thermocouple history log — #1913 Zone 7 deviation (Raj Patel observation, week of June 1)
• KIC Profile Verification Record — #1913, April 28, 2026 (doc ID TBD — to be attached to NCR#26-0244)
• NCR#26-0244 — opened June 5
• Traceability pull — last 8 days of paste lots, board lots, oven profile changes

D3 — Interim Containment Action (ICA)

ICA Taken:

1. New-build WIP halted on #1913 for this part number; builds diverted to #1911 and #1912

2. 100% X-ray inspection (Dan Kim) of all #1913 QFN output in suspect window; confirmed defective joints being reworked

3. NCR#26-0244 opened June 5; traceability pull executed on last 8 days of paste lots, board lots, and oven profile changes

4. Customer complaint queue reviewed — no complaints received as of NCR opening

Implemented By: Anna Novak (lead), Dan Kim (X-ray execution), NCR opened June 5

Date: Containment initiated June 5; 100% X-ray screen start date TBD — Anna Novak to confirm with Dan Kim

ICA Verified Effective: Partial. Production halt and diversion to #1911/#1912 are in place and functioning. 100% X-ray retrospective screen is in progress; rework of confirmed defects is ongoing. Two open containment actions remain unverified:

• OCA-1: Linda Chen — reconcile ship logs vs. #1913 build dates June 2–5. Due EOD June 11. If suspect product confirmed shipped → escalate to customer notification per Beacon Circuits procedure. Until resolved, it is unknown whether suspect product reached the customer.
• OCA-2: Linda Chen — confirm whether #1913 is currently running other part numbers. If yes → assess expanded hold pending root cause determination.

Evidence referenced:

• NCR#26-0244
• Dan Kim 100% X-ray records — #1913 suspect-window output
• Customer complaint queue review (no complaints received)
• Traceability pull — paste lots, board lots, oven profile changes (last 8 days)

D4 — Root Cause Analysis

Investigation Method: 6M Fishbone + 5-Why, building on the IS/IS NOT analysis in D2

6M Fishbone (breadth):

Machine

• Zone 7 thermal deficit: #1913 Zone 7 reading 5°C below programmed setpoint per oven thermocouple history log — CANDIDATE (RC-1). Hardware mechanism TBD pending Raj Patel's inspection (candidates: heater element degradation, zone thermocouple drift, controller issue).
• #1911/#1912 all zones within ±2°C — oven-specific contrast confirmed, supports RC-1.

Method

• Profile verification gap: KIC profiler run on gold-board thermocouple fixture — does not directly measure QFN thermal pad temperature; highest thermal-mass location is a blind spot in verification method — CANDIDATE, contributes to RC-2.
• No formal thermal margin analysis exists for zone deviation sensitivity at the thermal pad — CANDIDATE, contributes to RC-2.
• New squeegee installed May 30 (WO-26-0610): RULED OUT — SPI confirms paste volume within ±10% of target on boards to all three ovens; print-side cause would not be oven-specific.

Material

• PST-26-0339 Ag 2.8%: within J-STD-006 spec (2.5–4.0% for SAC305); low against internal 3.0% standard but not a nonconformance — RULED OUT as standalone cause (same lot on #1911/#1912 at 0.3%).
• PST-26-0339 liquidus 219°C vs. 217°C standard: CANDIDATE — retained as suspected aggravating factor (H2), not independently sufficient.
• Full COA #SAC-0339 parameter review (flux activity, metal load, particle size, viscosity): TBD — Maria Sanchez to complete.

Measurement

• Zone 7 thermocouple calibration status: TBD — Raj Patel to pull calibration record. Product evidence (X-ray-confirmed cold joints at thermal pad on #1913 only) argues that thermal deficit is real, not a sensor artifact — but independent temperature verification has not yet been performed.
• X-ray detection method (Dan Kim): probed further — no measurement system concerns identified.

Man (People)

• Operator/shift difference: RULED OUT — defect pattern tracks Reflow Oven #1913, not a specific operator or shift. Three ovens run same programmed recipe; no dedicated per-oven operator; no new hires or training gaps flagged. Linda Chen to confirm shift coverage for the record.

Environment (Mother Nature)

• Ambient conditions at #1913 location: RULED OUT — all three ovens in same climate-controlled SMT area; no HVAC anomalies or dock-door effects reported.
• Nitrogen purity: RULED OUT — checked, no deviation found.
• Stencil wear: RULED OUT — checked, no issues found.
• Process changes in suspect window: RULED OUT — none identified.

5-Why Analysis:

Chain 1 — Why did it happen (RC-1):

Root Cause Statement:

> RC-1 — Zone 7 Thermal Deficit (Hardware): Reflow Oven #1913 Zone 7 is delivering insufficient heat — 5°C below programmed setpoint — causing the QFN-48 thermal via array (highest thermal-mass location on board) to receive inadequate time above liquidus, producing cold joints and insufficient wetting. Hardware mechanism (heater element, thermocouple, or controller) — TBD pending Raj Patel's inspection.

> Verdict: LEADING — strongest IS/IS NOT support; consistent physical mechanism; oven-specific contrast confirmed. Advances to PROBABLE when zone 7 position in recipe confirmed and hardware inspected; advances to CONFIRMED when production board profile shows QFN thermal pad TAL below spec on #1913 under current oven condition.

> RC-2 — Absence of Oven Zone Monitoring Requirement (Systemic): The process control plan has never designated routine monitoring of oven zone thermocouple history logs as a required activity. No alarm thresholds, no review owner, no review interval exist for zone deviation detection. This gap allowed Zone 7's thermal deficit to develop and persist undetected until it caused product failures.

> Verdict: PROBABLE — finding stands independent of RC-1 hardware diagnosis; no contradicting evidence. Advances to CONFIRMED when control plan pull confirms the monitoring requirement was never defined.

Ruled-Out Hypotheses:

Escape Point:

> RC-1 Escape Point — Zone 7 thermocouple history log (unmonitored): The Zone 7 thermal deficit was recorded continuously in #1913's own oven data from drift onset. This signal should have triggered detection at or near the onset of the drift — well before product was affected. It did not, because no process requirement exists for reviewing this data (RC-2). The periodic KIC verification is not an effective backstop: the ~5-week verification interval allows drift to develop undetected, and the gold-board fixture does not measure QFN thermal pad temperature — the location with least thermal margin.

> Escape Point Verdict: PROBABLE — the unmonitored zone log as the missed detection opportunity is directly supported by Raj's chance-observation finding. Confirmed when control plan pull shows zone log monitoring was never a defined requirement.

Escape Surfaces:

• Occurrence: Zone 7 delivered insufficient heat to the QFN-48 thermal via array — hardware mechanism TBD pending diagnosis (RC-1).
• Detection at source: The oven's own zone thermocouple history log recorded the deviation continuously, but no in-process monitoring, alarm, or scheduled log review existed to catch it. The periodic KIC verification (last April 28) could not serve as an effective backstop — the ~5-week interval allowed drift to develop undetected, and the gold-board fixture does not measure the QFN thermal pad, the location with least thermal margin.
• Incoming acceptance: Not applicable, because the escaping condition (Zone 7 thermal deficit) is an internal equipment condition, not a purchased-material characteristic. Incoming inspection acceptance criteria for PST-26-0339 (COA review against J-STD-006) are not implicated in this escape — the paste lot meets its acceptance specification.

Cause Completeness:

• Systemic: The process control plan designates periodic KIC profile verification as the sole thermal control for reflow ovens; it never defined continuous or periodic zone-level thermal log monitoring as a required control. This systemic gap is what allowed both the occurrence (undetected drift) and the escape (no backstop) — it is the same finding as RC-2 stated above.
• Organizational: None identified, because no staffing, workload, competency, or scheduling-priority factor has been raised or evidenced in the investigation. Raj Patel discovered the deviation through routine attentiveness, not because of a workload or competency gap in his role.

Root Cause Verified: No — verification experiment designed and accepted by the team; results pending execution (see Open Items).

Evidence referenced:

• Dan Kim 100% X-ray results — #1913 suspect-window output (84/2,000 defective joints)
• Oven thermocouple history log — #1913 Zone 7 deviation (Raj Patel, week of June 1)
• KIC Profile Verification Record — #1913, April 28, 2026 (doc ID TBD — to be attached to NCR#26-0244)
• COA #SAC-0339 — PST-26-0339 (Ag 2.8%, liquidus 219°C cited; full parameter review TBD — Maria Sanchez)
• J-STD-006 — SAC305 silver content specification (2.5–4.0%)
• WO-26-0610 — squeegee replacement May 30; SPI data confirming paste volume within ±10%
• NCR#26-0244 — traceability pull (last 8 days, paste lots / board lots / oven profile changes)
• Process control plan — TBD pull (Anna Novak); expected to determine whether zone log monitoring is currently defined
• Reflow PFMEA — TBD pull (Anna Novak); expected to determine whether zone drift is currently listed as a failure mode

D5 — Permanent Corrective Actions (PCA)

Verification actions (not corrective): Root Cause Verified is No for both RC-1 and RC-2. The following actions exist to CONFIRM the suspected causes, not to correct them, and are not counted as PCAs:

Permanent Corrective Actions:

Note on Actions 3–7: These target RC-2, graded PROBABLE. The underlying finding — no monitoring requirement currently exists — is already directly evidenced by Raj's chance discovery of the deviation and the absence of any documented review trigger; V5/V6 (control plan and PFMEA pulls) will formalize the citation but the systemic gap itself is not in serious dispute. Actions 1–2 target RC-1 and are contingent on V1's diagnosis; the specific repair scope cannot be finalized until then.

PST-26-0339 Paste Lot Disposition:

• #1911/#1912: continue use — lot has run at normal yield (0.3%) throughout suspect window on ovens with zones verified in tolerance
• #1913: hold from use until full return-to-service protocol complete (Actions 1–2, plus return-to-service verification below) — oven-condition hold, not a nonconformance hold
• Supplier: no corrective action, no notification — lot meets J-STD-006; no nonconformance to report
• Conditional long-term action: if Cycle 2 drift-onset data (V4) and a paste-by-oven trial confirm the interaction hypothesis (H2), evaluate internal purchasing standard for tighter Ag band specification — requires trial data before initiating; supplier discussion, not rejection

Return-to-Service Criteria for #1913 (gates release back to unrestricted production; sequenced after PCA #1 repair):

1. Zone 7 verified at setpoint under independent measurement (V2 repeated post-repair) — all zones within ±2°C, matching #1911/#1912

2. Full KIC profile verification passed

3. Production board profile (V3 repeated post-repair) confirms QFN thermal pad peak temperature and TAL in spec

4. Re-qualification build with 100% X-ray on QFN joints — defect rate at baseline (≤0.3%) before unrestricted release. Owners: Dan Kim (X-ray), Linda Chen (production release), Anna Novak (process release decision)

Evidence referenced:

• J-STD-006 — SAC305 specification (basis for paste lot disposition decision)
• IPC-A-610 Class 2 — inspection standard for return-to-service re-qualification build
• KIC Profile Verification Record, #1913, April 28, 2026 — baseline for return-to-service KIC criterion
• WO-26-0610 — squeegee replacement record (ruled out as contributing factor)

D6 — Implement & Verify Effectiveness

No corrective actions have been implemented yet. No verification actions have been executed yet. This section will be populated in Cycle 2.

Effectiveness criteria to be checked in Cycle 2 (not yet met):

• RC-1: QFN defect rate on #1913 returns to ≤0.3% on re-qualification build (100% X-ray); Zone 7 independently verified within ±2°C of setpoint; production board profile confirms QFN pad TAL in spec
• RC-2: Zone deviation alarms active and tested on all three ovens; weekly zone log review records completed and auditable for a minimum period post-implementation; control plan and PFMEA updated and approved per document control

D7 — Prevent Recurrence

Similar processes/products to review:

• Reflow Ovens #1911 and #1912 — same oven model, same zone configuration; the zone-monitoring gap (RC-2) applies equally to them even though they are not currently showing defects. Alarm configuration and log review requirements should be implemented on all three ovens simultaneously, not #1913 alone.
• Any other high thermal-mass component types (e.g., BGAs, power modules, large connector pads) run through all three ovens — the profile verification method blind spot (gold-board fixture not representative of high-mass locations) applies to any assembly with such features.
• Any other convection reflow ovens at Beacon Circuits governed by the same process control plan — if that plan designates KIC verification as the sole thermal control fleet-wide, the monitoring gap potentially exists on all reflow equipment, not just this line.

Documents to update:

Lessons Learned:

1. An oven-specific defect pattern is diagnostic before the fishbone is built. When one oven fails and two run clean on the same material and print process, the investigation should focus on what is different about that oven — not on what changed in the shared process inputs. The shared stencil printer was the key structural fact that broke the paste-lot hypothesis early.

2. The signal existed. The oven was recording the zone deficit continuously. The failure was not that the data was unavailable — it was that no process consumed it. Process control plans that designate periodic verification as the sole thermal control create blind spots between verification intervals.

3. A profile verification that passes does not mean the process has adequate margin at every location on every board. The gold-board KIC fixture verifies oven thermal performance at fixture thermocouple locations — it does not verify joint temperature at high thermal-mass locations on production assemblies.

4. Timeline correlation is a hypothesis, not a conclusion. The paste lot change on June 1 and first defects on June 2 produced a compelling narrative that was not the root cause — but working it through the IS/IS NOT analysis (shared printer, same paste on good ovens) broke it efficiently without discarding the material observation that became H2.

5. PFMEA adequacy is not visible to conformance audits. A missing failure mode in a PFMEA does not trip a finding when audit scope checks conformance to the documented plan rather than adequacy of the plan itself.

D8 — Closure

Closure Criteria (all must be true):

• ☐ Zone 7 hardware mechanism diagnosed (D5 V1 · E15)
• ☐ Zone 7 confirmed real thermal deficit via independent measurement, not oven's own sensor (D5 V2 · E17)
• ☐ Production board profile confirms QFN thermal pad TAL/peak temperature deficit under current oven condition (D5 V3 · E16) — this is the RC-1 CONFIRMED trigger
• ☐ Zone 7 component repaired or replaced per diagnosis (D5 PCA #1)
• ☐ Return-to-service Steps 1–4 passed: independent zone check, KIC verification, production board profile post-repair, re-qualification build ≤0.3% (D5 return-to-service criteria)
• ☐ Zone deviation alarms configured and tested, all three ovens (D5 PCA #3)
• ☐ Weekly zone log review defined and running with auditable records (D5 PCA #4)
• ☐ Process control plan updated and approved (D5 PCA #5 · D5 V5)
• ☐ Reflow PFMEA updated and approved (D5 PCA #6 · D5 V6)
• ☐ Profile verification method updated to include production-board check at QFN pad (D5 PCA #7)

0 of 10 criteria met → closure NOT YET EARNED.

Closure Status: NOT YET EARNED. No D6 effectiveness data exists — no corrective action or verification action has been executed. The critical path is D5 V1 (Zone 7 diagnosis) and D5 V3 (production board profile), which together determine whether RC-1 advances to CONFIRMED and scope the repair. Until that data exists, this record states a grounded, evidence-based root cause analysis at LEADING/PROBABLE grade — it does not state a confirmed, corrected condition.

Investigation Summary:

QFN-48 thermal pad cold joints on Reflow Oven #1913 (4.2%, 84/2,000) were traced to a Zone 7 thermal deficit — 5°C below setpoint — delivering insufficient time above liquidus at the highest thermal-mass location on the board. The deficit went undetected because no process requirement exists for monitoring oven zone thermocouple history logs; the signal was present and unread. Paste lot PST-26-0339 was the initial leading hypothesis but was ruled out as a standalone cause by the shared stencil printer structure; it is retained as a suspected aggravating factor pending drift onset data. Two co-equal root causes have been identified and corrective/verification actions are structured; execution is pending.

Team Recognition:

Anna Novak led the investigation and revised the initial paste-lot hypothesis once the IS/IS NOT evidence did not support it. Dan Kim's 100% X-ray inspection established the quantitative foundation. Raj Patel's observation of the Zone 7 deviation was the pivotal finding. Maria Sanchez and Linda Chen hold open evidence and containment actions. Team recognition is the process owner's step — acknowledge these contributions when closing.

Open Items & Gap Impact

Evidence Index

This is the Cycle 1 8D Problem Solving Report for NCR#26-0244. Root cause analysis is at current evidence grade (RC-1 LEADING, RC-2 PROBABLE) with Root Cause Verified = No. No corrective actions have been implemented; D6 shows 0 of 10 D8 closure criteria met. This report is a draft investigation record for review — it requires verification, approval, and sign-off by the qualified responsible party (Anna Novak, 8D Lead) via the Review & Sign-off step before it is filed or submitted as a completed record.