Laser troubleshoot

Laser System Troubleshooting

Operates under the rules in ../../INSTRUCTIONS.md (reference-first, reference-only, interactive mode).

When to use this skill

When system performance has dropped (cut quality degraded, mark contrast lowered, weld depth inadequate, ablation rate down) and the operator doesn't yet know what's at fault. This skill enforces the correct triage order so alignment isn't pulled apart unnecessarily for a power or contamination problem.

References to load at start

• references/scanner_alignment.md §11 — primary triage order
• references/laser_optics_cleaning.md — Step 2 of the triage
• references/measuring-laser-power.md — Step 1 of the triage
• references/scanner_optimization.md §1, §6, §7 — Step 3 (focal plane) and laser-output sanity

The triage order is non-negotiable

Per scanner_alignment.md §11:

1. Power at the laser output
2. Optic cleanliness
3. Focus-plane parallelism
4. Then alignment checks

⚠️ Don't skip ahead. Most "alignment" complaints are actually one of Steps 1–3. Fixing them in order is faster than guessing.

Inputs to gather first

Ask the operator:

1. What is the symptom — quantitatively? (e.g., "cut depth dropped from 80 µm to 50 µm at the same recipe", "kerf is wider than yesterday".)
2. What changed recently? Laser swap, optic change, recipe update, environment change?
3. When did it start? Abrupt or gradual?
4. Any safety concerns or alarms that came up?

If the answer to #2 is "we just installed a new laser", the path is laser-replacement workflow — go to Step 4 directly with that branch in mind. Otherwise, run the full triage.

→ Pause: confirm inputs and triage path before starting Step 1.

---

Workflow — 4 ordered steps

Step 1 — Verify laser output power

Per scanner_alignment.md §11 + measuring-laser-power.md.

Measure directly at the laser output, before any downstream optic. This isolates the source from the rest of the system.

Two options:

• Hand off to laser-power-measurement for a fully guided 7-step routine.
• Do a quick check inline if the operator has a recent baseline reading.

Outcome paths:

Result	Next
Power is in spec	Go to Step 2
Power is low	Stop the alignment hunt. Source is the issue — engage laser vendor support, check chiller, fiber, pump diodes (laser-type-specific).
Power readings inconsistent	Repeat 3× to rule out warm-up or detector issue, then decide.

→ Pause: confirm Step 1 result before moving on.

Step 2 — Check every transmissive optic in the beam path

Per scanner_alignment.md §11 + laser_optics_cleaning.md.

Walk the operator through each optic, in order from source to work surface:

1. Beam expander input window.
2. Beam expander internal optics (if accessible).
3. Beam expander output window.
4. Scanner input window (if any).
5. Galvo mirrors.
6. Focus lens (F-θ or basic).
7. Any protective window above the work surface.

For each: visual inspection at 45°–85° per laser_optics_cleaning.md §2.

If contamination is found, hand off to laser-optics-cleaning for the cleaning workflow. Return here when complete.

⚠️ If power at the laser output (Step 1) was in spec but power at the work surface is low, the loss is in the optic chain. Find it before continuing. Don't proceed to Step 3 with a known optic loss unfixed.

→ Pause: confirm each optic checked. Confirm any cleaning was done and re-verified. Confirm power at the work surface is now in spec.

Step 3 — Verify focal plane is parallel to the work surface

Per scanner_alignment.md §11 + scanner_optimization.md §1.

Quick test:

1. Mark a small pattern at multiple points across the FoV (corners + center).
2. Compare mark width / quality at center vs. edges.
3. Equal quality across FoV → focal plane parallel to work surface.
4. Edge variation that the center doesn't show → not parallel.

If not parallel, possible causes:

• Work surface mechanically tilted.
• Z-stage motion axis non-parallel to focus-lens normal.
• Wrong working distance — could be a beam-expander divergence problem (per scanner_alignment.md §9 diagnostic).

→ Pause: confirm Step 3 result before moving to Step 4.

Step 4 — Now alignment

Only reached if Steps 1–3 all confirm OK.

Per scanner_alignment.md §11:

Situation	Procedure
Laser was just swapped	Laser-replacement procedure (per §11 — remove BEx and focus lens, align source to scanner aperture / focal plane, then reinstall)
First-time installation	First-time installation procedure — requires metrology for scanner perpendicularity / centering relative to work surface
Drift from working baseline	Generic alignment iteration — two-target principle (scanner_alignment.md §5)

Reality check (per scanner_alignment.md §11): Most production systems lack the metrology and access for a textbook alignment. If tool-specific work instructions exist, follow those. The triage order above is universal; the alignment procedure beyond is system-specific.

→ Pause: confirm operator wants to proceed to alignment, or whether to bring in service / vendor support.

---

Output

A triage report (markdown):

• Symptom + quantification + when it started
• What changed recently
• Step 1 (power): result, action taken
• Step 2 (optics): each optic checked, any cleaning done
• Step 3 (focus): parallelism verdict
• Step 4 (alignment): if reached, what was adjusted
• Where the issue was found
• Final verification (re-measure or re-cut sample)

Save to outputs/laser-troubleshoot/<YYYY-MM-DD>-<machine>.md.

---

Cross-skill links

• Step 1 → laser-power-measurement
• Step 2 → laser-optics-cleaning
• If a specific defect is observed (corner burn-in, recast, charred edges) instead of a generic performance drop, branch to laser-defect-diagnose first.
• If position-specific defects (line starts/ends, corners) → laser-delay-tuner directly.