123 Main Street, New York, NY 10001

Creepage & Clearance for PCB Isolation (Pollution Degree)

← Back to: Digital Isolators & Isolated Power

This page turns creepage & clearance into an auditable, board-level workflow: lock pollution degree/CTI/altitude inputs, then enforce surface/air spacing with PCB geometry and controlled coating processes. The outcome is a repeatable evidence pack and inspection criteria so separation passes review, production, and field rework with clear X/Y/N acceptance.

H2-1. Overview & Scope: What This Page Covers

Intent

This page focuses only on PCB separation distances (clearance and creepage) and the two real-world drivers that most often break them: pollution (contamination) and altitude (air pressure). It is written to prevent “safety surprises” at review time by making distance decisions repeatable, auditable, and layout-actionable.

Scope Guard

  • Covered: how to size clearance/creepage using inputs (OVC/PD/CTI/altitude/coating), and how to implement them on PCBs using slots, keep-outs, barriers, coatings, and tolerance-aware layout rules.
  • Covered: how to document and verify distances so they are measurable and review-ready (drawings, inspection evidence, worst-case accounting).
  • Not expanded here: insulation class selection (basic/reinforced), impulse/surge withstand, hi-pot and partial discharge test methods. Link out only: Basic/Reinforced · Impulse/Surge · Production Tests

Where clearance/creepage most often fails in the field

  • Condensation / wet pollution: a conductive surface film forms → creepage becomes the first limiting factor. Fix belongs to PD/coating/cleanliness and surface path control.
  • Dust + humidity (mixed contamination): PD effectively “upgrades” over time → creepage margin erodes silently. Fix belongs to CTI/material group, coating coverage, and inspection evidence.
  • High altitude: lower air pressure reduces breakdown margin → clearance is usually the first to fail. Fix belongs to altitude derating and air-gap enforcement.
  • Sharp edges / solder protrusions: local field concentration + effective distance shrink → both clearance and creepage can fail even when “nominal mm” looks correct on the PCB.
  • Assembly tolerance stack-up: placement drift + solder fillet growth + contamination residues reduce the real minimum distance → audit rejects “as-built” geometry.

What this page delivers

  1. A repeatable workflow: inputs → standards lookup → required distances → PCB implementation → verification evidence.
  2. PCB techniques that reliably increase effective creepage/clearance: slots, barriers, keep-outs, coating strategy, and tolerance-aware rules.
  3. Audit-ready documentation approach: how to “measure” and freeze separation distances in drawings, DRC rules, and inspection records.

Data Block · Design Inputs Checklist (fill before any distance sizing)

Input Value (X/Y/N) Evidence to keep (audit-ready)
Working voltage definition X System requirement / schematic annotation
Overvoltage category (OVC) X Product spec / installation environment note
Pollution degree (PD) X Field environment description + maintenance assumption
Altitude requirement X m Deployment spec / geography requirement
Material group / CTI X Laminate datasheet + coating datasheet (if used)
Coating applied? Y/N Process spec: coverage %, thickness target, inspection method

Rule: do not size distances until every input above is locked and traceable. “Unknown” is a design risk and should be treated as “worst-case” until resolved.

Figure · Inputs → Lookup → Layout → Verify (workflow preview)

Inputs OVC / PD / CTI Lookup Standards tables Required C&C X mm / margin Layout Slots / keep-outs Verify Evidence pack Altitude Coating Lock inputs → size distances → implement geometry → verify as-built

The workflow is intentionally linear: if inputs are not frozen, distance outputs cannot be defended. Verification is a required step, not a “nice-to-have.”

H2-2. Definitions: Clearance vs Creepage (and Why They Fail Differently)

Intent

Clearance and creepage are measured on different paths and fail through different mechanisms. Correct sizing starts with counting the right path and matching the path to the dominant stress.

H3 · Clearance (air gap)

  • Definition: the shortest path through air between conductive parts on different potentials.
  • Primary drivers: overvoltage environment (OVC), peak stress definition, and altitude (air pressure).
  • Typical failure pattern: air breakdown is triggered by insufficient margin under reduced pressure, sharp geometry, or underestimated peak stress.

H3 · Creepage (surface path)

  • Definition: the shortest path along an insulating surface between conductive parts.
  • Primary drivers: pollution degree, material group / CTI, surface condition (residues, moisture films), and coating quality (coverage consistency).
  • Typical failure pattern: surface leakage initiates on contamination, then progresses via tracking/carbonization, creating a conductive “highway” that bypasses nominal geometry.

H3 · Counting mistakes that cause audit failures

  1. Measuring creepage as a straight line. Creepage must follow the surface, including around slots, steps, and edges.
  2. Ignoring solder fillets, copper burrs, and sharp corners that reduce the real minimum distance and increase local field concentration.
  3. Letting silkscreen, solder mask bridges, or contamination cross the isolation boundary without a defined keep-out and cleanliness policy.
  4. Using nominal CAD distances as “proof.” Verification must account for worst-case stack-up (placement, process, and residue).

Data Block · Failure Mode Map (first suspect)

Condition Clearance risk Creepage risk Quick first check
High altitude requirement Yes No/Maybe Altitude correction applied to clearance (Y/N)
Condensation / wet pollution Maybe Yes PD/CTI locked + coating/cleanliness evidence exists
Sharp copper edge / solder protrusion Yes Yes Geometry review: radius/keep-out/slot near hot nodes
Flux residue / ionic contamination No Yes Cleanliness policy + process controls are defined (Y/N)
Tolerance stack-up concern Yes Yes Worst-case min distance computed from tolerance chain

Fast memory hook: Altitude → clearance first; Contamination → creepage first.

Figure · Air path (Clearance) vs Surface path (Creepage)

Primary Secondary PCB surface Slot Clearance (air path) Creepage (surface path) Altitude Pollution Count the correct path before selecting any distance numbers

Clearance is the shortest air route between conductive parts. Creepage is the shortest route along the insulating surface—often longer when geometry forces detours (for example, slots).

H2-3. Pollution Degree, Material Group & CTI: The Inputs You Must Lock First

Intent

Creepage sizing is not meaningful until pollution degree (PD) and surface robustness (CTI / material group) are locked. These two inputs define whether the PCB surface is expected to stay insulating or gradually turn into a conductive leakage path.

Gate rule (process)

  • PD must be declared with an explicit environment assumption (contamination type, condensation probability, maintenance policy).
  • CTI / material group must be traceable to a datasheet or supplier declaration for the actual surface (laminate + solder mask + coating).
  • If any item is “unknown,” treat it as worst-case until evidence is provided and recorded.

Engineering interpretation (what PD/CTI actually controls)

H3 · Pollution Degree (PD): how to decide it in real deployments

  • PD is an environment model: it reflects how likely an insulating surface becomes a conductive film due to contamination + moisture.
  • The practical split is condensation risk: repeated temperature cycling and sealed volumes can create moisture films even when air looks “dry.”
  • Maintenance assumptions matter: filtration, periodic cleaning, and enclosure strategy should be written down, otherwise PD cannot be defended.

H3 · CTI / material group: why it changes creepage requirements

  • CTI is a proxy for tracking resistance: higher CTI surfaces resist carbonized “tracks” that turn leakage into a permanent short path.
  • The relevant surface is not only the laminate: solder mask, residues, and coatings jointly define the effective tracking behavior.
  • If CTI evidence is missing, creepage should be sized conservatively and flagged as a risk until the surface system is documented.

H3 · “Pollution upgrade” mechanisms seen in the field

  • Dust + humidity → conductive moisture film → surface leakage rises first, often before any visible damage.
  • Flux / ionic residues → invisible contamination → conductivity appears only under humidity or after thermal cycling.
  • Salt fog / coastal air → electrolyte film → leakage becomes stable and aggressive even at moderate humidity.
  • Oil mist → persistent film → difficult to remove → creepage margin erodes without any geometry change.

Data Block · PD determination questionnaire (10× Yes/No → PD placeholder)

Purpose: capture assumptions and evidence. Output is a PD placeholder to be confirmed against the applicable standard and project requirements.

  1. Condensation expected during temperature cycling (Y/N)?
  2. Enclosure “breathes” moisture in/out (pressure changes, vents) (Y/N)?
  3. Dust accumulation likely (fans, open vents, industrial floor) (Y/N)?
  4. Salt fog / coastal exposure likely (Y/N)?
  5. Oil mist / coolant / lubricant aerosols likely (Y/N)?
  6. Cleaning/maintenance is periodic and specified (Y/N)?
  7. PCB surface is washed/verified for residues in production (Y/N)?
  8. Conformal coating is applied with coverage verification (Y/N)?
  9. Sustained high humidity (long seasonal exposure) expected (Y/N)?
  10. In-service contamination can remain for months without intervention (Y/N)?
Output (placeholder)

PD = X (1 / 2 / 3 / 4) · Assumptions recorded = Y/N · Evidence attached = Y/N

Data Block · Surface system record (laminate + solder mask + coating)

Item Spec / Value Evidence
PCB laminate Part / grade = X Datasheet / supplier declaration
CTI / material group CTI = X CTI report / datasheet reference
Solder mask Type = X Process spec / coverage rule
Coating (if used) Type = X · Applied = Y/N Thickness target / coverage verification
Cleanliness control Policy = X Wash spec / inspection record template

Note: creepage is governed by the surface system. A high-CTI laminate does not automatically guarantee a high-CTI surface if residues or coatings dominate the actual path.

Figure · Pollution evolution: dry dust → wet film → tracking path

Stage 1 Dry dust Stage 2 Wet film Stage 3 Leakage path I-leak Stage 4 Tracking Moisture + contamination turns “insulating surface” into a conductive route

Field failures often originate from a slow transition: a clean surface becomes a wet conductive film, then repeated stress creates a permanent tracking route.

H2-4. Clearance Sizing Workflow: Voltage, Transients, OVC, Altitude

Intent

Clearance is the shortest air path. Correct sizing depends on consistent voltage definitions, correct overvoltage environment (OVC), and explicit altitude derating. This section provides a step-by-step workflow without hardcoding standard table numbers.

Workflow foundations (definitions and decision links)

H3 · Define voltage terms (working vs peak/transient)

  • Working voltage (Vwork): the maximum continuous steady-state value for the insulation barrier.
  • Peak voltage (Vpeak): the highest expected voltage during normal operation, including controlled switching events.
  • Transient definition: record the project-specific peak event category (placeholder), but do not mix it with impulse/surge test waveforms handled elsewhere.

H3 · OVC determines the expected air-gap stress level

  • OVC is a classification of the installation environment and its overvoltage exposure. Higher OVC generally implies higher expected air-gap stress and tighter clearance requirements.
  • OVC should be justified by deployment notes (panel wiring, cable exposure, system boundary). If justification is absent, treat it as a risk and size conservatively.

H3 · Altitude reduces air breakdown margin (clearance must increase)

  • Reduced air pressure at altitude lowers breakdown strength, so the same geometry can fail even if it passes at sea level.
  • Altitude derating must be applied not only to PCB trace-to-trace gaps, but also to the tightest air routes around packages and connectors.

Data Block · Clearance sizing and lookup workflow (placeholders)

Step Input / action Reference Output
1 Lock Vwork / Vpeak definitions (X) Project voltage note (evidence) Inputs frozen
2 Select OVC (X) with deployment justification Installation notes (evidence) OVC locked
3 Lookup base clearance from table IEC 60664-1 table (placeholder) Base clearance = X mm
4 Apply altitude derating (Altitude = X m) Altitude correction method (placeholder) Derated clearance = X’ mm
5 Add margin policy (+Y%) and tolerance reserve Company policy / review note Required clearance = X mm
6 Verify the tightest air route on as-built geometry Inspection record template Pass = Y/N

Keep the table edition, method version, and evidence snapshots together; clearance sizing becomes non-auditable if references drift across revisions.

Figure · Clearance decision tree (OVC + altitude correction)

Define Vwork / Vpeak Select OVC Lookup Base clearance Altitude? Apply Correction Output Clearance = X mm No Yes Margin +Y% policy

The decision tree keeps clearance defensible: definitions first, then OVC selection, then table lookup, then altitude correction, then margin and verification.

H2-5. Creepage Sizing Workflow: PD, CTI, Surface Path, Coating

Intent

Creepage is governed by environment (PD), surface tracking robustness (CTI/material group), and the true shortest surface route. This section provides a strict workflow that separates required creepage (from tables) from effective creepage (from PCB geometry and surface strategy).

Workflow gates (must be locked in this order)

  1. PD locked with assumptions + evidence (Y/N).
  2. CTI / material group locked for the actual surface system (Y/N).
  3. Surface path rule locked (slot/step/mask bridge/edge counting).
  4. Coating claim locked: any “equivalent PD reduction” must follow the applicable standard and review policy.

H3 · Define the “true surface route” (what counts as creepage length)

  • Creepage is the shortest path along the insulating surface between conductive parts. It is not the straight-line spacing.
  • Slots force the route to detour and can increase effective creepage if the slot is continuous and the route cannot “jump” across surfaces.
  • Steps / ledges must be counted by “wrapping” the surface; the route follows the actual contour.
  • Solder mask bridges are a common audit dispute: a project must lock a counting rule and evidence for mask integrity and coverage.
  • Board edges can become the shortest surface route in dense layouts; check edge-to-edge paths explicitly.

Required vs effective creepage (the only pass rule)

Required creepage comes from PD/CTI/Vwork lookup. Effective creepage comes from the shortest realizable surface route on the as-built geometry. Acceptance is Effective ≥ Required, documented with the route and assumptions.

H3 · Coating and “equivalent PD reduction” (only with standard/review alignment)

  • Coating can reduce moisture film formation and contamination attachment, improving surface robustness — but any claim must be aligned with the applicable standard and audit policy.
  • If a project treats coating as a risk reducer, coverage must be verifiable (coverage area, thickness target, pinhole/void checks), otherwise it cannot be used as a sizing basis.
  • Record coating status as a binary, auditable input (Y/N) with evidence references; avoid informal “it should help” assumptions.

H3 · Typical board-level creepage failures (where tracking starts)

  • Sharp corners / narrow necks concentrate surface stress and are frequent starting points for leakage and tracking.
  • Local contamination sources (residue, dust traps, vent paths) create a stable film that shortens the effective surface route.
  • Progression: wet film → leakage current → localized heating → carbonized track → permanent short surface path.

Data Block · Creepage sizing output (Required vs Effective) — placeholders

Field Value Evidence / note
PD X Assumptions + questionnaire
CTI group X Datasheet / declaration
Vwork X Voltage definition note
Required creepage X mm IEC table placeholder + edition
Coating None / Conformal / Potting (Y/N) Process spec + verification
Geometry Flat / Slot / Barrier Route definition locked
Effective creepage X’ mm Measured / computed path
Pass Y/N (Effective ≥ Required) Review record / snapshot

Use this block as the project’s auditable creepage record: required value, effective route, coating status, and evidence references in one place.

Figure · Same Vwork, different PD/CTI → different required creepage (X1/X2/X3)

Input Vwork PD Low CTI High X1 mm PD Mid CTI Mid X2 mm PD High CTI Low X3 mm Geometry Effective X’ Required creepage (X1/X2/X3) comes from PD+CTI; effective creepage (X’) comes from geometry

H2-6. Altitude Derating: Practical Rules for High-Altitude Designs

Intent

Altitude must be converted from a vague concept into a project rule: when derating is required, how it is recorded, and which geometry must be covered (PCB, connector, module, enclosure). This section defines auditable policy blocks and common pitfalls.

H3 · Why altitude matters (clearance risk increases as air pressure drops)

  • Altitude ↑ → air pressure ↓ → air breakdown strength ↓.
  • A geometry that passes at low altitude can fail at high altitude without any PCB changes.
  • Derating must be explicitly applied via a standard-aligned method and recorded as part of sizing evidence.

H3 · Project workflow: how to declare altitude in requirements and reviews

  • Requirements must state Target altitude = X m (deployment maximum, not typical).
  • Design records must include the correction method / multiplier (placeholder) and the standard edition used.
  • Certification/test statements must reflect the declared altitude condition; if altitude is omitted, assumptions become undefined and difficult to audit.

H3 · Common pitfalls: derate PCB clearance but miss the true shortest air route

  • PCB spacing is updated, but connector pin-to-pin air gaps remain the minimum.
  • Slots are added on PCB, but the tightest path is around a package body / module edge.
  • Enclosure brackets or standoffs create an unexpected short air route.
  • Mechanical tolerance eats the minimum gap; altitude derating must include tolerance reserve.

Data Block · Altitude policy (project standard) — placeholders

Target altitude
X m
Clearance multiplier / correction method
(standard placeholder + edition)
Mechanical tolerance reserve
Y mm
Coverage checklist
PCB: Y/N Connector: Y/N Module: Y/N Enclosure: Y/N

This policy card prevents partial fixes: altitude derating must cover the true shortest air route across all physical elements.

Figure · Altitude ↑ → air breakdown margin ↓ → required clearance ↑ (3 levels)

Altitude Low Mid High Air breakdown margin Required clearance Declare target altitude, apply standard-aligned correction, and verify the shortest air route

H2-7. PCB Geometry Techniques: Slots, Barriers, Keep-Outs, Guard Rings

Intent

PCB safety spacing is built by controlling the shortest route, not just “moving pads apart”. This section turns creepage/clearance requirements into geometric techniques and an auditable layout rule set that can be enforced by DRC and verified by the shortest-path snapshot.

Technique map (what each knob controls)

  • Slots: force surface route detours → increases effective creepage.
  • Barriers: block or reroute the shortest path across assemblies.
  • Keep-outs: prevent copper/mask/silkscreen from creating a new shortest route.
  • Guard rings: manage surface leakage route on the isolation side (not an EMI topic).

H3 · Slots: increase creepage by forcing a detour

  • A slot is effective only when it forces the shortest surface route to go around it.
  • Lock edge-to-copper around the slot; a tight copper edge can become a new shortest route and tracking start point.
  • Ensure the slot is continuous across the intended isolation corridor; partial slots may not change the shortest path.
  • Acceptance requires a shortest-route snapshot showing the pre-slot path X and the post-slot path X’.

H3 · Barriers: coordinate PCB and mechanics to avoid a hidden shortest route

  • Barriers reroute the shortest path across assemblies; the barrier surface can also become a creepage surface.
  • Include mechanical tolerance reserve; the minimum gap is the as-built gap, not the nominal CAD gap.
  • Verify coverage at the true minimum points: near connector pins, package edges, and any mounting features.
  • Acceptance: document the minimum air route and surface route with a labeled geometry snapshot.

H3 · Keep-outs: prevent copper/mask/silkscreen from creating a new shortest path

  • Copper keep-out must include traces, pours, vias, test pads, thermal copper, and any copper features that could reduce the minimum spacing.
  • Mask opening keep-out prevents exposed edges from becoming a weak surface region; lock the project’s mask counting rule consistently with creepage sizing policy.
  • Silkscreen keep-out prevents bridge-like markings across the isolation corridor that can accumulate contamination.
  • Acceptance: DRC must flag any layer that crosses the isolation corridor (copper/mask/silkscreen).

H3 · Guard rings: control surface leakage route on the isolation side

  • A guard ring provides a controlled collection route for surface leakage, reducing unpredictable leakage paths across the isolation corridor.
  • Place the guard on the intended side only; it must not cross the isolation gap or violate keep-out rules.
  • Guard is not a substitute for spacing; it must be applied only after required/effective creepage and clearance are satisfied.
  • Acceptance: guard continuity and keep-out compliance are checked by DRC and by geometry snapshot.

Data Block · Layout Rule Set (copy into PCB DRC) — placeholders

Rule item Value Scope
Min clearance rule X Copper-to-copper (air)
Min creepage along surface X Surface shortest route
Copper keep-out width Y Traces/pours/vias/pads
Mask opening keep-out Y2 Solder mask openings
Silkscreen keep-out Y3 Legend across corridor
Slot width Z Manufacturable minimum
Edge-to-copper (slot/board edge) Z2 Prevent new shortest route
Min corner radius / fillet R Reduce sharp tracking starts

The rule set must be enforced on copper, mask, and silkscreen layers to prevent “shortest-path hijacks”.

Figure · Slot changes the creepage route: straight X → detour X’

Before After Copper A Copper B Keep-out X Copper A Copper B Keep-out Slot Z2 Z2 X’ Slot is effective only when it forces the shortest surface route to detour

H2-8. Surface Treatments: Conformal Coating, Potting, and What They Do NOT Fix

Intent

Coating and potting are engineering trade-offs, not “automatic passes”. This section clarifies what surface treatments can improve for creepage robustness, what they cannot replace (especially air clearance), and which process records must exist for production auditability.

H3 · Conformal coating: improves surface robustness only when coverage is consistent

Helps
  • Reduces moisture film and contamination attachment on the surface route.
  • Stabilizes surface condition across humidity swings when process control is strong.
Does NOT fix
  • Does not “erase” air clearance requirements.
  • Does not compensate for a geometry that violates required creepage by design.
Audit hooks
  • Coverage target and verification method must be recorded (Y/N).
  • Pinhole/void inspection must be defined (placeholder).

H3 · Potting/encapsulation: stronger coverage, heavier trade-offs

  • Serviceability: potting reduces repair access and changes field diagnostic strategy.
  • Thermal: potting alters heat paths and can create new stress points; thermal policy must be documented.
  • Voids/bubbles: local voids can become weak spots; inspection/controls must be defined.
  • Aging: shrink/crack over life can re-expose a surface route; lifetime assumptions must be consistent with the safety case.

H3 · Common misjudgments: coating does not remove clearance, thin edges and pinholes are risks

  • Clearance remains an air rule: surface treatment cannot be used as an informal substitute.
  • Edge thinning: coating is often thinner at edges and near tall components; thin regions can become leakage weak points.
  • Pinholes: a single pinhole can create a continuous leakage route under humidity; inspection is mandatory for any safety claim.

Data Block · Production process record (must exist) — placeholders

Coating / potting process
Type: X Thickness: X Coverage: Y% Cure: (placeholder) Rework: Y/N
Inspection / verification
Pinhole/void method: (placeholder) Coverage verification: (placeholder) Witness coupon: Y/N
Cleanliness (before coating)
Ionic contamination: (placeholder) Cleaning process ref: (placeholder)

A surface-treatment claim is auditable only when process and inspection records exist and are linked to production control.

Figure · Uneven coating → leakage route still exists (thin edge + pinhole)

PCB substrate Copper A Copper B Coating Thin edge Pinhole Leakage path Air gap Surface treatments help creepage robustness, but clearance remains an air requirement

H2-9. Component & Assembly Constraints: Packages, Connectors, Tolerances

Intent

Board spacing is only the starting point. Real minimum separation is often defined by package geometry, connector/module pin pitch, and assembly tolerances. This section converts nominal spacing into an auditable worst-case clearance and documents the measurement points.

Shortest-path hijackers (what to check first)

  • Package body / lead root becomes the creepage surface.
  • Connector housing creates a new surface route between terminals.
  • Solder fillet growth reduces air gap (clearance) vs CAD nominal.
  • Placement/tilt shifts the minimum gap toward one side (worst-case stack-up).

H3 · Package constraints: use datasheet package drawings as the geometry source

  • Minimum separation is defined by the as-built geometry, not only copper-to-copper spacing.
  • Datasheet package drawings define body/lead dimensions that can create the true shortest surface route.
  • Document the measurement point pairs (pin-to-pin / pad-to-pad / body-to-body) to make review and inspection reproducible.

H3 · Connectors & modules: terminal pitch and housing surface paths can dominate

  • Connector terminal pitch and housing surface can form a shorter route than the PCB corridor.
  • For isolation modules/transformers, treat the external body surface as part of creepage routing; do not assume PCB-only routing.
  • Promote key connectors/modules to critical characteristics for incoming and outgoing inspection.

H3 · Production reality: placement shift, solder growth, and residues reduce the minimum gap

  • Clearance can be reduced by solder fillet growth and by sharp solder peaks near the isolation corridor.
  • Worst-case stack-up must assume placement shift toward the same side (do not average away the risk).
  • Residues and localized contamination change the surface condition; inspection points must record evidence (photo ID + point pairs).

Data Block · Stack-up of tolerances (worst-case) — placeholders

Term Symbol Value Note
Nominal gap X (placeholder) CAD nominal
Placement tolerance ±a (placeholder) Worst-case toward hazard
Solder fillet growth b (placeholder) Air gap reduction term
Tilt/warp allowance c (placeholder) If applicable
Worst-case clearance X-(a+b+c) (compute) Must meet Xreq + margin

Apply the same point pairs used in inspection to ensure the stack-up model matches real measurement geometry.

Figure · Nominal gap vs worst-case gap (dimension chain)

Package / Pad Side A Package / Pad Side B X a b c Worst-case gap X-(a+b+c) Pass gate ≥ Xreq ?

H2-10. Verification & Documentation: How to Make Separation Auditable

Intent

Separation must be reproducible and auditable. This section defines what must be labeled in design files, how to measure and record creepage/clearance consistently, and how to close the loop when a review says “not enough”.

H3 · Design artifacts: what must be explicitly labeled

  • Isolation boundary: show the corridor and the point pairs that define the minimum gap.
  • Keep-outs: list copper/mask/silkscreen constraints; show slots and edge-to-copper rules.
  • Inputs: link PD/CTI/coating policy (Y/N) and altitude assumption to the drawing set.

H3 · Inspection: how to measure creepage/clearance consistently

  • Use fixed measurement point pairs and a defined method (microscope/fixture template — placeholders) so results are reproducible.
  • Layered checks: incoming (connectors/modules), in-process (solder shape), outgoing (critical minimum gaps).
  • Record X/Y/N plus photo ID for each critical point pair; avoid “freehand” interpretation during audit.

H3 · Remediation: if review says “not enough”, prioritize fixes with controlled impact

  1. Tighten keep-outs (fastest, low risk): prevent copper/mask/silkscreen crossings.
  2. Add/extend slots (strong geometry): force surface route detours for creepage.
  3. Change package/connector (higher impact): remove the shortest-path hijacker.
  4. Add barriers (needs mechanical alignment): document tolerance reserve and minimum points.

Data Block · Evidence Pack Checklist (audit-ready) — placeholders

Design files
  • PCB fabrication drawing (isolation boundary + key dimensions)
  • 3D snapshots / point-pair list (reproducible minimum points)
  • Rule set (keep-outs + slots + edge-to-copper)
Manufacturing files
  • Assembly drawing (critical parts + tolerance notes)
  • Incoming inspection plan for connectors/modules (X/Y/N)
  • Process controls relevant to gap risk (placeholders)
Inspection records
  • Inspection report template (point pairs + X/Y/N + photo ID)
  • Outgoing QC samples (min gap evidence)
  • Nonconformance & disposition records (if any)
Change control
  • Change log (why + what + affected points)
  • Re-verification results (X/Y/N)
  • Final audit closure statement (placeholder)

An evidence pack is complete only when minimum point pairs, measurement method, and photo IDs are tied to the revision history.

Figure · Evidence pack loop: design → build → inspect → approve (closed loop)

Design Files Fab / Asm Files Inspect Records Audit Pass Change Log If fail → loop Re-verify

H2-11. Engineering Checklist: Design → Bring-up → Production

Intent

Convert the entire page into a gate-based, audit-ready checklist. Every item includes Owner, Evidence, and Pass criteria (X/Y/N placeholders) so spacing compliance is reproducible across design, prototype, and production.

Data Block · Checklist item template (copy/paste format)

Action
(imperative verb) + measured point pairs + affected corridor
Owner
(EE / PCB / ME / QA / CM / Compliance) — placeholder
Evidence
(drawing link / point-pair list / microscope photo ID / inspection report rev) — placeholder
Pass criteria
X/Y/N placeholders (e.g., clearance ≥ X mm; creepage ≥ Y mm; coverage ≥ N%)

Design checklist (Design Gate)

Lock design inputs
  • Owner: (placeholder)
  • Evidence: input sheet rev (placeholder)
  • Pass criteria: all fields populated (X/Y/N)
Freeze isolation boundary + measurement point pairs
  • Owner: (placeholder)
  • Evidence: fab drawing + point-pair list
  • Pass criteria: point pairs defined for every critical corridor (X/Y/N)
Implement DRC rule set (clearance / creepage / keep-outs / slots)
  • Owner: (placeholder)
  • Evidence: DRC rules export + review sign-off
  • Pass criteria: no DRC violations in isolation corridors (X/Y/N)
Apply tolerance reserve policy (worst-case gap stack-up)
  • Owner: (placeholder)
  • Evidence: stack-up table + calculation note
  • Pass criteria: worst-case clearance ≥ Xreq (X/Y/N)
BOM examples (design-time enablers)
  • Pluggable terminal block: Phoenix Contact MSTB 2,5/ 2-ST-5,08 (PN 1757019)
  • Nylon standoff/spacer: Keystone Electronics 1902C
  • High-temp masking/insulation tape: 3M 5413

Bring-up checklist (Bring-up Gate)

Measure minimum gaps at defined point pairs
  • Owner: (placeholder)
  • Evidence: microscope photo ID + point-pair log
  • Pass criteria: measured clearance/creepage ≥ required (X/Y/N)
Validate solder morphology near isolation corridors
  • Owner: (placeholder)
  • Evidence: close-up photos (worst corner)
  • Pass criteria: no peak/bridge that reduces air gap below X (X/Y/N)
Check coating coverage (edges, pinholes, thin zones)
  • Owner: (placeholder)
  • Evidence: coverage record + inspection method
  • Pass criteria: coverage ≥ N% and no critical pinholes (X/Y/N)
BOM examples (coating options for prototype validation)
  • Silicone conformal coating: Dow DOWSIL 1-2577
  • Urethane conformal coating: HumiSeal 1A33 (e.g., SKU 51383 1 L can / 51391 aerosol)
  • Silicone-modified coating: MG Chemicals 422B-340G

Production checklist (Production Gate)

Lock process window for spacing-critical steps
  • Owner: (placeholder)
  • Evidence: process control plan rev
  • Pass criteria: parameters within window (X/Y/N)
Enforce sampling plan for minimum point pairs
  • Owner: (placeholder)
  • Evidence: outgoing inspection reports (point pairs + photo IDs)
  • Pass criteria: zero critical misses per lot (X/Y/N)
Quarantine and trace failures back to evidence pack
  • Owner: (placeholder)
  • Evidence: SN → lot → process params → photo IDs
  • Pass criteria: root-cause closed with change record (X/Y/N)
BOM examples (production-friendly separation helpers)
  • Repeatable mechanical spacing: Keystone 1902C standoff (fixture-able, consistent height)
  • Standard connector family: Phoenix Contact 1757019 (pitch-controlled interface)
  • Optional cable insulation for exposed conductors: TE Connectivity / Raychem heat shrink DR-25-1-0-SP

Figure · Three-stage gates (Design → Bring-up → Production)

Design Gate Inputs + Rules Bring-up Gate Measure + Photos Production Gate Process + QC Pass Pass Fail → loop Rework + re-verify

H2-12. Applications & Selection Logic (Board-Level Separation Playbook)

Intent

Map environmental drivers (PD, altitude, CTI/material, serviceability) to board-level knobs (clearance-first vs creepage-first) and the practical techniques (slots, keep-outs, barriers, coating, tolerance reserve). This section stays at the board/separation playbook level (no device-level selection deep dive).

Bucket · Motor / Inverter

PD: (placeholder) Altitude: (X m) Tolerance: conservative

Primary knob: clearance-first when altitude/transients dominate; reinforce creepage with slots where surface routes are unavoidable.

Technique set
  • Clearance margin + worst-case stack-up
  • Slots to force longer surface routes
  • Strict copper/mask/silk keep-outs across corridors
BOM examples
  • Phoenix Contact terminal block PN 1757019 (MSTB 2,5/ 2-ST-5,08)
  • Keystone nylon spacer PN 1902C
  • Coating option: Dow DOWSIL 1-2577

Bucket · BMS / HV Systems

Wet contamination risk CTI group: (placeholder) Service: (Y/N)

Primary knob: creepage-first when pollution/condensation dominates; enforce cleanliness + coating coverage gates.

Technique set
  • Slots + surface keep-outs for creepage routing
  • Coating with documented coverage + pinhole checks
  • Incoming inspection for pitch-critical connectors
BOM examples
  • HumiSeal urethane coating 1A33 (e.g., SKU 51383 / 51391)
  • MG Chemicals coating 422B-340G
  • 3M high-temp tape 5413 (masking / controlled coverage edges)

Bucket · Medical

Leakage-sensitive Serviceability important Process consistency

Primary knob: balanced—avoid relying on “coating as a blanket fix”; prioritize geometry + auditable inspection points.

Technique set
  • Clear separation corridor with labeled point pairs
  • Conservative keep-outs + tolerance reserve
  • Coating only with documented coverage + rework plan
BOM examples
  • Dow coating DOWSIL 1-2577
  • Keystone spacer 1902C (repeatable mechanical spacing)

Bucket · Industrial I/O

Dust + humidity Connector-dominated High-volume

Primary knob: creepage-first for surface contamination; enforce incoming connector pitch + outgoing minimum-gap sampling.

Technique set
  • Slots + keep-outs to control surface routes near connectors
  • Inspection templates with point pairs + photo IDs
  • Coating when field contamination is unavoidable (process gated)
BOM examples
  • Phoenix Contact connector PN 1757019
  • MG Chemicals coating PN 422B-340G

Bucket · EV Onboard Charger (OBC)

High altitude possibility Process locked Rework limited

Primary knob: clearance-first when altitude is a requirement; creepage is maintained by geometry and controlled coating where appropriate.

Technique set
  • Altitude correction policy captured in requirements
  • Worst-case gap stack-up reserved in mechanical + PCB
  • Process-gated coating for contamination control
BOM examples
  • Dow coating DOWSIL 1-2577
  • Raychem heat shrink DR-25-1-0-SP (for exposed conductor insulation where needed)

Data Block · Quick decision matrix (Condition → Knob → Technique → Trade-off)

Condition Primary knob Recommended technique Example BOM lines Expected trade-off
High altitude (X m) + OVC-driven transients Clearance-first Corridor widen + conservative tolerance reserve + keep-outs Keystone 1902C (spacing control) Board area / mechanical constraints
Wet contamination / condensation risk Creepage-first Slots + surface keep-outs + coating coverage gate Dow DOWSIL 1-2577 / HumiSeal 1A33 / MG 422B-340G Process control + rework planning
Connector-dominated spacing (pitch-limited) Depends Control the connector family + incoming inspection + keep-outs Phoenix Contact 1757019 (MSTB series) BOM lock-in / mechanical envelope
Masking / edge control during process Support Controlled masking to prevent unintended surface bridges 3M polyimide tape 5413 Labor + process discipline

Use the matrix as a fill-in policy: it records why a knob was chosen and which evidence must be produced later (photos, reports, point-pair measurements).

Figure · Scenario → Strategy matrix (board-level playbook)

Motor / Inverter BMS / HV Systems Medical Industrial I/O EV OBC PD Altitude CTI Service Clearance-first Creepage-first Process gates Tolerance reserve

Request a Quote

Accepted Formats

pdf, csv, xls, xlsx, zip

Attachment

Drag & drop files here or use the button below.

H2-13. FAQs (Audit, Acceptance, and Field-Rework Criteria)

Intent No new domains 4-line fixed answers Data placeholders (X/Y/N)

These FAQs close on review wording, inspection criteria, and field rework for creepage/clearance, pollution degree, CTI/material group, altitude correction, coating/process evidence, and worst-case tolerance stack-up.

Reviewer says clearance is OK but creepage is not — what did we likely miscount?
Likely cause
The creepage path was counted as a straight line, ignoring the actual surface route (around mask openings, pad edges, steps, slots, or board edge contours).
Quick check
Compare (1) CAD “shortest surface path” vs (2) the defined point-pair list: confirm whether the measured route crosses solder-mask dams, pad corners, via-in-pad, or the board edge. Evidence IDs: PointPair={PP-__}, DrawingRev={D__}, PhotoIDs={P__..P__}.
Fix
Re-define the creepage route on the surface: add slot/keep-out to force detours, widen the isolation corridor, and lock a “surface-only” rule in DRC for the corridor.
Pass criteria
Effective creepage (surface route) ≥ Y_req mm for PD={__} & CTIGroup={__}; clearance (air) ≥ X_req mm (if applicable). Evidence complete: PointPair log + photos + drawing notes. Result: X/Y/N.
A slot was added, but audit still rejects creepage — first check what geometry rule?
Likely cause
The slot did not actually force a longer surface route due to edge-to-copper being too small, slot ending in the wrong place, or the measured creepage path bypassing the slot.
Quick check
Verify three dimensions on the corridor: SlotWidth={S_w} mm, SlotEndOffset={S_e} mm, SlotEdgeToCopper ≥ {Z} mm. Re-run the “surface path” measurement using the same PointPair={PP-__}.
Fix
Extend the slot so the shortest surface route must detour; increase edge-to-copper keep-outs; add a “no copper / no mask opening / no silk” band around the slot edge.
Pass criteria
Effective creepage after slot ≥ Y_req mm; SlotEdgeToCopper ≥ Z_min mm; no bypass route shorter than Y_req. Photos: {P__..P__}, DrawingRev={D__}. Result: X/Y/N.
Same PCB, different lab comments — what documentation is usually missing?
Likely cause
The evidence pack is incomplete: missing isolation boundary definition, point-pair list, measurement method, or a clear link between CAD geometry and inspection records.
Quick check
Confirm these exist and are cross-referenced: DrawingRev={D__} includes isolation corridor callouts; PointPairs={PP-1..PP-n}; InspectionReportRev={R__} logs min values + PhotoIDs={P__..P__}; ChangeLogRev={C__} explains any spacing edits.
Fix
Publish a single “Spacing Evidence Pack” index: drawings + point-pair definitions + measurement procedure + inspection template + revision mapping. Use the same naming scheme across labs.
Pass criteria
All required artifacts present and consistent: {D__, PP-list, R__, C__}. Min clearance/creepage results trace back to the same point pairs (no ambiguity). Audit outcome: X/Y/N.
Clearance passes at sea level, fails for high-altitude requirement — fastest correction?
Likely cause
Altitude correction was not applied (or was applied only to PCB copper spacing but not to connectors, module cavities, or board-edge routes).
Quick check
Recompute requirement with Altitude={A_m} and method={AltCorrMethod__}. Identify the minimum air gap point pair (often near board edge or connector pins): PointPair={PP-__}, Clearance_meas_min={X_meas} mm.
Fix
Increase the air-gap corridor locally (move copper/edge route back, widen keep-out, adjust footprint or mechanical spacing). Update requirement documents to explicitly declare Altitude={A_m}.
Pass criteria
Clearance_meas_min ≥ X_req(A_m) mm for all critical point pairs {PP-1..PP-n}; requirement statement includes Altitude={A_m}. Evidence: photos + drawing callouts. Result: X/Y/N.
Coating applied, but leakage is still high — first suspect what process gap?
Likely cause
Coating coverage is incomplete (edge thin zones, pinholes, shadowed regions), or contamination remained under the coating (flux residue + humidity → conductive film).
Quick check
Inspect the isolation corridor for pinhole count and edge thickness: Coverage={N%}, ThicknessTarget={T_um}, PinholeCount={n}. Cross-check cleaning record: CleanlinessMetric={C__}. PhotoIDs={P__..P__}.
Fix
Add a gated process step: pre-clean + verify cleanliness, then controlled coating with documented coverage and pinhole inspection. Where feasible, add geometry help (slot/keep-out) so coating is not the only barrier.
Pass criteria
Coverage ≥ N_min%; Thickness within T_target ± ΔT μm; PinholeCount ≤ n_max; leakage metric meets acceptance threshold {L__}. Evidence: process record + photos. Result: X/Y/N.
Nominal distance is met, but worst-case stack-up fails — where does margin get eaten?
Likely cause
Worst-case gap is reduced by placement tolerance, solder fillet growth, and edge machining/copper swell that were not included in the stack-up budget.
Quick check
Compute: X_worst = X_nom − (a_place + b_solder + c_edge + d_mech). Compare with requirement X_req: X_worst={__} mm vs X_req={__} mm. Identify which term dominates and link to evidence (photos/assembly spec).
Fix
Increase nominal corridor or add geometry protection (keep-out band, slot). Tighten assembly controls where feasible (placement window, solder volume control), and lock the worst-case formula into the evidence pack.
Pass criteria
X_worst ≥ X_req and (if creepage relevant) Y_worst ≥ Y_req. Stack-up table is published with terms {a,b,c,d} and values. Result: X/Y/N.
Silkscreen or solder mask crosses the isolation boundary — does it matter and why?
Likely cause
A “non-conductor” layer was treated as harmless, but it can change the auditable surface path, create unintended bridges/contamination traps, or violate keep-out rules required by the spacing definition.
Quick check
Check the corridor layers: CopperKeepOut={Y} mm, MaskKeepOut={Y_m} mm, SilkKeepOut={Y_s} mm. Confirm whether mask openings or silkscreen inks create a shorter creepage route on the surface. Evidence: CAM outputs + photos.
Fix
Enforce a full-layer keep-out band across copper, solder mask openings, and silkscreen in the isolation corridor. Update fabrication notes so the corridor is unambiguous in production.
Pass criteria
No mask openings or silkscreen ink within the corridor keep-out band; measured creepage/clearance unchanged or improved. CAM + fab drawing explicitly defines keep-outs. Result: X/Y/N.
Connector footprint meets pitch, but creepage fails — what’s the typical path mistake?
Likely cause
Pitch is not the same as creepage: the surface route may run along pad edges, mask openings, housing contours, or via fields, producing a shorter effective creepage than expected.
Quick check
For the critical pins, trace the actual surface path: pad-to-pad along solder mask openings and around the connector body. Record: ConnectorRef={J__}, PointPair={PP-__}, Creepage_path={Y_eff} mm.
Fix
Add keep-outs around the footprint, increase pad-to-corridor distance, and remove bypass features (unnecessary vias/copper) within the creepage route. If the housing creates a short path, adjust the corridor boundary around the connector outline.
Pass criteria
Y_eff (surface) ≥ Y_req mm for PD={__}, CTIGroup={__}; corridor keep-outs documented for ConnectorRef={J__}. Measurement + photos attached. Result: X/Y/N.
Dusty environment field returns — increase creepage or change coating strategy first?
Likely cause
Dust plus humidity forms a conductive film on the surface; creepage becomes the dominant risk driver unless the surface is reliably protected and kept clean.
Quick check
Identify return condition: PD_assumed={__} vs PD_observed={__}; check whether contamination is visible in the isolation corridor. Confirm current creepage margin: Y_meas_min={__} mm vs Y_req={__} mm.
Fix
If Y_meas_min is close to Y_req, increase creepage first (slots/keep-outs) to create structural margin. Then apply coating as a controlled process with coverage/pinhole gates if the environment cannot be kept clean.
Pass criteria
Creepage margin: Y_meas_min ≥ Y_req + M mm (M = field robustness reserve placeholder); if coated: Coverage ≥ N_min% and PinholeCount ≤ n_max. Field corrective action documented. Result: X/Y/N.
After rework, failures increase — what contamination mechanism is most likely?
Likely cause
Rework introduced flux residues or rework splatter near the isolation corridor; under humidity, residues become a conductive film and reduce creepage robustness.
Quick check
Inspect post-rework surfaces in the corridor and around pad edges: ResidueVisible={Y/N}. Compare cleanliness record pre/post rework: CleanlinessMetric_pre={C1}, CleanlinessMetric_post={C2}. PhotoIDs={P__..P__}.
Fix
Add a rework gate: mandatory clean + verify cleanliness + re-inspect minimum point pairs. If coating is used, re-apply with documented coverage in the repaired region (do not leave partial thin edges).
Pass criteria
CleanlinessMetric_post meets threshold {C_req}; minimum creepage/clearance re-measured at point pairs {PP-__} and meets {Y_req/X_req}. Rework log links to photos and measurements. Result: X/Y/N.
Board-edge routing reduces clearance locally — what is the fastest DRC catch?
Likely cause
The board-edge keep-out was not enforced, so copper or pads encroach the air-gap corridor near the edge, creating a smaller local clearance.
Quick check
Enable a board-edge distance rule: EdgeToCopper ≥ {E_min} mm inside the isolation corridor. Run DRC on the corridor region only and list violations by reference: {Net/RefDes/Coord}.
Fix
Pull back routing and copper pours from the board edge, widen the corridor keep-out, and lock an “edge corridor” rule into the standard DRC set for all revisions.
Pass criteria
Zero DRC violations for EdgeToCopper in the corridor; minimum clearance at board edge point pairs ≥ X_req. DRC report saved as {DRCRev__}. Result: X/Y/N.
Pass criteria for inspection — how to define “minimum measurable clearance/creepage”?
Likely cause
Inspection fails due to an ambiguous “minimum point”: different inspectors measure different locations or different path definitions (air vs surface) without a locked point-pair map.
Quick check
Confirm the existence of a point-pair list and measurement method: PointPairs={PP-1..PP-n}; Tool={microscope/caliper/fixture}; Resolution={R_um}; Photo evidence required={Y/N}.
Fix
Define “minimum measurable” as the minimum across a fixed set of point pairs, with a fixed tool and resolution. For creepage, specify the surface route constraints (no shortcuts). Require photo IDs for each minimum point.
Pass criteria
For each PP-i: Clearance_meas(PP-i) ≥ X_req and/or Creepage_meas(PP-i) ≥ Y_req. Measurement tool resolution ≤ R_max μm; PhotoIDs captured for minima. Inspection template Rev={R__}. Result: X/Y/N.