H2-1. What makes a Landscape/Floodlight driver different (Definition + boundaries)

Purpose of this chapter: Define the unique constraints of landscape/floodlight luminaires, lock scope boundaries, and establish an evidence-first style for the rest of the page.

1) Outdoor “stress stack” that dominates real failures

Landscape and floodlight luminaires typically run in sealed or potted enclosures where environmental stress is the primary driver of field returns. The dominant stress stack is not a single factor, but a combination of moisture, thermal cycling, contamination, and cabling transients that interact over years.

• Moisture/condensation: creates leakage paths and corrosion, leading to intermittent open-circuit symptoms and visible flicker.
• Thermal cycling: repeated expansion/contraction increases risk of micro-cracks, connector fretting, and potting delamination.
• Salt fog / dust / insects: accelerates corrosion and contamination, raising contact resistance and local heating.
• Sealed housing + potting: improves ingress resistance but increases thermal resistance, making derating and stability more critical.
• Long cabling / imperfect grounding: turns normal switching and plug events into damaging voltage/current spikes.

2) Two common system archetypes (classification only)

This page treats the driver as a system block and focuses on the constraints that remain true across implementations. Two archetypes cover most landscape/floodlight deployments:

• A) AC mains integrated luminaire: driver is inside the fixture, facing line variability, surge/EMI pressure, and sealed thermal conditions.
• B) Low-voltage landscape CC module: a constant-current module powered by 12/24VAC or 24/48VDC distribution, often with long cable runs and connector aging.

3) What this page optimizes (three design axes)

• Reliability axis: sealing/potting choices must not create thermal or mechanical failure modes; derating must be stable (no “hunting” flicker).
• Waveform axis: LED current is treated as a waveform, not just an average—its ripple spectrum impacts EMI, visible flicker, camera banding, and thermal cycling.
• Lumen maintenance axis: brightness consistency over life requires bounded, traceable compensation (runtime/temperature history), with rollback and safety caps.

Evidence fields (measurement & field data rules for this page)

All later chapters reference these evidence categories. If evidence cannot be measured or logged, the claim is considered weak for outdoor reliability work.

• Environment targets: temperature range, sealing/potting method, salt/contamination expectations, installation style (wall/ground/landscape).
• Field failure statistics: water ingress, corrosion, intermittent open/short, flicker complaints, connector damage—tracked as counts/percentages.
• Behavior evidence: waveforms (VIN spikes, LED current ripple), plus event logs/counters (restart reason, OVP/UVLO trips, open/short detections).

H2-2. Electrical environment model (input, cabling, transients)

Purpose of this chapter: Build a measurable electrical model of real outdoor installations so later reliability, protection, and waveform decisions are grounded in probe-able evidence.

1) Input variability: what changes outdoors and why it matters

Outdoor luminaires often see supply conditions that are “good enough” for simple loads but unstable for tightly regulated LED current. Input variability matters because it can repeatedly push the driver across undervoltage thresholds, restart states, or protection boundaries—creating visible flicker and long-term stress.

• Line sag / long distribution runs: increases brownout events, especially during cold start or high load.
• Generator or shared circuits: adds low-frequency ripple and step changes that can trigger restart loops.
• Switching events: create fast edge spikes that stress clamps, insulation, and switching devices.

2) Cable model: the “hidden energy storage” that creates spikes

Outdoor cabling is not a perfect conductor. Long runs behave as distributed inductance and capacitance. During plug/unplug, load switching, or intermittent opens, stored energy can turn into voltage overshoot and ringing. Grounding and shielding are often imperfect in field installations, making common-mode disturbance more likely.

• Inductance (L): converts fast current change into voltage overshoot (di/dt → spike).
• Capacitance (C): forms resonant ringing with input filters and protection elements.
• Imperfect return paths: increases common-mode noise and radiated EMI risk.

3) Load model: LEDs + connectors age into intermittent faults

In outdoor fixtures, “intermittent” is the default hard problem. Moisture and contamination increase connector contact resistance, which raises local heating and accelerates degradation. Multi-string arrangements can amplify symptoms when one string becomes intermittent, shifting current/voltage stress onto the remaining strings.

• Corrosion / fretting: increases contact resistance → heat → intermittent opens → flicker complaints.
• Intermittent open events: can trigger output overshoot and protective retries.
• Short events: require controlled current limiting; aggressive restart behavior can create visible “blinking.”

Evidence fields: what to measure (probe plan)

Measure both ends (input and LED side) and keep conditions recorded so results are reproducible.

• Input (VIN side): VIN_min/VIN_max under real load, cable length, spike peak amplitude, dv/dt, and ringing behavior.
• Driver behavior evidence: restart reason (UVLO/OVP/OTP), retry cadence, open/short detection counters.
• LED side: LED current ripple amplitude and spectrum, open-circuit overshoot voltage, short-circuit current-limit waveform shape.

Why this model matters: Cabling transients and input sags are the root causes behind many “mystery flicker” cases. Later chapters (protection stack, thermal/derating, waveform shaping, and lumen maintenance) assume this model and reference its evidence fields.

H2-3. Potting & sealing strategy (reliability first, not cosmetics)

Purpose: Treat potting/sealing as a system-level reliability decision that reshapes thermal paths, mechanical stress, leakage paths, and serviceability. Focus on selection trade-offs, process risks, and evidence.

1) What potting is trying to achieve (and the hidden costs)

Outdoor potting is rarely “one goal.” Each intended benefit introduces secondary risks that must be designed around—especially under thermal cycling and moisture exposure.

• Moisture barrier: reduces condensation-driven leakage paths → cost: trapped moisture can remain inside longer; edge sealing becomes the weak link.
• Vibration / shock damping: improves mechanical robustness → cost: cure shrink and CTE mismatch can load solder joints and component bodies.
• Corrosion isolation: slows contamination and salt ingress → cost: connectors and cable entry points become single-point failures.
• Tamper resistance: discourages field modification → cost: repairability drops sharply; diagnostic access becomes mandatory.
• EMI damping (sometimes): can reduce micro-movement/noise coupling → cost: often worsens thermal resistance and makes failure analysis harder.

2) Material selection: engineering trade-offs (not a materials catalog)

Material properties should be chosen by how they change predictable failure modes after years of cycling and humidity—not by a single “best” metric.

• Thermal conductivity vs. thermal resistance: potting usually increases system thermal resistance unless the path to the housing is designed end-to-end.
• Dielectric strength: protects against high humidity leakage and transient stress, especially near edges and cable entry.
• Water absorption: high absorption can turn the compound into a moisture reservoir, destabilizing insulation resistance over time.
• Cure shrink / modulus: higher shrink or stiffness increases mechanical stress concentration around large components and solder joints.
• Reworkability: “non-reworkable” is acceptable only when diagnostic access remains possible (test points/log readout).

3) Structure & process: where potting fails in the field

The dominant risks come from geometry and process variability: voids, uneven fill height, and stress concentration at edges and tall components.

• Fill height & edge sealing: too high → thermal and serviceability penalties; too low → creepage/leakage risk at boundaries.
• Voids/bubbles: act as thermal insulators and stress concentrators → local hotspots, cracking, and moisture accumulation.
• Stress concentration points: connector bodies, inductors/transformers, MOSFETs, and current-sense shunts.
• Moisture traps: enclosure cavities and cable-entry pockets can retain condensation and accelerate corrosion.

4) Designing a “serviceability window” (diagnose even if rework is limited)

Outdoor reliability work improves fastest when failures remain classifiable in the field. A small diagnostic window can prevent repeated blind replacements.

• Keep key test nodes accessible: VIN sense, LED current sense, protection flags, and ground reference points.
• Do not fully bury critical interfaces: programming/test pads, serial debug pins, or sealed access ports (mechanical access only).
• Partition potting: avoid fully encapsulating connectors and high-stress components unless the mechanical load path is controlled.

Evidence fields (process & reliability proof)

• Void rate: post-cure void/bubble assessment (X-ray or sectioning if available) and trends by batch/process.
• Thermal impact: before/after potting thermal resistance change (steady-state ΔT vs power).
• Insulation robustness: insulation resistance after damp-heat exposure; leakage sensitivity near edges/cable entry.
• Failure localization: correlation between field failures and stress points (connector, inductor, MOSFET, shunt).

H2-4. Wide-temperature thermal design (potting makes it harder)

Purpose: Build a wide-temperature thermal closed loop: model → measurement points → derating strategy → protection behavior → validation evidence. Potting increases thermal resistance and slows sensing dynamics, so stability (anti-hunting) is a core requirement.

1) Two ends of the temperature range: different failure signatures

Wide-temperature operation is not a single spec. Cold and hot conditions produce different symptoms, and the evidence required to diagnose them differs.

• Cold end: startup instability, parameter drift, and cable/connector stiffness can amplify contact issues and input sag.
• Hot end: potted assemblies trap heat, raising LED board temperature and accelerating lumen depreciation; driver hotspots can trigger derating or protection.

2) Thermal path planning: LED side vs driver side (and their coupling)

Outdoor luminaires must manage both the LED thermal path and the driver thermal path. Potting can improve environmental robustness while making localized hotspots harder to remove.

• LED path: LED board → MCPCB → housing → ambient. Housing temperature is a proxy; the board-to-housing delta indicates thermal resistance changes over time.
• Driver hotspots: MOSFETs, rectifiers, and magnetics can become localized hotspots under potting; hotspot control must remain stable across airflow and mounting variations.
• Coupling: driver bay temperature may be elevated by LED bay heat soak; isolation vs coupling should be deliberate (layout, conduction paths).

3) Derating strategy: where to sense, how to avoid visible “hunting”

Temperature-based derating can introduce visible brightness oscillation if the sensor is slow, noisy, or poorly located. Anti-hunting behavior should be designed explicitly.

• Sensor placement (NTC/proxy): choose whether the control target is housing, LED board, or driver hotspot; mixing targets leads to unpredictable derating.
• Time constant awareness: potting increases thermal lag; fast control updates can chase delayed readings and oscillate.
• Stability mechanisms: hysteresis, rate-limited current adjustments, and minimum dwell time after entering derating to prevent flicker.
• User-visible constraint: derating should be monotonic and smooth within the perceptible range to avoid “breathing” effects.

Evidence fields (thermal loop proof)

• Tj proxy set: LED board temperature, housing temperature, and driver hotspot temperature (at least three points).
• Thermal time constant: time-to-steady-state under controlled power and ambient.
• Temperature rise curves: cold-start to steady-state profiles with aligned LED current/brightness changes.
• Derating behavior evidence: trigger points, slope, and whether derating changes introduce measurable ripple or visible flicker artifacts.

H2-7. Light-decay compensation (lumen maintenance) strategy

Purpose: Lumen maintenance is not “blindly increasing current.” A correct strategy is provable, rollback-safe, and traceable across runtime, temperature history, and inspection records.

1) Why “just add current” backfires

Outdoor luminaires age under humidity and thermal cycling. If compensation is uncontrolled, it can create a positive feedback loop: higher current → higher temperature → faster aging → even more compensation, leading to instability, early protection trips, and inconsistent brightness.

• Thermal runaway risk: compensation increases dissipation in LEDs and power devices, raising board/housing temperatures.
• Seal aging interaction: as sealing degrades, thermal resistance can worsen; continuing to compensate can accelerate failure.
• Batch/environment spread: identical compensation tables can behave differently across LED bins, optics contamination, and mounting conditions.

2) What “light decay” really contains (kept practical)

Lumen loss is typically a mixture of multiple mechanisms. Compensation must be designed to remain valid under mixed causes rather than assuming a single aging mode.

• LED package aging: gradual lumen depreciation accelerated by temperature history.
• Temperature history: “hot hours” carry disproportionately higher aging impact than cool operation.
• Optics yellowing / contamination: lens/cover aging and dirt can reduce output without being solved by higher LED current.

3) Three-layer compensation strategy (from robust to advanced)

Use layered strategies so the system can achieve predictable lumen maintenance without over-reliance on sensors or aggressive current growth.

Layer A — Runtime table (runtime counter → current trim)

A versioned lookup table trims the current reference based on accumulated runtime. This is the most controllable approach because it is deterministic and easy to audit.

Layer B — Temperature-history weighting (“hot hours”)

Runtime alone is not enough: high-temperature operation accelerates lumen decay. A weighted accumulator (“hot hours”) can adjust compensation more accurately than runtime-only methods.

• Hot-hours definition: use a consistent temperature threshold and accumulation rule.
• Cross-check with thermal loop: ensure compensation does not conflict with derating behavior (protection always wins).

Layer C (optional) — Light-sensor closed loop (use only when worth it)

A light sensor can reduce unit-to-unit variation, but it introduces sensor drift and contamination risks. It is only worthwhile when sensor stability, mounting geometry, and calibration/maintenance are controlled.

• Worth using: high uniformity required; maintenance access is limited; sensor can be protected and calibrated.
• Not worth using: sensor contamination is likely; geometry varies; long-term drift cannot be bounded.

4) Safety boundaries: cap, rollback, and anomaly detection

A safe lumen maintenance design defines when compensation must stop, reduce, or revert—especially when environmental aging changes thermal behavior.

• Max compensation cap: limit the maximum trim to avoid pushing the system beyond thermal/optical stress boundaries.
• Protection-first rollback: on repeated over-temperature, frequent restarts, or abnormal OVP/OCP events, freeze or roll back compensation.
• Anomaly detection: if the same current produces rising temperature over time (thermal resistance degradation), compensation should not continue upward.

Evidence fields (traceable lumen maintenance)

• Runtime counter: accumulated operating time with a clear persistence rule.
• Hot-hours accumulator: weighted time above a defined temperature threshold.
• Compensation curve version: table ID/version used for the trim decision.
• Current trim record: applied trim value and timestamped updates.
• Inspection records: spot lux/lumen checks before/after compensation (production + maintenance sampling).

H2-8. Moisture ingress & corrosion fault model (what fails in the field)

Purpose: Outdoor returns often come from water vapor, corrosion, leakage, and intermittent connectors rather than IC defects. This section provides a reusable fault model and an actionable “symptom → evidence → isolation” template.

1) Typical failure chain (from boundary loss to visible symptoms)

A common outdoor fault chain is progressive and non-linear: sealing weakens → condensation forms → leakage/corrosion grows → intermittent open/variable contact resistance → flicker, protection trips, or random shutdown.

• Condensation/leakage: insulation resistance drops; noise and leakage paths become unpredictable.
• Corrosion: contact resistance rises; intermittent conduction appears under vibration and thermal cycling.
• Intermittent open: output voltage spikes and open-detect events increase; protection may cycle (hiccup/restart).

2) Connectors and cables are primary fault sources

Cable entry and connectors often dominate field failures because they combine mechanical load, environmental exposure, and electrical sensitivity.

• Resistance increase → local heating: higher contact resistance creates hot spots that accelerate oxidation/corrosion.
• Mechanical intermittency: movement + thermal expansion can create brief opens that repeatedly trigger protections.
• Moisture transport: water can migrate along cables and into enclosures, worsening condensation and leakage paths.

3) “Symptom → evidence → isolation” templates (field-reusable)

Template A — Flicker

• First split: low-frequency current modulation vs protection-driven restart cycling.
• Evidence to collect: I_LED waveform (low-frequency modulation), restart reason counters, OVP/OCP trip counts.
• Isolation steps: hold a fixed brightness level; check whether flicker correlates with humidity/rain or with thermal derating events.

Template B — Random “off” (sometimes recovers)

• First suspect: intermittent open and output over-voltage clamping.
• Evidence to collect: open-detect log count, V_OUT spike waveform, OVP trigger flags, brownout flags.
• Isolation steps: gently perturb cable/connector; observe if open events spike or if V_OUT shows transient overshoot.

Template C — Rain/seasonal spike in failures

• First suspect: insulation/leakage degradation and corrosion acceleration.
• Evidence to collect: brownout/restart counters aligned to environment, leakage/insulation indications (if available), repeated clamp/limit events.
• Isolation steps: compare dry vs damp conditions; check whether failures reduce after drying/warming.

Evidence fields (logs + waveforms)

• Event logs: open/short count, restart reason, OV/OC triggers, brownout flags.
• Waveforms: V_OUT spikes during intermittent opens; I_LED entering hiccup/limit patterns during protection cycling.
• Correlation records: timestamps linking symptoms (flicker/off) with measured waveforms and trip counters.

H2-9. Protection & control behaviors (open/short, thermal, brownout) tuned for outdoor

Purpose: Outdoor protection is not only about safety. Its retry timing and recovery rules can become visible flicker. This section defines behavior contracts and tuning logic so protections converge safely without creating user-visible “restart blinking.”

1) Behavior contracts: what each fault is allowed to do

Define the action for each fault type first (shutdown vs hold-limited), then tune timing. A clear contract prevents uncontrolled interactions with dimming waveforms and lumen maintenance.

• Open-circuit (intermittent open included): prevent sustained V_OUT rise and repeated OVP spikes; prioritize a fast safe transition into a controlled retry rhythm.
• Short-circuit (hard short and “soft short” from moisture leakage): choose “limit and hold” vs “shutdown” based on thermal boundary and enclosure safety; avoid repetitive hot pulsing.
• Thermal (over-temperature / rising thermal resistance): prefer smooth derating over abrupt off/on toggling; prevent temperature hunting that becomes visible brightness modulation.
• Brownout / undervoltage: avoid frequent restarts on long cables and poor mains by using UVLO entry/exit hysteresis and a controlled restart window.

2) Hiccup timing: the difference between “safe retry” and “visible blink”

Hiccup behavior is a timing problem. If retry intervals fall into human-visible cadence, the luminaire looks like it is flickering even though it is “protected.” Tuning must consider:

• Retry period: too short looks like flicker; too long looks like a dead light. The goal is to exit the most noticeable cadence while still allowing recovery from transient events.
• On/off duty: the “on time” during retry determines whether the user sees a quick flash, a breathing effect, or a stable low output.
• Thermal impulse: repeated high-current retry bursts can create thermal shock and accelerate aging—avoid “pulsing hot.”

3) Recovery policies: auto-retry, cool window, latch vs non-latch

Outdoor failures are often repeatable (moisture, corrosion, cable intermittency). Recovery must be designed to converge instead of cycling forever.

• Auto-retry: use for transient disturbances (momentary brownout, short spikes) where success probability is high.
• Cool window: use for thermal events so the system does not re-enter fault immediately after recovery; prevents brightness hunting.
• Latch / lockout: use when faults repeat frequently (water ingress, intermittent open) to stop visible blinking and limit damage from repeated stress.

4) Coupling rules with waveform shaping and lumen maintenance

Protection strategies must not amplify the problems addressed in earlier sections.

• Waveform shaping (H2-6): avoid transitions into pulsed/unstable current modes during fault retries that can increase camera banding and visible modulation.
• Lumen maintenance (H2-7): compensation must freeze or roll back when thermal margin shrinks; never allow “compensate → overheat → derate → compensate” loops.
• Moisture/corrosion faults (H2-8): intermittent open triggers can cause repeated OVP spikes; use a recovery rhythm that limits V_OUT excursions and reduces stress.

5) Practical tuning order (portable method)

1. Set the experience target: no fast restart blinking; controlled low-output derate is preferred over repeated off/on.
2. Choose fault actions: shutdown vs hold-limited vs smooth derate for each fault class.
3. Tune timing: retry period, cool window, and latch thresholds so repeated faults converge to safe behavior.
4. Lock coupling rules: freeze compensation on repeated faults; prioritize protection and logging.

Evidence fields (prove the behavior, not assumptions)

• Restart timing: retry period distribution and retry count over time (not a single value).
• Reason codes: fault cause flags (open/short/OVP/OCP/OTP/UVLO) time-aligned with events.
• Threshold drift vs temperature: evidence that triggers change across cold/hot operation.
• Visible event correlation: timestamped flicker/blink observations aligned to restart logs and protection flags.

H2-10. BOM/IC selection checklist (what matters for this application)

Purpose: This is not a parts encyclopedia. It is a Landscape/Floodlight selection checklist that maps directly to wide temperature reliability, protection behavior, logging evidence, waveform shaping, and lumen maintenance.

1) Input & protection (surge, cold start, insulation reality)

• Surge coordination: clamp/filter/limit must act as a stack, not as isolated “strong parts.”
• Cold-start robustness: startup must remain stable at low temperature without false trips or repeated resets.
• Leakage tolerance: humidity-aged insulation and potting materials must not create nuisance triggers or ground contamination.
• Degradation behavior: after repeated events, clamp/leakage drift must remain within a predictable window.
• Boundary clarity: ensure protection components support a clean “energy boundary” (return path matters, without turning this into PCB training).

2) Power-stage controller (wide-temp control + configurable behavior)

• Wide-temp startup: consistent start across cold/hot, including brownout recovery scenarios.
• Configurable protections: support for retry/lockout logic, controlled soft-start, and fault reason reporting.
• EMI-friendly control options: controllable switching behavior and stability under cable input dynamics.
• Deterministic fault signaling: fault flags/reason codes usable for field correlation and service decisions.
• Supply sensitivity: stable behavior under long-cable drops and noisy mains (avoid restart blinking).

3) Current sensing (accuracy, drift, and potting stress risks)

• Temp coefficient & drift: sensing stability over thermal cycling (affects waveform and compensation accuracy).
• Thermal coupling: sense element temperature must not be dominated by nearby hot spots.
• Potting stress sensitivity: shunt/contacts must tolerate cure shrink and mechanical stress (crack risk).
• Noise immunity: robust measurement under high dv/dt environments and EMI coupling.
• Fault signature clarity: intermittent open/short should produce stable, diagnosable evidence instead of random toggling.

4) Thermal sensing (placement, time constant, anti-hunting)

• Sensor placement: measures the right proxy (LED board, enclosure, or driver hot spot) for safe derating.
• Time constant: too fast causes hunting; too slow misses protection. Pick a predictable response window.
• Wide-temp accuracy: avoid threshold drift that changes behavior between winter and summer conditions.
• Potting effect: potting changes thermal paths and delays; the sensing strategy must remain valid.

5) NVM & logging (runtime/hot-hours, reason codes, and versioning)

• Runtime + hot-hours: counters must survive power loss and remain consistent for lumen maintenance.
• Log coverage: open/short counts, restart reasons, OV/OC triggers, and brownout flags should be recordable.
• Write strategy: logging should not silently exhaust endurance; update rules must be bounded.
• Versioned tables: compensation curve versions and rollback records must be traceable.

6) Optional minimal MCU/logic (local-only, no ecosystem)

• Scope-limited role: only for compensation tables and local logs (no networking assumptions).
• Robust reset behavior: stable under brownout with readable reset reasons and watchdog protection.
• EMI discipline: minimize additional interfaces that become new coupling paths in a harsh enclosure.

Evidence fields (selection must map to proof)

The selection checklist should be auditable against: wide-temp behavior, protection state-machine behavior, log fields, and lumen maintenance strategy.

• Wide-temp proof: cold/hot startup evidence, threshold drift evidence.
• Behavior proof: retry timing distribution, lockout decisions, reason codes.
• Logging proof: open/short, OV/OC, brownout, restart reasons.
• Compensation proof: runtime/hot-hours counters, table version IDs, trim logs.