System Boundary & Intent Map (What this page solves)

This page is scoped to the opener control electronics and its evidence chain: motor actuation, position sensing (Hall/optical/limit), safety interlocks, and local wireless control. It intentionally avoids app walkthroughs and cloud/platform architecture.

Most garage door opener failures can be routed into four engineering chains. Each chain is defined by: where the signal starts, what it must prove, and which two measurements can discriminate the root cause quickly.

A fast way to use this page is to start with the symptom and immediately bind it to a chain:

• “Stops mid-travel / reboots” → Motion + Power evidence first (bus droop vs true stall).
• “Learns limits but later drifts” → Position chain first (pulse quality + repeatability).
• “Reverses with no obstacle” → Safety chain first (photo-eye/edge vs current-sense false trip).
• “Remote unreliable, worse during movement” → Connectivity chain first (RF noise floor vs antenna/ground).

Architecture Overview (Power + Motor + Sensors + Radio)

A garage door opener is best understood as a set of domains with clear evidence anchors: power integrity (does the controller stay alive), actuation (does the drive produce torque), position capture (does the system know where it is), safety interlocks (is a close allowed), and wireless coexistence (does RF reliability degrade during switching events).

The “minimum architecture” below is sufficient to explain and diagnose the majority of field failures without drifting into unrelated platform topics.

When diagnosing field issues, the fastest discriminator is to align measurements by time: V(bus), V(logic), I(motor), and RF counters. If a failure coincides with a rail dip or reset event, it is a power-domain issue first; if rails hold but current spikes and the safety chain triggers, it is an actuation/safety discriminator issue.

Motor Drive Choices & Control Strategy (DC vs BLDC)

The motor drive determines how smoothly the door moves and how reliably obstruction logic can discriminate true resistance from switching noise or power dips. The engineering goal is not “maximum torque” but repeatable motion segments: start, cruise, and stop—each with measurable current and speed signatures.

Drive strategy should be evaluated with four measurable indicators:

• Start current: peak and duration during the first 50–200 ms (correlate to bus droop).
• Stall current: steady-state current under true obstruction or mechanical jam.
• Motion curve: acceleration → cruise → deceleration shape and repeatability across runs.
• Noise/vibration proxy: current ripple and switching-event timing relative to vibration reports.

Current Sensing & Stall/Obstruction Detection (Evidence-based)

Obstruction detection is only reliable when current evidence is aligned with power integrity evidence. A “reverse event” can be caused by a real load increase or by a brownout/UVLO episode that corrupts current shape and triggers protection. The discriminator is time-aligned I(motor) + V(bus) + V(logic).

A robust decision uses a three-layer criteria stack:

• Layer 1 (instant guard): hard overcurrent threshold for protection.
• Layer 2 (envelope): averaged current or envelope rise for noise immunity.
• Layer 3 (segment logic): different thresholds for Start / Cruise / Stop to avoid start-spike false trips.

Position Sensing: Hall vs Optical vs Limit (Calibration included)

Position control is only trustworthy when the signal chain can prove pulse integrity, repeatability, and recoverability after power events. “Sensor present” is not enough; the system must detect missing pulses, edge jitter, and endpoint drift using device-side evidence.

Calibration should be treated as a verifiable state machine, not a one-time action:

• First-run learn: establish the travel baseline (total counts + segment timing) with validity checks.
• Endpoint learn: endpoints are a distribution, not a single point (repeatability window across multiple runs).
• Power-loss recovery: if a reset/brownout can lose counts, position must downgrade to “untrusted” until re-anchored.
• Re-cal triggers: drift trend, missing-pulse counter, or unstable stop segment must trigger re-learn before hard failures appear.

Safety Interlocks & Redundancy (Photo-eye, Edge, Door State)

Safety interlocks must be fail-safe and fault-detectable. A safe system treats open circuits, shorts, stuck inputs, or blocked sensors as unsafe states. Actions (inhibit close, stop, reverse) must be traceable via device-side fault codes and timestamps.

A simplified interlock truth model keeps behavior deterministic:

Redundancy is most effective when faults are explicitly detectable:

• Open circuit detection: pull-up/pull-down defines “invalid” state; transition absence is logged.
• Short detection: input stuck at rail; self-test stimulus produces no change.
• Blocked/occluded detection: sensor remains in unsafe state beyond timeout; treat as persistent unsafe.
• Power events: if brownout/reset occurs near a trip, motion must not resume without re-check.

Wireless Links: Wi-Fi / BLE / Sub-GHz Roles (Device-side only)

Link reliability is often limited by coexistence: antenna placement, ground reference stability, and coupling from motor PWM and return currents into RF-sensitive nodes. This chapter focuses on device-side evidence (RSSI, retry count, packet loss) and hardware requirements.

Coexistence must be treated as a measurable noise path:

• Antenna placement: keep distance from high-di/dt loops (H-bridge switching node, bus capacitor return); preserve an RF keep-out zone.
• Ground reference: avoid PWM return currents sharing RF ground reference; ground bounce commonly reduces effective sensitivity.
• Power domains: RF/MCU rails must remain above brownout thresholds during motor start and during Wi-Fi Tx bursts.
• Evidence anchors: correlate RSSI, retry counter, and packet loss against motor PWM timing and door motion segments.

Power Integrity: Brownout, Inrush, and Backup Power (if present)

Many “half-travel stop”, “random reboot”, and “false obstruction” incidents are power events: motor start inrush causes VBUS droop, triggering UVLO or MCU brownout/reset. Power integrity must be proven with voltage waveforms, reset-cause records, and event counters.

Protection and mitigation should be specified with “how to prove it” evidence:

• UVLO (driver + logic): confirm with UVLO status/fault pin aligned to VBUS waveform.
• Soft-start / inrush control: prove by reduced VBUS dip duration and lower peak current at start.
• Energy buffer: bulk capacitance keeps logic/RF rails above reset threshold during motor start and Wi-Fi Tx bursts.
• Domain isolation: separate motor return from RF/MCU reference; verify reduced RSSI/retry correlation during motion.
• Reset supervision: log reset cause (BOR/POR/watchdog) and brownout count; never treat a reset as “obstruction”.

EMC / ESD / Surge & Rugged I/O (Garage door opener boundaries)

Practical stress cases are dominated by long sensor cables (photo-eye / edge / limit), motor harness switching, and relay contact spikes. Design boundaries here focus on return paths and evidence: waveforms, reset causes, and glitch counters that prove an EMC-triggered event rather than a software fault.

Rugged I/O protection patterns (port → minimal network → layout boundary):

• Photo-eye / limit inputs: series-R + RC bandwidth limit + Schmitt/clean capture; keep protection-to-ground return short and local.
• Edge sensor (contact/cap/resistive): debounce + fault detect (open/short/stuck); avoid routing the clamp return through sensitive MCU ground.
• Relay coil / contact environment: coil clamp near the coil; contact snubber near the contact; minimize loop area on high dv/dt nodes.
• Power entry: surge clamp + inrush control + domain isolation; verify that logic/RF rails remain above reset thresholds during motor start.

Validation & Field Debug Playbook (symptom → evidence → isolate → fix)

This playbook uses a repeatable template: Symptom → First 2 measurements → Discriminator → First fix. Each block is short and evidence-driven.

IC / BOM Selection Map (Concrete categories + key parameters + example MPNs)

This chapter is a selection map, not a product catalog. Each bucket lists the must-have parameters, the evidence anchors used for validation (waveforms/counters), and 2–4 example MPNs to cover common cost/performance tiers. Sub-GHz parts must match the regional band (e.g., 315/433/868/915 MHz).

Notes for MPN usage: verify rail voltage class (12/24V), peak/stall current, thermal path, and Sub-GHz band legality per target region. Keep IO protection returns local to avoid injecting noise into MCU/RF reference.