3.2 Functional & Load Tests
Simulators/fixtures that catch wiring and contact issues under load.
Functional and load testing goestransforms beyondwire “pinsharness arevalidation connected” to prove thatfrom a harnessstatic performschecklist underinto reala electricalreal-world performance trial. By pushing current through contacts, monitoring voltage drop, and mechanicalapplying conditions.motion Bystress, drivingthese powertests netsreveal at rated loads, measuring voltage drop and temperature rise, and flexing the loom during operation, you uncover weak crimps, bad contacts, and intermittentsweaknesses that continuity alone won’tcannot catch.expose. SimulatingThe productprocess interfaces—whethersimulates CAN, USB, or sensors—verifies correct wiring and terminations without running full firmware. Well-designed fixtures use mating connectors, proper current ratings, Kelvin sensing, and variant lockouts to protect pins and ensure accuracy. With results logged to each serial number, load testing becomes the step that catches hidden faults beforehow the harness everwill reachesbehave inside the product.product, ensuring not just electrical connectivity but reliable performance under operating conditions.
3.2.1 Why load matters (the 30-second pitch)
Continuity says “the pins touch.” Load tells you whether they work: contacts heat, voltage drops grow, intermittents show up when you flex the loom. A good functional test simulates the product and makes problems appear before the box does.
3.2.2 What a functional tester looks like (block view)
- Mating fixtures: real connectors or guided pogo arrays; keyed, with replaceable wear parts.
- Switch matrix / relays: route loads/stimuli to any net; isolate sensitive lines.
- Programmable loads & sources: DC supplies, electronic loads, PWM drivers, coil drivers, lamp loads.
- Sense & protection: Kelvin sense on power nets, high-side current sense, fast fusing/foldback.
- Comms emulators: CAN/LIN/RS-485/UART/USB/Ethernet loopbacks or nodes.
- Motion/perturbation: wiggle bars / mild shaker / flex mandrels to reveal intermittents.
- Guarding: interlocked lid, E-stop, HV/HOT indicators.
- Software: selects program by scan of PN–Rev–Variant; pushes results to MES by SN (20.5).
3.2.3 Core tests (in a safe order)
- Low-current functional
- Switches/sensors: debounce and logic states.
- Comms physical: presence/ID, proper termination (e.g., CAN ≈ 120 Ω bus).
- Load & drop
- Drive each power net at 50–100% rated current; measure Vdrop end-to-end and per contact (Kelvin).
- Compute R = V/I, log per path.
- Thermal settle
- Hold worst-case load 2–5 min; measure ΔT at connectors/suspect splices (IR camera or stick-on dots).
- Hold worst-case load 2–5 min; measure ΔT at connectors/suspect splices (IR camera or stick-on dots).
- Wiggle under load
- Flex to the drawing’s min radius; cycle bends 30–60 s while logging dropouts & Vdrop jitter.
- Flex to the drawing’s min radius; cycle bends 30–60 s while logging dropouts & Vdrop jitter.
- Dynamic profiles (if applicable)
- Crank profile, PWM ramps, relay chatter, inrush.
- Crank profile, PWM ramps, relay chatter, inrush.
Stop on unsafe current/heat; record partial data for analysis.
3.2.4 Power & ground paths (numbers that help)
- Target drop: as engineered in 19.1; if none given, start with ≤ 5% of rail at full load and ≤ 50 mV per mated contact on low-voltage power (tune to vendor spec).
- Temperature rise: ≤ 30 °C over ambient at steady load for tin contacts (use vendor limits if tighter).
- Symmetry: parallel returns should be within 10% drop of each other.
Tip: log Vdrop vs current at 25/50/75/100% to catch non-linear contact behavior (fretting/oxidation).
3.2.5 Intermittents & contact fretting (how to catch the sneaky ones)
- Glitch monitor: sample continuity at ≥1 kHz while flexing; flag any open > 1 ms (choose per product risk).
- Jitter metric: standard deviation of Vdrop during wiggle; sudden spikes imply micro-opens.
- Connector cycling: mate/unmate 5–10× during NPI; watch contact R drift.
3.2.6 Protocol-aware checks (quick, honest)
Keep it functional-light: we prove harness health, not firmware features.
3.2.7 Loads & sources (recipes that behave)
- Electronic loads for DC rails; slew-rate limit to avoid violent inrush unless you are testing inrush.
- Relay/coil banks for inductive nets; add flyback as in the real product.
- Lamp/filament simulators for automotive style loads (inrush ≈ 10×).
- PWM drivers for dimmers/motor leads (test at duty steps 10/50/90%).
- Current clamps with 1% accuracy for cross-check; log at 10–100 Hz.
3.2.8 Fixture design (fast, durable, and kind to pins)
- Use mating connectors whenever possible; pogo only for robust round pins.
- Rated current on every contact/pogo; short, heavy bus bars for high A.
- Kelvin points brought out near each contact for accurate R.
- Replaceable tips/inserts; count cycles and PM (18.1).
- Color/shape coding by variant; scanner blocks the wrong fixture.
3.2.9 Example starter limits (tune for your product)
3.2.10 Safety first (load test edition)
- Interlocked lid; E-stop; guarded HV/HOT.
- Fused outputs and foldback on supplies.
- Discharge any bulk capacitance before lid unlocks.
- One-hand rule; no ESD strap when high energy is exposed.
- Daily self-test: fixture ID, relay click test, load zeroing.
3.2.11 Recording (what to store with the SN)
- Program/fixture IDs, operator, ambient Temp/RH.
- For each loaded net: I, Vdrop, R, ΔT, pass/fail.
- Glitch counts/timestamps; comms pass bits.
- Photos if the station requires.
- All tied to the harness SN (20.5).
3.2.12 Common traps → smallest reliable fix
3.2.13 Pocket checklists
Before test
- Program selected by scan (PN–Rev–Variant)
- Fixture mated; fans/guards OK; loads zeroed
- Kelvin clips on power paths; thermals ready
Run
- Low-current functional PASS (switches, sensors, IDs)
- Load at 25/50/75/100%; log V, I, Vdrop
- Hold worst case 2–5 min; record ΔT
- Wiggle at min radius; glitch monitor 0 >1 ms
Close
- Results to SN; any fails to NG-QUAR with plot
- Fixture PM counter ticked; worn tips replaced
- One kaizen note if time was lost