09 - Battery and full-load regression

Prerequisite: Stage 8 accepted. Firmware: breadboard-record-2ch. Status: NOT TESTED.

Sequence overview | Exact contacts | 3D assembly

Add and test

Remove both lab supply leads. Check protected pack polarity before connection. Log actual runtime for at least two hours, voltage/current/temperature and low-battery events; repeat 20 starts/stops. Firmware session limit is two hours: document restart if runtime testing continues.

Acceptance checks

Send for review

Battery protection does not provide patient isolation. Battery warning/stop thresholds are firmware decisions, not proof that the pack protection trips correctly. Do not bypass protection or intentionally overdischarge.

Continuous regression

Before power: inspect the changed wiring against the contact table and photograph it. After power: remeasure every connected rail at the load, verify no unexpected heat/current, and rerun the last accepted diagnostic. Repeat earlier affected tests whenever a supply, cable, firmware or grounding connection changes. Record failures and repairs; retain the previous accepted evidence.

If the gate fails

Stop acquisition and remove power before changing wiring. Disconnect only the last-added module and restore the last accepted configuration. Repeat its test. If it now fails, check the shared supply/ground and changed connections before replacing components. Never change multiple modules at once to chase a fault.

Results

Fill measurements.csv; retain raw instrument files alongside screenshots. Outcome starts NOT TESTED. Include units, uncertainty, instrument model/settings, source voltage/load, firmware hash and filenames. Mark unmet criteria FAIL; missing evidence PENDING. Submit the whole stage folder, including failures. A completeness check is not engineering approval.

Stage files