Write one record for every important choice (REQ-3): cell, LED, optic, driver topology, MCU, charger, housing material, sealing, firmware layout. Start the record when the question comes up (status “proposed”), list at least three options, and score them with decision-matrix.md if the choice is not obvious. Give every source with a link and the date you read it, because datasheets and prices change. Never delete a record: when a decision changes, write a new record and set the old one to “replaced by D-xx”. Number records D-01, D-02, … and refer to requirement IDs (R1, R2, …) from requirements.md and rule IDs from docs/design-rules.md.
| ID | Date | Question | Decision | Status |
|---|---|---|---|---|
| D-01 | ||||
| D-02 | ||||
| D-03 |
Copy this block for each decision.
<nn> <short title>| Field | Entry |
|---|---|
| Date | <YYYY-MM-DD> |
| Author(s) | <...> |
| Question | <what must be decided, in one sentence> |
| Requirements and rules | <R.., R..; BAT-1, ...> |
| Status | <proposed / accepted / replaced by D-xx> |
Options considered (at least 3)
| Option | Short description | Passes must-have limits? |
|---|---|---|
| A | <...> |
<yes / no: why> |
| B | <...> |
|
| C | <...> |
Criteria and weights: <criterion (weight %), ...>, matrix: <link to the filled
decision-matrix.md or .csv>
Data and sources
| Data | Value | Source (link) | Type | Accessed |
|---|---|---|---|---|
<parameter of option A> |
<value with unit> |
<link> |
<datasheet / measurement / article / retailer> |
<YYYY-MM-DD> |
Decision: <option chosen>
Why: <2-4 short points, linked to the criteria>
Consequences and risks: <what this choice forces elsewhere; what could go wrong;
mitigation>
Verification: <test or check that will prove the choice works; requirement and test
report section>
Part of an imaginary 18650 EDC light, > 1000 lm. Cell data: only the values below, from the sources listed.
| Field | Entry |
|---|---|
| Date | 2026-10-15 |
| Author(s) | <...> |
| Question | Which 18650 cell do we specify and test with? |
| Requirements and rules | R18, R20, R24; BAT-1, BAT-2, CHG-3 |
| Status | accepted |
Options considered
| Option | Short description | Passes must-have limits? |
|---|---|---|
| A | Molicel P28A | yes |
| B | Samsung 30Q | yes, if turbo cell current <= 70 % of 15 A |
| C | Murata VTC6 | only retailer data, no maker datasheet found (BAT-1) |
Criteria and weights: current headroom (40 %), capacity (30 %), charge temperature
window (15 %), data quality (15 %). Weights are illustrative; matrix: <link>.
Data and sources
| Data | Value | Source (link) | Type | Accessed |
|---|---|---|---|---|
| P28A capacity | 2800 mAh typ | Molicel datasheet | datasheet | 2026-10-15 |
| P28A max continuous discharge | 35 A | same | datasheet | 2026-10-15 |
| P28A charge temperature | 0-60 °C | same | datasheet | 2026-10-15 |
| 30Q capacity | 3000 mAh | Samsung datasheet | datasheet | 2026-10-15 |
| 30Q continuous discharge | 15 A | same | datasheet | 2026-10-15 |
| 30Q charge temperature | 0-50 °C | same | datasheet | 2026-10-15 |
| VTC6 capacity | 3000 mAh | retailer page | retailer | 2026-10-15 |
| VTC6 continuous discharge | 15 A | same | retailer | 2026-10-15 |
Decision: Molicel P28A.
Why:
Consequences and risks:
Verification: cell current at turbo from status (cell_mA) and a clamp meter, compared
with BAT_MAX_MA; runtime test per mode; both in test-report.md.