Przebieg r20261030-0443-v003-zadymiarka-concept-1

Requirements — compact mains fog machine (concept tier)

Owner: REQ. Source of every given requirement: INPUT/requirements.md (v0) and INPUT/brief.md. Verification conditions are published as facts req.* (board facts file of REQ) and quoted here. Status of each id: compliance.csv in this folder. At the concept tier every status is a prediction from first-order estimates; nothing in this run is a physical test.

Common verification conditions (apply to every row unless the row says otherwise):

• Ambient 20 °C for performance rows (req.ambient_nominal), re-checked at 40 °C (req.ambient_max, ZAD-REL-02 target) and 5 °C (req.ambient_min). READY-hold power at 25 °C (req.ambient_ready_hold).
• Supply: EU variant 230 V 50 Hz nominal (req.supply_eu_nominal), 207 V and 253 V limits (±10 %, req.supply_eu_min / req.supply_eu_max); US variant 120 V 60 Hz (req.supply_us_nominal), input ≤ 1,440 W (req.us_input_max). Heater power scales with V²: 207 V gives 0.81 × and 253 V gives 1.21 × the 230 V power (CALCULATION, 0.9² and 1.1²).
• Fluid: ANSI E1.5 water-glycol fog fluid; reference = a standard-density type; the densest (long-hang) type sizes the energy per ml (req.fluid_reference, ASSUMPTION).
• Test runs: 30 min after READY (req.test_run_duration); design duty: events of up to 8 h at up to 100 % output command (req.event_duty_hours, SOURCE brief).

1. Given requirements — interpretation and design consequence

2. Derived requirements

Areas checked (handbook section 2): function/performance (DR-PRF-01..03), UI and ergonomics (DR-UI-01), power and standby (DR-PWR-01, ZAD-PWR-03/04), mechanical interfaces and mounting (DR-MEC-01), environment (DR-ENV-01/02, DR-PKG-01), safety: shock (ZAD-SAF-06), heat and burns (ZAD-SAF-02/03, DR-SAF-05), fire (DR-SAF-01/02), mechanical hazards (ZAD-SAF-04, DR-MEC-01), chemical (DR-MAT-01, fluid exposure in ZAD-SAF-08), optical and other radiation (not applicable: no laser or UV source; display and LEDs only), noise (ZAD-UX-01/02), foreseeable misuse (DR-SAF-03/04/05); EMC and radio (DR-EMC-01, DR-RAD-01); materials and restricted substances (DR-MAT-01); lifetime and duty (DR-REL-01); maintenance and repair (DR-SRV-01, ZAD-SRV-02); manufacturing at 1/100/1000 (DR-MFG-01); production assembly and test (ZAD-MFG-02, DR-MFG-03); packaging and drop (DR-PKG-01); labelling and manuals (DR-LBL-01); unit cost and price (DR-CST-01); prototype schedule (DR-SCH-01).

3. Conflicts, decisions and assumptions

Screening estimate used below (CALCULATION, hand, REQ): energy to vaporise 1 ml of a water-glycol fluid ≈ 0.8 × (4.19 J/gK × 80 K + 2,257 J/g) + 0.2 × (~2.5 J/gK × 170 K + ~900 J/g) ≈ 2.34 kJ/g; with superheat of the vapour and density ≈ 1.04 g/ml → ≈ 2.4-2.6 kJ/ml. MECH owns the design value; this figure only screens conflicts.

Chosen concept (TALLY M-0012: A1, integrated cast-Al thermal-mass block, pump-follows-heat, firmware-independent cut-outs, two-way 2.4 GHz remote). Predicted against the thresholds (facts, CALCULATION): ZAD-PRF-02 20.8 ml/min at 230 V and 16.7 ml/min at 207 V (mech.continuous_rate_230v / _207v) meets the 15 ml/min threshold and DR-PRF-01 but misses the 22 ml/min target (about 1,000 W heater needed, mech.heater_need_for_prf02_target; C1 kept the 970 W input of arch.power under the ZAD-PWR-02 target). ZAD-PRF-04 17.4 ml and ZAD-CST-01 72.9 EUR meet thresholds, miss targets. No predicted threshold miss; open evidence gaps are listed as not_addressed in compliance.csv.

Assumptions register

Specification sheet — compact fog machine (concept A1, PREDICTED values)

Every number below is a design prediction of the concept tier, quoted from its fact and labelled; none is measured. Values marked "to be confirmed" have no evidence yet and are not claimed. Output is fluid vaporised in ml/min (weighed tank); no volume (m³/min) figure is claimed.

No certification mark (CE, FCC, NRTL listing) is claimed: the conformity plans are open (ZAD-CRT-01, ZAD-CRT-02).

Benchmark - compact DMX fog machines with wireless remote (COST)

Prices: ASSUMPTION - typical EU street prices recalled from distributor listings (Thomann, manufacturer pages); NOT re-read on 2026-10-30 (no time in this smoke run). URLs point to the product family pages to re-check.

Realistic targets derived

• Market gap: 150-250 EUR segment has DMX but either no included wireless remote or reheat pauses; 350-450 EUR segment has pro output.
• Our differentiator: pro-grade continuous output (heater sized above pump flow, closed-loop PID on block temperature, no reheat pauses), DMX + included wireless remote, at <= 229 EUR incl. VAT (ZAD-CST-05 target).
• Cost implication: retail 229 EUR incl. VAT leaves a unit-cost budget of about 60-80 EUR at 1000 units (see sections/COST.md).

Concept - compact mains fog machine (concept tier)

Status: final at CLOSE. Numbers are facts on board/facts/*.yaml (homes named inline). Decision record: board/decisions/D-ARCH-01.md (A1, TALLY M-0012).

1. Product in one paragraph

A compact Class I mains fog machine for ANSI E1.5 water-glycol fluids. The envelope is arch.dimensions [310, 165, 185] mm (ASSUMPTION from MECH's layout) and rated input is about 1 kW (arch.power 970 W, CALCULATION). An insulated cast-aluminium heat-exchanger block stores heat, which lets the pump keep running while the heater catches up ("pump follows heat"). Output therefore derates smoothly instead of stopping to reheat. It has isolated DMX512/RDM, a paired two-way 2.4 GHz remote and an on-unit display with an encoder. It is for mobile DJs, small venues and theatres, at EUR 229 retail (cost.retail_price, COST).

2. Functions and flows

F1-F12 with energy, material and signal flows: board/ideas/ARCH.md.

• Energy: mains -> heater 950 W (elec.heater_power_nom) -> fluid (arch.e_ml 2470 J/ml), with a loss of 49.3 W (mech.ready_hold_loss).
• Material: tank -> filter -> pump -> block bore -> nozzle.
• Signal: DMX / remote / UI -> MCU -> pump + triac; the cut-outs override in hardware.

3. Alternatives (full table and risks: board/concepts.md)

Script: calc/arch_energy.py -> calc/arch_energy.results.json, rerun with the ELEC/MECH values. The hand estimate (21.6 ml/min, 180 s) agrees with the script. The ballot was cast on the first run (975 W heater, 60 W loss); the rerun moves every option by 1-3 % and keeps the order.

4. Evaluation

The blind concept ballot is board/votes/ARCH-concept/BALLOT.yaml: requirement weights come from the input, RISK has share 0.15 and SAFETY 0.10.

• Knock-outs: none. A2's peak of 13.6 ml is above the 12 ml threshold; A3's warm-up of 4.44 min is within 5 min.
• Result (TALLY M-0012): A1 7.26, A3 5.50, A2 4.44. The winner holds under ±20 % on every weight.
• Caveat: only 2 of 6 voters (REQ, ARCH) rated, on 5 of 58 criteria.
• Details: board/decisions/D-ARCH-01.md, mesh/votes.md.

Why each loser lost:

• A2 has a high spitting risk (ZAD-PRF-07, weight 5) and its peak output is close to the threshold.
• A3 misses the warm-up target: 4.44 min at 230 V and 5.53 min at 207 V. Its thermostat band also threatens ZAD-PRF-06, and it needs two boards.

5. Chosen concept A1 - architecture

flowchart LR
  MAINS[Mains 230/120 V] --> INLET[IEC C14 inlet, fuse, switch, PE]
  INLET --> CUT[cut-outs: auto-reset + thermal fuse] --> RELAY[relay] --> TRIAC[zero-cross triac] --> HEATER[950 W heater in Al block]
  INLET --> PSU[isolated PSU] --> MCU[MCU PID + pump-follows-heat]
  NTC[block sensor] --> MCU
  TANK[1.29 L tank + level] --> FILTER[filter] --> PUMP[vibratory pump] --> CHECK[check valve] --> HEATER --> NOZZLE[nozzle + guard]
  MCU --> PUMP
  MCU --> TRIAC
  DMX[isolated DMX/RDM] --> MCU
  RADIO[2.4 GHz module] <--> MCU
  UI[display + encoder] <--> MCU

Source: architecture.mmd.

Subsystems, make or buy:

• Make: block + bore (die-cast at 1000 units, CNC for the prototype); main PCB; enclosure (Al shell + PC V-0 end caps, MECH); translucent moulded tank.
• Buy: cast-in tubular heater, vibratory pump, cut-outs, PSU module, pre-certified radio module.

Candidate part numbers were not looked up by ARCH (none invented); see the ELEC/COST sections and 09-bom-cost/. Interfaces: interfaces.md. Owner sections: sections/ELEC.md, sections/COST.md.

6. Key estimates

7. Safety concept

Inherently safe design first:

• The heater is energised only through a relay and series cut-outs that act without firmware: an auto-reset cut-out and a one-shot thermal fuse on the block.
• The pump runs only when READY.
• Class I, with PE bonding of the block and of the metal shell.
• DMX, radio and UI are on SELV behind reinforced insulation.
• The nozzle sits behind a touch guard.
• The tank is above no live parts, with a vented cap and a drip path away from the electronics.

Hazard list and FMEA: SAFE (08-safety-certification/).

8. Compliance route (owner SAFE)

• EU: CE under the radio-equipment route, covering safety (EN 60335-1, named in ZAD-SAF-01) and EMC, plus RoHS, WEEE and standby ecodesign.
• US/CA: FCC Part 15 (module grant plus unintentional radiator) and an NRTL listing.
• ESTA: E1.11 (DMX), E1.20 (RDM) and E1.5 (fluid), as named in INPUT/requirements.md.
• Exact standard editions, costs and timeline belong in SAFE's section.

9. Cost

COST facts:

• Unit cost: cost.unit_cost_q1 338.1, cost.unit_cost_q100 122.5, cost.unit_cost_q1000 72.9 EUR.
• Retail price: cost.retail_price 229 EUR.
• One-off costs: cost.one_off_total 43500 EUR.
• Gross margin per unit: cost.direct_gross_margin 93.0 EUR direct, cost.dealer_gross_margin 33.6 EUR via dealers.

Details: 09-bom-cost/, sections/COST.md.

10. Risks

11. Decisions taken without a human

See mesh/decision-log.md. ARCH's:

• Fluid composition: 75/25 water/glycol (ASSUMPTION).
• Heater 950 W (ELEC), so rated input is 970 W.
• Set point 290 C.
• Radio: a pre-certified 2.4 GHz module (POLICY 4) rather than sub-GHz.

12. Open issues for detailed design

• Heater and pump part selection, with second sources.
• Bore and nozzle design against spitting.
• SAFE FMEA.
• Per-variant BOM (ZAD-PWR-01).
• ZAD-PRF-02 target: CR M-0023 (MECH) asks for a heater of about 1000 W. ARCH's answer (M-0033): within the 1000 W budget the heater can reach 980 W; ELEC decides the heater power.

Interfaces (ARCH) - concept A1

Decision log of the peer mesh

Generated 2026-10-30T04:54:39+01:00 by method/gates/board_check.py export from the board log (34 messages: ANS 1, BLOCKED 1, CALL 1, CLOSE 1, CR 2, DECISION 2, DONE 11, INFO 7, NUDGE 1, PHASE 2, Q 1, TALLY 1, VOTE 3). Every message between peers was logged before it was sent; the raw log is log.jsonl in this folder.

Roster

Decisions (re = the CR, question, objection or tally it settles)

Change requests (2: 0 accepted, 1 answered or rejected, 1 open; median answer time 2 min)

Objections, vetoes, lifts, calls and tallies

Blocked (needs a human)

• M-0030 SAFE: Legal confirmation that Machinery Regulation (EU) 2023/1230 does not apply to the fog machine and that Ecodesign 2023/826 scope covers it :: needs a legal opinion (human)

Still open at export

• M-0023 CR MECH->ELEC,ARCH (3 min, waiting for ELEC): ZAD-PRF-02 target 22 ml/min needs heater about 1000 W at 230 V (mech.heater_need_for_prf02_target; worst fluid 2.59 kJ/ml, loss 49 W). At elec.heater_power_nom 950 W: 20.8 ml/min = PARTIAL. Request 10
• M-0028 CR SAFE->REQ (2 min, waiting for REQ): Add derived safety reqs DR-SAF-01..14 (board/ideas/SAFE.md): firmware-independent heater cut-outs (HW relay drop + thermal fuse), HW pump-enable temp interlock, ext. watchdog, fluid/electronics separa

Key numbers register (merged board facts)

Generated 2026-10-30T04:54:39+01:00. Each number has one owner; source is its home (script, results file, URL with date, or the reason of an ASSUMPTION).

Interface constraints and budgets

• arch.mass le arch.budget.mass
• arch.power le arch.budget.power
• elec.max_input_us_127 le req.us_input_max
• elec.rated_input_eu le req.eu_input_threshold
• budget mass: 2.37 kg from 3 contributions; limit 3.0
• budget power: 970.0 W from 3 contributions; limit 1000

Consistency findings at export

• none

Concept fragments posted by the peers

Each domain owner posted its own fragments; the system alternatives were assembled from them (see ../concept.md).

<!-- board/ideas/ARCH.md -->

Ideas - ARCH

Function structure (verb-noun; every fragment names the functions it serves)

Energy: mains -> F1 -> F5 (~975 W) / F2 (~18 W) -> heat into fluid (F6) + losses (~60 W) through F12. Material: fluid F3 -> F4 -> F5/F6 -> fog. Signal: F9/F10/F11 -> F7 -> F4/F5; F8 overrides F5 in hardware.

F-ARCH-1 - Integrated single main PCB, software regulation + hardware backstop

• Working principle: one main PCB on the mains side carries PSU, heater triac+relay, pump driver, MCU, isolated DMX and a pre-certified radio module; MCU runs PID and pump-follows-heat; two firmware-independent cut-outs in series with the heater.
• Functions served: F2, F7, F8, F9, F10, F11
• First-order numbers: arch.power, arch.a1_* (calc/arch_energy.py)
• Needs: block NTC/thermocouple to ELEC, heater 975 W/230 V and 975 W/120 V from MECH, cut-out on block.
• Risks: single PCB mixes mains and SELV -> reinforced creepage (SAFE); one board spin.

F-ARCH-2 - Modular: mains power module with hardware thermostat + low-voltage logic module

• Working principle: heater regulated by a capillary/bimetal thermostat (EN 60730) on the block, no MCU in the heat loop; LV module (MCU, DMX, radio, UI) only meters the pump against block temperature read by an NTC; large thermal mass rides through thermostat hysteresis.
• Functions served: F5, F7, F8 (heat in hardware), F9-F11 (logic module)
• First-order numbers: arch.a3_* (calc/arch_energy.py)
• Risks: thermostat hysteresis +/-10 K -> output stability (ZAD-PRF-06); slower warm-up; two PCBs + harness.

<!-- board/ideas/COST.md -->

COST ideas / summary

• Benchmark (deliverables/02-concept/benchmark.md): 6 products, 70-400 EUR; pro output today costs 350-450 EUR.
• Price point: 229 EUR incl. VAT (ZAD-CST-05 target) with dealer channel still positive.
• Unit-cost anchors (facts cost.*): q1000 ~76 EUR (EU assembly), q100 ~116 EUR, q1 ~346 EUR.
• Cost levers: purchased die-cast heater block with embedded element, off-the-shelf vibratory pump, pre-certified RF module (saves FCC/RED radio cost), sheet-steel enclosure (low tooling).

<!-- board/ideas/ELEC.md -->

Ideas - ELEC

Physics of my domains (handbook section 4)

F-ELEC-1 - AC drives with zero-cross triacs, 3 W supply, sub-GHz radio

• Working principle: 3 W isolated offline module (MEAN WELL IRM-03-5, LCSC C6969425, $3.93, stock 66, read 2026-10-30) feeds an STM32G071 (C529341, $1.31 @1k); heater by BTA16-600BW triac (C87396 ST $0.63 @1k / C19184935 KTP) via MOC3063 zero-cross opto (C77950 Lite-On / C3014302 UMW) with full-cycle burst control, in series with a safety relay held by a hardware watchdog; AC vibratory pump by a second triac with half-wave stroke metering, stroke rate compensated for measured mains voltage; PT1000 thin-film sensor on the exchanger; CC1101 868/915 MHz two-way radio (C29953, $1.19 @1k; a pre-certified module preferred, ASSUMPTION); DMX via SN65HVD72 (C48125) behind a B0505S-1WR3 isolated DC-DC (C7465178 EVISUN / C2992394 YLPTEC) and digital isolator.
• Numbers: rated input 970 W at 230 V (elec.rated_input_eu), standby 0.29 W (elec.standby_power), radio margin 18.8 dB at 50 m.
• Needs: MECH pump with <= 0.05 ml/stroke; bay for 100 x 70 mm board.
• Risks: pump noise (UX-01), open-loop stroke volume (PRF-06 target 5 %).

F-ELEC-2 - 12 V DC gear pump with closed-loop speed, 10 W supply, 2.4 GHz SoC module

• Working principle: IRM-10-12 (C42418573 MW / C20618065 LZTEC) + TPS54202 buck (C191884); DC gear pump PWM with tacho loop (flow independent of mains voltage); K thermocouple + converter; ESP32-C3-MINI-1 (C2838502, pre-certified module) as MCU + ESP-NOW radio.
• Numbers: rated input 957 W, standby 0.83 W (target 0.5 W missed, threshold 2 W met), continuous metering down to < 1 ml/min.
• Risks: brushed pump life (REL-01), higher standby, +4-6 EUR cost (ASSUMPTION).

Hardware protections without firmware (for SAFE)

• Heater stuck on / firmware hung: bimetal manual-reset cut-out on the exchanger in series with the heater, plus one-shot thermal link at a cooler mounting point, plus safety relay dropped by a hardware watchdog (retriggerable monostable) - three independent breaks.
• Overcurrent: mains fuse T6.3A (EU, 4.5 A heater) / T12A (US, 8.4 A heater) in the IEC C14 fused inlet; varistor + X2 capacitor for transients; NTC not needed (resistive heater, zero-cross switching limits inrush).
• Pump with heater cold: pump triac gated by an AND of the MCU output and a hardware READY comparator on the exchanger sensor.

<!-- board/ideas/MECH.md -->

Ideas - MECH (stand-in; numbers: deliverables/02-concept/calc/mech_concept.py, facts mech.*)

Physics of my domains (handbook section 4)

F-MECH-1 - high-mass cast-Al block (heat-storing), horizontal layout

• Principle: 0.55 kg A380 block, cast-in heater and 316 coil, 20 mm mineral wool, stainless shroud.
• Numbers at 950 W: warm-up 2.61 min, continuous 20.8 ml/min (worst fluid), peak 17.4 ml, hold loss 49 W, 2.62 kg.
• Needs: heater about 1000 W for the ZAD-PRF-02 target; PCB bay 150 x 110 x 40 mm above the tank.

F-MECH-2 - low-mass extruded exchanger (power-limited)

• Principle: 0.25 kg extrusion, 2 cartridge heaters, brazed tube.
• Numbers: warm-up 1.25 min, continuous 21.1 ml/min, peak 13.9 ml, hold loss 39 W.
• Risks: spitting on slugs (290 J/K), faster coking.

<!-- board/ideas/SAFE.md -->

SAFE - hazard list (ISO 12100 manner) and regulatory map - FRAME

Scale: S severity 1 first aid, 2 reversible injury, 3 serious/irreversible, 4 fatal or fire of the venue. P probability 1 improbable, 2 remote, 3 occasional, 4 frequent. R = S x P before -> after measures. Measures in the order: D = inherently safe design, G = protective device/guard, I = information for use.

Regulatory map (SOURCE where read; editions to be confirmed by the lab at test time)

• EU: radio remote -> RED 2014/53/EU (Art 3.1a safety via LVD objectives, 3.1b EMC, 3.2 spectrum; LVD/EMCD then not applied separately); RoHS 2011/65/EU + 2015/863; WEEE 2012/19/EU; REACH 1907/2006; GPSR (EU) 2023/988 (consumer sales, Kickstarter/web shop); Batteries (EU) 2023/1542 (remote); Ecodesign standby (EU) 2023/826 (applicability to check); PPWR (EU) 2025/40. Machinery Regulation (EU) 2023/1230: assessed not applicable (only an enclosed fluid pump moves) - legal confirmation BLOCKED for a human.
• EU standards: EN 60335-1 (+A's; cl. 11 heating, 13 leakage, 16 electric strength, 19 abnormal operation, 20 stability, 22 construction, 27 earthing, 29 creepage/clearance, 30 heat and fire), EN 62233, EN 55014-1/-2, EN IEC 61000-3-2, EN 61000-3-3, ETSI EN 301 489-1 + -17 (2.4 GHz) or -3 (868 MHz), ETSI EN 300 328 or EN 300 220-2, EN 62479, EN IEC 63000; remote handset EN IEC 62368-1. Route: Module A (RED Annex II) self- declaration, no notified body when harmonised standards are applied in full.
• US/Canada: FCC 47 CFR Part 15 Subpart B Class B (SDoC) for the host, Part 15 Subpart C (15.247) via a pre-certified module with FCC ID; NRTL listing (ETL/UL/CSA) to ANSI E1.29-2009 (R2024) - product safety standard for theatrical fog generators, based on ANSI/UL 998 (SOURCE: https://tsp.esta.org/tsp/documents/published_docs.php, read 2026-10-30); Canada: CSA certification legally required (SOURCE: https://www.ontario.ca/document/safety-guidelines-live-performance-industry/fog-and-haze-effects, read 2026-10-30), ICES-003 Class B, RSS-Gen + RSS-247, bilingual labels; California Prop 65; 16 CFR 1263 if coin cells.
• ESTA/ANSI: E1.11 (DMX512-A), E1.20 (RDM, optional), E1.27-1 (cables), E1.5-2024 (fluids), E1.14-2023 (fog equipment manuals), E1.23-2023 (atmospheric effects), E1.29 (product safety) (SOURCE: https://asepo.org/FogSafety, read 2026-10-30).

Derived safety requirements (sent to REQ as CR)

DR-SAF-01 heater protection independent of firmware: HW over-temp cut on 2nd sensor + series relay, plus non- self-resetting thermal fuse (no reliance on software -> no EN 60335-1 Annex R class B software). DR-SAF-02 pump supply hardware-interlocked to exchanger >= minimum vaporisation temperature. DR-SAF-03 external window watchdog removes heater and pump enable <= 1 s after MCU hang. DR-SAF-04 fluid path and tank separated from mains electronics (drip barrier, no joints above PCBs). DR-SAF-05 reinforced insulation mains-SELV: creepage >= 5.0 mm, clearance >= 3.0 mm, 3000 V a.c. hipot. DR-SAF-06 line/neutral-to-PE Y capacitance <= safe.y_cap_max_nF; heater hot leakage spec <= 0.05 mA. DR-SAF-07 heater switched zero-cross only (no phase angle); switching rate keeps Pst <= 1 (EN 61000-3-3). DR-SAF-08 insulating parts carrying > 0.2 A: GWFI >= 850 degC, GWIT >= 775 degC; metal enclosure around exchanger. DR-SAF-09 exchanger/fluid path withstand >= 2x pump dead-head pressure or relief to tank (blocked outlet). DR-SAF-10 truss bracket + secondary safety-cable eye, static proof load 10x unit mass. DR-SAF-11 remote on AAA cells (no coin cell). DR-SAF-12 normal exchanger temperature within the fluid makers' limit; thermal fuse below the lowest fluid autoignition temperature. DR-SAF-13 tilt sensor stops pump outside the allowed operating angles (ZAD-FUN-08). DR-SAF-14 minimum distances in the manual derived from a plume-temperature test.

<!-- board/concepts.md -->

System alternatives (ARCH, CONVERGE) - morphological table

Note (CLOSE): the tables below are the ballot snapshot (975 W heater, 60 W loss). The current values after the rerun with ELEC/MECH facts are in board/facts/ARCH.yaml and deliverables/02-concept/concept.md (A1 21.9 ml/min, 2.89 min, 20.6 ml).

Numbers: deliverables/02-concept/calc/arch_energy.py -> facts arch.* (CALCULATION on ASSUMPTION inputs: fluid energy arch.e_ml = 2470 J/ml, heater 975 W at 230 V). Fragments: F-ARCH-n in board/ideas/ARCH.md; MECH/ELEC fragments credited where their board/ideas files define them (stand-ins; placeholders noted).

First-order numbers vs top-weighted requirements (230 V unless noted)

Main risks and hazards

• A1: fluid energy per ml uncertain (+/-17 %) -> continuous target margin thin; mains+SELV on one PCB (reinforced insulation).
• A2: spitting at load steps (low thermal mass), heater surface load -> life (ZAD-REL-01), gear pump cost and glycol wear.
• A3: misses ZAD-PRF-01 target, 207 V warm-up > 5 min; output stability under thermostat hysteresis (ZAD-PRF-06); two boards.

Votes of the peer mesh

Blind votes: each voter wrote its ratings before seeing the others'; the tally script applies knock-outs, the safety veto, the requirement weights of the input and a +/-20 % weight sensitivity.

Tally - ARCH-concept

Tallied 2026-10-30T04:50:07+01:00 by method/gates/board_check.py (run by the convener, who never votes). Question: Which system concept do we take into design? Kind: concept; criteria: 58 (5 rated); requirement weights from INPUT/requirements.md; ballot sha 12763e41b09a, CALL M-0004 sha 12763e41b09a (unchanged) Voters: 2/6; missing: MECH, ELEC, SAFE, COST

Result: A1 leads by 1.76 Sensitivity (each criterion weight x0.8 and x1.2, 10 cases): winner holds in every case Contested ratings (spread >= 4 between voters): none Unrated criteria: ZAD-PRF-01, ZAD-PRF-02, ZAD-PRF-03, ZAD-PRF-04, ZAD-PRF-05, ZAD-PRF-06, ZAD-PRF-07, ZAD-PRF-08, ZAD-PRF-09, ZAD-SIZ-01, ZAD-SIZ-02, ZAD-WGT-01, ZAD-PWR-01, ZAD-PWR-02, ZAD-PWR-03, ZAD-PWR-04, ZAD-FUN-01, ZAD-FUN-02, ZAD-FUN-03, ZAD-FUN-04, ZAD-FUN-05, ZAD-FUN-06, ZAD-FUN-08, ZAD-UX-01, ZAD-UX-02, ZAD-UX-03, ZAD-UX-04, ZAD-SAF-01, ZAD-SAF-02, ZAD-SAF-03, ZAD-SAF-04, ZAD-SAF-05, ZAD-SAF-06, ZAD-SAF-07, ZAD-SAF-08, ZAD-CRT-01, ZAD-CRT-02, ZAD-CRT-03, ZAD-CST-01, ZAD-CST-02, ZAD-CST-03, ZAD-CST-04, ZAD-CST-05, ZAD-MFG-01, ZAD-MFG-02, ZAD-REL-01, ZAD-REL-02, ZAD-REL-03, ZAD-REL-04, ZAD-SRV-01, ZAD-SRV-02, ZAD-SRV-03, SAFETY

Votes (revealed after the tally)

REQ

• ZAD-CRT-04: {'A1': 8, 'A2': 6, 'A3': 6} - A1 headline numbers all have CALCULATION facts (arch.a1_warmup_230v, arch.a1_continuous, arch.a1_peak30s, arch.a1_ready_hold); A2/A3 continuous rates exist only in board/concepts.md, no facts yet
• ZAD-MFG-03: {'A1': 8, 'A2': 7, 'A3': 7} - EOL limits of DR-MFG-03 fit all; A2 adds a flow-loss cut-out test, A3 a thermostat check and a second module -> longer test
• RISK: {'A1': 7, 'A2': 4, 'A3': 5} - A1: e_ml +/-17 % thins PRF-02 target margin (arch.e_ml); A2: spit risk at load steps on PRF-07/PRF-03 (w5) + dry-heater runaway + CST-01 +10-20 EUR on cost.unit_cost_q1000=72.9 straddles 85 EUR; A3: PRF-01 4.32 min misses target, 207 V warm-up 5.38 min, thermostat hysteresis vs PRF-06, two boards

ARCH

• ZAD-FUN-07: {'A1': 8, 'A2': 8, 'A3': 8} - all three carry the same display + encoder UI on the logic side (board/ideas/ARCH.md F11); no difference in completeness
• ZAD-UX-05: {'A1': 8, 'A2': 8, 'A3': 7} - A1/A2 one warm-up then fog; A3 warm-up 4.32 min (arch.a3_warmup_230v) and one-way remote pairing step adds handling
• RISK: {'A1': 7, 'A2': 3, 'A3': 5} - A1 proven class, thin margin only on fluid energy (arch.e_ml band 2100-2900); A2 spit risk from C 150 J/K droop + gear-pump wear + heater surface load; A3 thermostat hysteresis vs ZAD-PRF-06 and two boards

Physics map (ARCH, from board/SCAN.md and the owners' tables)

ELEC - electrical architecture, power budget, protection and firmware (concept)

Numbers are quoted from board/facts/ELEC.yaml; their home is deliverables/02-concept/calc/elec_power.py -> elec_power.results.json (run ddpy elec_power.py in that folder). Status words: PASS (analysis) only. Recommended fragment: F-ELEC-1 (lower standby and cost); it implements concept A1 (single main PCB, cast-Al block) chosen by tally M-0012. F-ELEC-2 (closed-loop DC pump) is the fallback for the ZAD-PRF-06 stability risk.

1. Block diagram (F-ELEC-1)

IEC C14 fused inlet (T6.3A EU / T12A US) -> 2-pole mains switch -> X2 cap + MOV + bleeder
   |-- L/N --> bimetal cut-out (manual reset) --> thermal link --> safety relay (HW watchdog) --> BTA16 triac --> HEATER 950 W (PE bonded block)
   |                                                                          \--> pump triac (gated by HW READY comparator) --> AC vibratory PUMP
   |-- IRM-03-5 (3 W, reinforced) --> 5 V --> LDO 3.3 V  (SELV side)
                                     |-- STM32G071 MCU: PI heater loop (PT1000, ratiometric ADC), stroke metering, state machine, IWDG
                                     |-- zero-cross + mains-voltage sense (opto)          |-- MOC3063 x2 (zero-cross opto-triacs, reinforced)
                                     |-- UI board: display + encoder + lit level LED       |-- CC1101 868/915 MHz two-way radio, paired 32-bit ID
                                     |-- DMX: B0505S isolated DC-DC + digital isolator + SN65HVD72, 5-pin + 3-pin XLR in/thru, switchable 120 ohm
                                     |-- level sensor (MECH: capacitive/float at pick-up) ; USB-C service port (FW update, log)

2. Power budget (both variants, all modes)

3. Protection concept (works without firmware)

• Heater stuck on / firmware hung: manual-reset bimetal cut-out on the block + one-shot thermal link + safety relay dropped by a hardware watchdog (retriggerable monostable fed by an MCU toggle): three independent breaks (ZAD-SAF-01, owner SAFE for set points).
• Pump runs while cold: pump triac gate is a hardware AND of the MCU output and a READY comparator on the sensor (ZAD-SAF-07 "no pumping when not READY"); sensor open or short drives the comparator to "not ready".
• Overcurrent: fuse in the inlet; transients: MOV + X2, TVS on DMX lines and USB; ESD: TVS on all user-touchable SELV I/O.
• Inrush: resistive heater switched at zero-cross; burst control in full cycles with a spread pattern to stay inside EN 61000-3-3 flicker limits (ASSUMPTION, pre-compliance test).
• Isolation: reinforced barrier (supply module, opto-triacs, digital isolator, >= 8 mm creepage slot, ASSUMPTION per EN 60335-1 table); DMX also isolated from SELV (ZAD-FUN-01, ZAD-SAF-06). Earth leakage: heater element rated <= 0.25 mA class (SAFE to confirm).
• Safe state on power loss: relay and triacs open (de-energised = off); settings in flash.

4. Firmware architecture

States: OFF -> HEATING -> READY <-> FOG -> COOLDOWN, plus FAULT (latched, code on display/remote/RDM) and CLEAN (guided purge, ZAD-SRV-02).

• Pump-follows-heat (ZAD-PRF-03): pump stroke rate is capped by the measured heater margin (block temperature above the fluid-type minimum); output derates smoothly instead of stopping.
• DMX: 2 channels (level, mode/timer), loss timeout 1.0 s (elec.dmx_loss_stop) unless hold mode; RDM (E1.20) responder for address, status, identify (ZAD-FUN-02).
• Remote: paired unique ID, 4 machines in groups, two-way status (ZAD-FUN-03, -04); continuous mode auto-off.
• Timer: interval 1-600 s, duration 1-120 s, level 1-100 % on unit and remote (ZAD-FUN-05).
• Low fluid: pump stops on level signal before the pick-up uncovers, auto-resume after refill (ZAD-FUN-06).
• Safety: IWDG watchdog, plausibility of sensor (open/short/rate-of-rise), mains-voltage compensation of stroke rate (ZAD-PRF-06), hours and fault log, USB-C bootloader (ZAD-SRV-03).

5. Key parts (LCSC offline catalogue, read 2026-10-30, data 0.7 d old)

6. Radio and EMC

868 MHz (EU) / 915 MHz (US) sub-GHz link: 50 m margin 18.8 dB (elec.radio_margin_50m, CALCULATION with ASSUMPTION inputs) -> ZAD-FUN-03 target PASS (analysis), range test in the test plan. Prefer a pre-certified module (RED / FCC Part 15.247 or 15.249) to avoid a full radio certification. EMC: zero-cross switching limits conducted emissions; X2 + common-mode choke on the SELV supply input.

7. Risks

1. Open-loop stroke metering: ZAD-PRF-06 target 5 % may be missed (threshold 15 % likely met) - mitigation: voltage compensation, EOL calibration; fallback F-ELEC-2 closed-loop DC pump.
2. Flicker of burst-fire heater control at 950 W (EN 61000-3-3) - pre-compliance test.
3. IRM-03-5 stock 66 at LCSC - second source needed (COST).
4. Board size 100 x 70 mm (elec.pcb_size, ASSUMPTION) not yet checked against the MECH bay.

MECH section - concept A1 (high-mass block), first-order

All numbers: CALCULATION in 02-concept/calc/mech_concept.py -> mech_concept.results.json (facts mech.*), inputs ASSUMPTION as labelled there; heater 950 W = elec.heater_power_nom (ASSUMPTION of ELEC).

Layout (mech.envelope 310 x 165 x 185 mm, 9.46 L)

Front bay: insulated exchanger module (block 120x70x50 mm, 20 mm mineral wool, stainless shroud) at the top; nozzle recessed 15 mm behind a slotted guard; vibratory pump on 4 elastomer isolators below it, behind a heat shield. Rear bay: translucent PP tank (internal 140x115x95 mm, lit level window), tool-free vented fill cap at the rear top, recessed carry handle at the rear (more than 150 mm from the outlet), electronics bay 150x110x40 mm above the tank and sealed from it. 1 mm Al U-shell (powder-coated) + PC UL94 V-0 end caps; 4 rubber feet, track 145 mm.

Governing estimates

Materials and processes

q1: 3D-printed PC end caps, laser-cut and bent Al shell, CNC block. q100: same plus sand-cast block. q1000: die-cast A380 block, formed sheet shell, injection-moulded PC V-0 caps, blow-moulded PP tank. Wetted parts: 316 stainless coil, PP tank, FKM seals (glycol-compatible).

Critical purchased parts

Vibratory solenoid pump (ULKA class), tubular heater, 2 thermal cut-outs, FKM seals. Part numbers and prices: lookup NOT RUN in this time box (ASSUMPTION ranges only; COST and ELEC to confirm).

First tasks for detailed design, and risks

1. Spitting physics: coil length against flow-boiling heat flux on a test rig (highest risk, ZAD-PRF-07).
2. Heater 1000 W against the ZAD-PWR-02 target. 3. Enclosure hot spot and vent path (ZAD-SAF-02).
3. Pump noise (ZAD-UX-01). 5. Tank vent valve for side-lying (ZAD-SAF-05).

SAFE section - safety and compliance (concept)

You can find the full hazard list (H01-H23, S x P before and after measures) in board/ideas/SAFE.md. The numbers are facts in board/facts/SAFE.yaml. The leakage and flicker numbers come from 02-concept/calc/safe_leakage.py, whose results are in safe_leakage.results.json. This section identifies and plans compliance work. It does not certify anything.

Top hazards and measures (inherently safe design -> protective device -> information)

The following protections work without the firmware. As a result, the design does not need EN 60335-1 Annex R (class B software).

• Thermal fuse.
• Hardware over-temperature relay drop.
• Hardware pump-enable temperature interlock.
• External window watchdog.
• Mains fuse.
• PE bonding.
• Reinforced isolation barrier.
• Nozzle guard.
• Fluid/electronics separation.

Residual risks are carried by the information for use:

• Fog setting off smoke detectors.
• Slippery residue.
• Aerosol exposure in small rooms.
• Misuse with unsuitable fluids.

Key numbers:

• Leakage: 0.226 mA EU and 0.155 mA US, against the 0.25 mA target (safe.leakage_eu, CALCULATION).
• Line/neutral-to-PE capacitance: at most 2.5 nF.
• Heater step: d = 0.76 % (CALCULATION). Zero-cross burst switching with a slow period keeps Pst <= 1.

Regulatory route

EU (CE): the remote is radio equipment, so RED 2014/53/EU applies. Its Article 3.1a covers safety (EN 60335-1 clauses 11, 13, 16, 19, 20, 22, 27, 29 and 30, plus EN 62233). Article 3.1b covers EMC (EN 55014-1/-2, EN IEC 61000-3-2, EN 61000-3-3, ETSI EN 301 489-1/-17). Article 3.2 covers the spectrum (ETSI EN 300 328 for 2.4 GHz; EN 300 220-2 if 868 MHz is chosen). Further standards: EN 62479, and EN IEC 62368-1 for the remote. Other EU rules that apply:

• RoHS (EN IEC 63000).
• WEEE registration per country.
• REACH.
• GPSR 2023/988.
• Batteries Regulation 2023/1542.
• Ecodesign 2023/826, applicability to be checked.
• PPWR 2025/40.

Conformity route: Module A internal production control with a lab test report, a technical file and an EU DoC. No notified body is needed.

US/Canada:

• FCC Part 15B Class B by SDoC for the host.
• The radio uses a pre-certified module under Part 15.247 (FCC ID / IC ID). The remote needs its own grant unless it is built on a certified module.
• Canada: ICES-003 Class B, RSS-Gen and RSS-247.
• Safety listing: NRTL listing (cETLus/cULus) to ANSI E1.29-2009 (R2024), which is based on UL 998. SOURCE: tsp.esta.org published documents, read 2026-10-30. The Canadian counterpart is to be named by the NRTL.
• Prop 65: specify DEHP-free cords and lead-free fittings.
• Manuals and labels in English and French.

ESTA (ZAD-CRT-03):

• E1.11 for the DMX512-A port.
• E1.20 for RDM (optional).
• E1.5 for approved fluids.
• E1.14 for the manual.
• E1.23 for the user guidance.

What an accredited lab would test (rough cost and duration, ASSUMPTION: no quotes requested)

Total: about EUR 40-65k (safe.compliance_cost = 55k) over 10-16 weeks if EU and US run in parallel. That is roughly EUR 55 per unit over the first 1,000 units, which is a cost risk for COST. Pre-compliance before the lab: thermocouple heating run, flickermeter, conducted-emissions scan, and a single-fault bench on the hardware cut-outs.

Open

• Legal confirmation that Machinery Regulation 2023/1230 does not apply (needs a human: BLOCKED).
• Fluid autoignition and decomposition limits from the fluid makers' SDS.
• Plume-temperature test to set the minimum distances in the manual.
• The derived safety requirements DR-SAF-01..14 are in board/ideas/SAFE.md and were sent to REQ.