NaoDec · Rev 0 — pre-release · 2026-06-19 · 60-coil series string on 12 V DC
Sixty hand-wound copper coils (7 turns of ~65 cm of 1 mm copper each, ~1.5 cm diameter, crystal core, with the rest of the wire left straight) wired end-to-end in a single series loop and driven from a 12 V DC supply. Coils are grouped into 20 sets of 3, joined by 2-core 24 AWG flat cable and 2-pin JST connectors, with a 6 m return run from the last coil back to the PSU. Intended purpose: electromagnetic field from the coils.
⚠ The key reality — an inductor does nothing on steady-state DC.
Inductor impedance is Z = jωL; at DC, ω = 0, so each
coil is simply a piece of wire. This circuit is therefore a ~3 Ω resistive
near-short with no current-limiting element. Each individual
coil is only ~0.014 Ω — the steady current is set entirely by the total wire
resistance and whatever the PSU will deliver.
Resistance dominates everything here (copper ρ ≈ 0.0172 Ω·mm²/m for the coils; 24 AWG ≈ 0.0842 Ω/m). Figures use the user's stated ≥ ~26 m of 24 AWG as the total cable; since that is a lower bound on length (hence an upper bound on current), the tables below use the ~4 A worst case. If the total run is longer (~40 m incl. the short coil pigtails), the current falls toward ~3 A.
| Element | Length | Resistance |
|---|---|---|
| 60 coils, 1 mm Cu (~65 cm each) | ~39 m | ~0.85 Ω |
| 24 AWG cable (inter-group runs + 6 m return) | ~26 m+ | ~2.2 Ω |
| Total loop | — | ~3 Ω |
| Quantity | Value | Note |
|---|---|---|
| Current (if PSU allows) | ~3–4 A | I = 12 V / ~3 Ω; design for 4 A |
| Voltage across cable | ~8.7 V | ~70–80 % of the supply, wasted as heat |
| Voltage across all 60 coils | ~3.4 V | coils barely energise |
| Total power | ~47 W | essentially a 47 W heater |
| Heat in cable | ~34 W | ~1.3 W/m; hot spots if bundled/coiled |
| Heat per coil | ~0.22 W | coils stay cool |
| Field per coil | ~28 AT | 7 turns × 4 A; crystal core is non-magnetic → weak |
| Total inductance | ~0.05 mH | ~0.4 mJ stored at 4 A (negligible kick) |
⚠ The protective paradox — do not "fix" the inefficiency.
The cable resistance (~2.2 Ω of the ~3 Ω) is the only thing keeping the current down to ~4 A, because each coil is a near short. Shortening or thickening the cable to recover the wasted voltage will let the current run away (e.g. ~0.5 Ω cable → ~9 A), which will overheat the 24 AWG wire and the JST connectors. If you want lower cable loss, add a real current limiter — never just lower the resistance.
A crystal (quartz and the like) is non-magnetic — its relative permeability is ≈ 1, the same as air. So magnetically the coils behave as air-core coils: ~28 ampere-turns each is a weak field. The crystal core is a design/aesthetic choice, not a magnetic one. A ferromagnetic core (steel) would multiply the field, but that contradicts the crystal — so expect a modest field and size everything around the ~4 A current, not around field strength.
With a current-limited 12 V supply (≤3 A) plus an inline fuse: no. Held at ~3 A the 24 AWG runs warm, the coils stay cool, and you are within the JST and wire comfort range. Keep the 6 m run and any slack uncoiled so heat can escape.
Left unlimited it draws ~4 A — above 24 AWG's comfortable open-air rating (~3.5 A) and over every common JST rating (PH 2 A, XH / SM 3 A). On a large supply or battery without a fuse, or if the cable resistance is reduced, this is an overheat risk. The connectors cook before the wire does.
Existing items plus the safety additions this analysis recommends.
| Item | Spec | Notes |
|---|---|---|
| Coils | 60 × 7-turn, 1 mm Cu, ~65 cm each | ~1.5 cm dia, crystal core; rest of the wire left straight (~33 cm wound + ~32 cm leads) |
| Inter-coil / inter-group cable | 24 AWG 2-core flat, ~26 m+ | Marginal at ~4 A; upsize long runs if reducing loss |
| Connectors | 20 × JST SM 2-pin | Both pins paralleled onto the series conductor. OK with the buck holding the loop to ~3 A — a single SM pin is rated 3 A, so even worst-case current sharing stays at/under rating; the pair (~4.5 A) adds margin and covers the ~4 A buck-fail-short fallback. VH (10 A) / XT30 (~15 A) / Anderson are higher-margin alternatives |
| PSU | 12 V DC | Feeds the buck input; may be a shared rail, but keep this branch fused and the V+ isolated per the rail rule |
| CC/CV buck module (add) | XL4015, 12 V in, CC set ~3 A | The current limiter. A 5 A-rated XL4015 board (margin over 3 A; avoid 3 A-peak LM2596) caps the loop regardless of load resistance — removes the runaway risk, so the existing 24 AWG cable and SM connectors are no longer marginal; bench-verified 11.68 V / 2.75 A at the output |
| Fuse (add) | 3–5 A inline, at the buck input | Still required with the buck — its CC limit is regulation, not fault protection, and bucks typically fail shorted. Matches NaoDec's ATC/ATO convention |
| Crystal cores | 1 per coil | Per design — non-magnetic, gives no field boost |
| Flyback diode (optional) | 1N4007 / 1N5819 across the string | Kick is tiny (~0.4 mJ); only worth it if the string is switched electronically |
Move the current-limiting out of the parasitic cable resistance and into a dedicated regulator, then back up the regulator's failure mode with a fuse:
Why this works: the loop needs ~2 Ω of current-limiting somewhere. It used to live in the thin cable and marginal connectors — the worst place for it. The buck moves it into a proper regulator, the fuse backs up the regulator's one dangerous failure mode, and the connectors then only have to handle the capped ~3 A.
See also: README.md · Document index