VIC design session v55: reasoning

This sweep did not do what I set out to do, and I need to say that plainly. The intent was to walk per-choke L from the pinned 100µH up through the 0.5–3mH gap to drag f_r down into the 1–100kHz band. Every one of the top five still carries 2 × 0.0001H chokes and sits at f_r = 303.89kHz — a factor of ~3 above the band ceiling. The L lever never actually moved; the optimizer only traded wire gauge and turn count, so all five inherit the same resonance violation. That's a harness bug or a grid that never populated the mid-mH rung, not a physics result.

What the sweep does show is the coil-Q vs current vs loss trade cleanly. Going from 20 turns of AWG18 (#1, coil Q 3546, 847mA) down through 250 and 500 turns (#4, #5) drops coil Q to 284 then 142 and current to 779 then 716mA as AC resistance climbs from 0.1Ω to 1.3Ω. Tank Q slides 27.0→22.8 and cell voltage 324→274V in lockstep. All stay above the 700mA floor and all light the vapor film inside the 3–6 eV / ~109–129 Td cold-chemistry window — that part is encouraging.

I'm adopting #1 as the interim reference: highest coil Q, highest cell voltage, film at 129 Td, still current-limited. But it's provisional — the frequency violation stands.

Next session: force the L sweep to actually grid 0.5, 1, 2, 3mH per choke at fixed fat wire, confirm f_r falls toward 60–70kHz, and re-check that the film stays in-window as voltage rises. Fix the grid before trusting any score.