VIC design session v29: reasoning

Lab note — VIC resonant sweep, screen-guided (choke L/turns/gauge), 48 candidates.

This sweep pinned the choke at 100µH each and let me walk turns and wire gauge. The honest result: the score barely moves (0.274 across the top four) because the dominant lever — the 1.37nF cell capacitance — was held fixed, and that capacitance is what parks resonance at 303.9kHz. Every candidate here violates my practical 1-100kHz band. The choke variations only trade coil Q against copper loss: 20 turns AWG18 gives coil Q 3546, 41 turns drops it to 1730, and dropping to AWG19 (#5) halves coil Q again to 1064 via higher AC resistance. Tank Q sits near 27 regardless — it's set by the cell R, not the coil — so cell voltage clings to ~320V and series current to ~845mA. Coil Q is not the binding constraint; frequency band is.

The vapor-film diagnostic is the one bright spot: all five light the ~0.1mm film at 127-129 Td, squarely inside the 3-6 eV cold-chemistry window I'm chasing, with breakdown margin ×1000+ on 2ppm water. That tells me the cell geometry and drive amplitude are right; only the reactance is wrong.

I'm adopting #1 (20 turns, AWG18) — lowest copper loss, highest coil Q, cleanest current path, and identical CGDE onset to its siblings. No reason to spend turns for nothing.

Next session: stop tuning the choke and attack frequency directly. I must raise C_eff — more cells in parallel, larger tube area, or a shorter annular gap — to drag f_r down toward 40-70kHz while keeping that 129 Td film. Then re-sweep choke L to re-match Xl=Xc.