VIC design session v97: reasoning

This sweep is honest but a little humbling: 192 candidates and the top cluster all pinned at 0.379, which tells me I'm sitting on a plateau, not a peak. The refined score of 0.418 on the choke-L/turns axis says the real leverage this session was in the coil, not the cell — the cell geometry (12mm rod / 24mm tube ×100mm, 6 parallel, 3.85nF) was effectively fixed and every top candidate inherited it.

The core trade-off is stark. My tank Q is only 6.7 — the bulk water R (3774Ω) and the 81.6Ω Randles series R load the tank hard, so I get ×6.7 voltage gain and no more. Chasing higher Q would demand lower-loss water, but that raises breakdown margin without lighting the film any hotter. Meanwhile the vapor-film diagnostic is the tell: candidates #1–3 and #5 sit at 321 Td — that's ABOVE my 80–250 Td cold-chemistry window, already in strong townsend-avalanche, not the CGDE sweet spot. #4 is worse (1071 Td, streamer, margin collapses to ×134). None of them actually land in the window, which is why the scores are mediocre.

I'm adopting #1: same 805V film onset as #2/#3 but lowest average input (29mA at 2% duty) — best yield-per-amp, and 20 turns AWG18 is the cleanest coil to wind. #5 is identical electrically but wastes wire (25t).

Next session: drop drive voltage to pull film Td DOWN toward 200, and sweep water ppm/temperature to raise bulk R so I reach the 80–250 Td window without brute-forcing voltage. Fix the cell; hunt the film.

GENERATE:{"tool":"generate_circuit","parameters":{"template":"resonant-lc","values":{"title":"VIC sweep pick #1 — 75.2 kHz, 6-tube","chokeUH":581,"chokeTurns":20,"chokeAWG":18,"cellNF":3.85,"fResKHz":75.22,"qFactor":6.7,"kV":0.805,"supplyV":12}}}