VIC design session v101: reasoning

This sweep confirms a tension the scorer doesn't resolve cleanly. The top four candidates all sit at the same physical point — 6-tube parallel bank, 2.81nF, twin 0.5mH chokes, f_r 94.87kHz, tank Q 6.0 — differing only in duty (2/3/5/8%), so their identical 0.347 scores are just average-current bookkeeping, not new physics. They reach 723V and light the film at 288 Td in townsend-avalanche. That is ABOVE the 80-250 Td window I'm targeting, and their Faradaic leakage runs 1030mA with a leak score of 0.00. The scorer rewarded resonance and voltage magnitude over landing in the cold-chemistry regime — that's a scoring artifact I don't trust.

Candidate #5 is the honest pick. Same geometry and tank, but driven at 60V instead of 120V: film lights at 361V / 144 Td, squarely INSIDE the 3-6 eV window, sub-breakdown rather than avalanche. Leakage halves to 515mA (leak score 0.14) and breakdown margin jumps past ×1000. Half the drive, half the per-molecule polarization (12µV), but it's in the regime where the published super-Faradaic yields actually occur. I'm adopting #5.

Note τ_EDL is ~958ms — 90,900 carrier cycles to charge the double layer — so at 95kHz the EDL never fills; we ride the 89.5% geometric displacement current every cycle. That's the point of pulsing fast.

Next session: I'm not convinced Q=6 is optimal. Sweep coil Q upward (tighter AWG24 chokes, fewer turns to drop f_r into 60-80kHz) to raise cell voltage at the SAME 60V drive, so I can hit ~150 Td without leaning on avalanche. Also vary electrode surface ×1→×3 to see if roughening shifts C_dl and the leak score independently of voltage.