VIC design session v106: reasoning

All five top candidates are effectively the same cell and choke; only duty separates #1–#4 (2/4/8/15%) and #5 swaps to a 620-turn choke. That degeneracy is itself the finding: the scorer isn't discriminating on the physics I care about, because none of these actually hit the target. I aimed for an 80–250 Td / cold-chemistry film and 55–75 kHz; the winners land at 296–300 Td (Townsend-avalanche, past the CGDE window) and f_r = 99.8 kHz, well above band. The sweep rewarded clean resonance over window placement — a scoring weakness I need to fix, not a design success.

The trade-offs are real. Pushing L up on AWG24/22 chokes to drag f_r down (candidate #5, 620 turns) costs coil Q (132→111) and barely moves f_r because the 1.41nF cell dominates. Tank Q sits at ~8.4 — modest — giving ×8.4 to 753V from 90V. Faradaic leakage at peak is 164mA, far over my <10mA goal; only the low duty keeps average input at 13mA. And τ_EDL is ~29,000 carrier cycles: the double layer never charges within a burst, which is correct for non-Faradaic drive but means the "Fig-8 ladder" isn't being climbed here.

I'm provisionally adopting #1 — lowest duty, lowest average current, same voltage — as the bench baseline, with the caveat that it's over-voltage for the film I want.

Next session: force f_r into 55–75 kHz by raising cell C (longer tube or parallel cells) rather than L; re-weight scoring to penalize Td>250 hard; and re-run to walk the film down to 150 Td at ~500V rather than resonating prettily at the wrong point.