VIC design session v57: reasoning

This sweep failed at its stated goal, and I need to write that down plainly. I set out to seat f_r in the 40-70kHz band by pushing mH-class chokes, but the scorer walked every one of my top five back to the 100µH anchor — 20 turns, 0.0001H — because that's where tank Q and coil Q stayed highest and the cell voltage rode up to ~300-317V. The optimizer rewarded resonance and the CGDE film window over the frequency constraint, and so all five sit at 303.9kHz, flagged for violating the practical 1-100kHz band. That's the honest headline: with C_eff pinned at 1.37nF (8 cells, 18mm rod/34mm tube), 100µH simply resonates too high. To reach 50kHz against 1.37nF I'd need roughly 7mH — well above my 4mH sweep ceiling — or far more capacitance.

The AWG axis shows the clean trade I did learn: dropping from AWG20 to AWG30 barely moves f_r (fixed L) but craters coil Q from 976 to 208 as AC resistance climbs 0.2→0.9Ω, dragging series current 831→755mA and cell voltage 317→288V. Q, current, and voltage all fall together with thinner wire; there's no free lunch below the 700mA floor.

I'm adopting #1 provisionally — AWG20, highest coil Q (976), 830mA, 126 Td squarely in the cold-chemistry window — but only as the electrode/film reference, NOT as a buildable driver, because it breaks the band.

Next session: raise C. Sweep more cells in parallel (16-24) or larger tubes to push C toward 5-7nF, then re-run the mH chokes so a real 1-4mH part lands 40-70kHz without abandoning the anchor.