VIC design session v16: reasoning

Lab note, 7/9/26 — L-decade-down sweep, gap-shortened C_eff

Pushed L to 0.1mH hoping the shortened gap capacitance would finally drag f_r into the 40-70kHz band. It didn't — all 24 candidates lock at 326.8kHz, roughly 5x too high, because C_eff barely moved (1.19nF) once I shortened the gap; L and C are both now too small together, and f_r = 1/(2π√LC) doesn't care which one I shrink if the ratio stays wrong. This confirms the earlier finding: I need to shrink L by pushing C_eff UP (wider gap or longer tube), not by cutting both down in the same direction.

Within this outside-band batch, the trade-off is exactly what I'd expect from the choke topology in my notebook (chokes as current limiters, series with the water capacitor): thinner AWG raises coil resistance, which craters coil Q (669→446) while tank Q barely drops (28.0→27.4) since tank Q is capped by the 14.1Ω Randles cell series R, not the coil. Series current stays pinned near 800-820mA regardless — the ionic branch is doing the limiting, not the choke. Breakdown margin actually improves slightly with lower coil Q (356→363), a minor consolation.

Adopting #1 provisionally only as the low-resistance/high-Q reference point (AWG 20, coil Q 669.6) since it's the best "current-limiting choke" instance, though it fails the band constraint and shouldn't run.

Next session: hold L fixed at 0.1mH, sweep tube diameter/length upward to inflate C_eff directly until f_r crosses into 40-70kHz — stop shrinking both L and C together.