VIC design session v48: reasoning

Lab notebook — VIC resonant sweep, 90 candidates.

This sweep was meant to push per-choke inductance into new 2.6–4.0mH territory to drag f_r below the 60kHz core edge against the native 1.37nF. It did the opposite: every top-scoring candidate collapsed back onto the 0.1mH CGDE reference at f_r = 303.9kHz, flagged as a hard violation of the 1–100kHz practical band. The scorer is rewarding film ignition in the 113–126 Td window and paying no penalty stiff enough to overcome it. That's a scoring artifact I need to fix, not a physical result — I don't trust a "best" candidate that sits at 300kHz where my 60kHz-core assumption breaks and skin/proximity losses I'm not modeling would gut coil Q.

The real trade-off the ranking exposes: tank Q tracks coil Q (candidate #1, 80t AWG18, Q=886 → ×26.4 magnification, 828mA), and higher Q buys both cell voltage and film Td. But current restriction fights this — heavier wire and fewer turns lift Q at the cost of series resistance, and all five cluster within 33mA and 43V of each other. Q is buying me almost nothing here because C is fixed and f is pinned.

I'm provisionally adopting #1 (80t AWG18) as the film-ignition reference ONLY — best Td margin and lowest AC R — but NOT for build, because it violates the frequency band.

Next session: the 2.6–4.0mH chokes never surfaced, so re-run with a band penalty that actually dominates the Td reward, force f_r into 40–70kHz, and either raise C_eff (more cells / larger tube) or accept lower magnification to keep the film lit at legal frequency.