VIC design session v7: reasoning

Lab notebook — VIC sweep, single 15.9mm/22.3mm×101.6mm cell, 1.31nF, water R 87.6kΩ.

The sweep confirms what I'd expect from the series-RLC picture: tank Q tracks coil AC resistance almost linearly at fixed L (candidates #1-3, same 1mH choke, Q falls 62.4→59.9→54.9 as wire gauge thins and R climbs 1.2→2.5Ω), and that Q multiplies straight into cell voltage (748V→659V). But going to 5mH chokes (#4-5) drops f_r from 98.3kHz to 44kHz, and even though coil Q is far higher there (1377 vs 518), tank Q collapses to ~31 because Xl/Xc scale with L while Rp's series-transformed contribution stays roughly fixed — so the water's own dissipation dominates the tank Q at lower frequency, not the coil losses. That's the real trade-off: bigger inductors buy quieter coils but starve the tank of reactance ratio, cutting voltage gain almost in half.

Current is not the binding constraint anywhere in the top five — all sit 130-606mA, comfortably under 1A, so I'm not sacrificing voltage for current headroom here.

Adopting #1: AWG24, 1mH chokes, 98.33kHz, 748V from 12V, 605mA. Best voltage gain of the set with legitimate margin on the current ceiling, and it stays in the operating band Stan specifies for step-charging.

Next session: hold L at 1mH, sweep AWG20-22 for lower R and see if tank Q pushes past 62 before current approaches the 1A limit; also test multi-cell parallel configs to see how Cd scaling shifts f_r and whether gain holds.