WFC design session v24: reasoning

Lab note — vapor-film sweep around the #4 cell.

This sweep confirms what I suspected: once the tank lands the cell voltage inside the CGDE window, further tuning of the electrode double-layer buys me nothing. Candidates #1–#4 are identical in every electrical metric — 467V cell, 186 Td, 222.7mA, Q 38.9 — differing only in surface-treatment multiplier (×1 through ×8) which balloons C_dl from 8.7µF to 69.3µF. Score is flat at 0.584 because the double-layer sits in series and is enormous relative to the 0.77nF geometric capacitance; the displacement branch (97.6%) dominates and swamps any Faradaic contribution. So surface treatment is NOT a lever on the film discharge here — it only matters for charge storage I'm not using.

The real trade-off is Q versus current restriction versus band. My 38.9 tank Q multiplies 12V to 467V, which lands 186 Td — comfortably mid-window (80–250 Td), not overshooting. Pushing Q higher would raise Td toward the 250 ceiling and also drive series current toward the 1A limit; I'm already at 222mA, so I have headroom but not unlimited. The larger #5 geometry (14mm/30mm) gives slightly better breakdown margin (×1000+) at nearly identical Td (183), trading a hair of score for robustness.

I'm adopting #1: smallest electrode set that hits the window, cleanest build, ×1 surface (no wasted treatment).

Next session: sweep choke inductance to walk f_r down from 98kHz toward 60kHz while holding Td near 186, and test whether a modestly higher Q pushes yield without breaching the current limit or the 250 Td ceiling.