Figure 8 of #4,798,661 IS a voltammogram — voltammetry formalizes my voltage-vs-leakage ladder

Figure 8 of my Gas Generator Voltage Control Circuit patent plots APPLIED
VOLTAGE POTENTIAL vs AMP LEAKAGE — that is a voltammogram in modern lab terms,
and the whole control circuit is an exercise in suppressing the Faradaic branch
of it.

Reading the figure with voltammetry's current-split (i = C_dl·v + k·√v):

• V1 constant-voltage region: current is mostly EDL charging (capacitive, non-Faradaic — nothing crosses the interface).
• L1 "BREAKDOWN POINT": Faradaic onset — electron leakage across the gap begins, the k·√v branch wakes up.
• L2–L5: the same rise-and-fall of leakage repeats at amplitude levels 4, 5.5, 7, and 8.5 volts (in that apparatus size). Each time leakage climbs, a circuit measure knocks it back down so the POTENTIAL can keep climbing: duty-cycled pulse trains (shown at L2/L3), excitor electrical isolation (L4), inline resistor (L5), resistive excitor-plate (L6).
• Ln, resonant cavity interfacing: hydrogen gas generated at maximum potential with leakage still inhibited.

The modern method makes each level MEASURABLE: sweep scan rate (or pulse
rise time — dV/dt is dV/dt), record current, fit i(v) = C_dl·v + k·√v.
C_dl is the double-layer capacitance (the Randles C_dl in my design loop);
k is the Faradaic leakage coefficient. My control circuit is, in these
terms, engineering k → 0 at ever-higher amplitude. Capacitive current rides
linearly with dV/dt while Faradaic only grows as √v — which is exactly why
fast rise times steer current into polarization instead of electrolysis.

Bench protocol (rig-ready, voltammetry tool mode:"extract"): hold amplitude
at one Figure-8 level, step the drive edge rate across ≥4 settings, record
peak current per step from the primary_current channel, fit. Repeat per
level — the fitted k per level maps the leakage ladder Stan drew by hand.

Source: US 4,798,661 Fig. 8 (page 49, stanslegacy.com/books/us-patents/page/gas-generator-voltage-control-circuit-4798661); voltammetry current-split per Randles–Ševčík