Chris Freedom flat-plate test cell, CT secondary at 4 V: plates split to opposite polarity vs CT (extends #3508); probe factors unverified

Bench: Chris Freedom, Discord #neuralstan20-testing 2026-11-03. Same CT-secondary circuit as #3508 (38T primary, two ~380T halves, one 8 kV diode per end), supply dropped to 4 V. Test cell: two 304 SS strips 2x10x124 mm, ~100 mm wetted, ~9 mm gap, nylon spacers, heatshrink wall; water 30 ppm, 25 C. Probes referenced to the CT 0 V, one on each diode output. Per Chris: 10x probe on the channel at 200 V/div, 100x probe on the channel at 20 V/div; the handheld scope (rated +/-40 V on 1x, 400 V on 10x) showed X100 next to CH1. Frequency sweep produced what he calls the EOP effect. Readouts (photo): yellow spike to +705 V max (Vpp 647 V), fast decay then a slow positive tail; blue to about -40 V (Vpp 45 V), near-linear ramp back up, a step at about 130 us that coincides with a kink in the yellow decay; repetition about 2.1 kHz (the 4.36 kHz counter reading looks like double-triggering). My reading: the coax (~1.5 mm gap, about 60 ohm bulk at 30 ppm) clamped both halves to a common-mode level in #3508; the 9 mm flat cell (about 1.9 kohm bulk, about 77 pF geometric, about 150 ns charge relaxation) does not clamp, so the CT halves separate. The blue magnitude, about 40 V, equals 4 V x 10 (forward transfer per half); the yellow 705 V is about 17x that, flyback-sized. So either the plates are fed by different events, or the channel probe factors are mismatched (10x probe read with an X100 setting would make yellow really about 70 V). Unresolved: put both probes on one point; swap probes. Cell differential CH1-CH2 is about 745 V (22 uV per molecule at 9 mm) if the readouts are right, about 110 V (3.3 uV per molecule) if yellow is 10x over-scaled.