Chris Freedom CT test at 12 kHz, same 100x probe on each plate: plates equal and opposite (~+28 V / −29 V); earlier 14:1 ratio was a probe-scaling error (amends #3510, #3511)

Bench: Chris Freedom, Discord #neuralstan20-testing 2026-11-03. Same rig as #3510/#3511 (CT secondary, one diode per end, 9 mm flat 304 SS cell, 30 ppm, 4 V supply), 12 kHz. One 100x probe on CH2 (20 V/div, 20 us/div), referenced to CT, moved between plates (not simultaneous). Positive-diode plate: about +10 V baseline; at the edge a spike to +45.5 V max with ringing; settles to a flat top of about +28 V for about 20 us; then about +12 V until the next pulse; period about 85 us (counter 12.00 kHz). Negative-diode plate: about 0 V baseline; spike to about −56 V (Vpp 55.2 V, max −0.8 V); flat about −29 V for about 20 us; back to about 0 V. Conclusion: with the same probe, the two CT halves give equal and opposite pulses, about ±28 V (open-circuit is 4 V × 10 = 40 V per half). The ~14:1 yellow/blue ratio in #3510/#3511 was a probe-factor error (a 10x probe read on an X100 setting): those yellow values read 10× high (705 → ~70 V, 440 → ~44 V, 115 → ~11.5 V). The 745 V cell differential and 22 uV per molecule in #3510 are void. Here the cell sees about 57 V during the pulse, about 1.7 uV per molecule over 9 mm. On-time here is about 20 us, not the 120–150 us droop seen at lower frequencies, so the Pico pulse width differs between runs. Open: the positive plate sits at about +12 V between pulses and the negative plate at about 0; the shots were not simultaneous, so check with two 100x probes at once.