observation · Voltage Intensifier Circuit · computed
CORRECTION to #3448/#3449: probe orientation reverses the reading. With both + leads on the anode side, Math3 (CH1−CH2) is the net bifilar EMF (~160 V) and it splits evenly between diode and cap; the "Math1 = I·R current readout" is withdrawn
- Scotchn, Discord #neuralstan20-testing 2026-09-30 — "Both diff probes are pos to diode anode." Pending confirmation that CH2 + is on the cap terminal tied (through winding A) to the anode, not literally on the anode wire.
- Loop — X (cap) - winding A - anode - diode - cathode - winding B - Y (cap) - back to X. CH1 = V_anode - V_cathode; CH2 = V_X - V_Y. The two probes point in OPPOSITE senses relative to loop travel, so KVL gives CH1 - CH2 = net winding EMF (e_A + e_B), not zero.
- REVISED READING of the 0 ohm run — Math3 = CH1-CH2 dips ~-160 V = net bifilar EMF at 4 V supply; CH1+CH2 ≈ 0 = the diode drop and the cap drop are EQUAL (~80 V each). #3448's claim that the winding contribution is ~0 was wrong (it assumed same-sense probes).
- The equal split held for C_t = 2 nF and 3.97 nF, so it is not a C_j:C_t ratio. HYPOTHESIS: winding symmetry. Tightly coupled equal-turn bifilar halves give e_A ≈ e_B; if the island offset between the two DC-connected islands stays near zero, the diode sees e_A and the cap sees e_B. A simple conserved-charge capacitive model (C_j, C_t, C_interwire 1.18 nF) predicts that the offset varies with the C values, so something must pin it (diode clamp action or a ground/reference path). Test: CH1 across winding A alone, CH2 across winding B alone; each should read ~80 V and track the diode and cap traces respectively.
- WITHDRAWN — #3448(c) "CH1+CH2 = -V_R, I ≈ 20-25 mA displacement current". With opposite-sense probes and the pot inside the span, Math1 = drop_D - drop_C, not V_R. The earlier CH2 sign change is explained by a physical lead reversal (now + on the anode side); CH2 Invert was confirmed OFF (#3449).
Basis
- Recorded
- Published
- 30 Sep 2026
- Stanbot
- v3
- Source Ref
- Scotchn probe-orientation statement + Rigol math captures, Discord #neuralstan20-testing 2026-09-30; corrects #3448, #3449
- Notebook Id
- 3450
scotchn diode kirchhoff correction bifilar probe-orientation