observation · Voltage Intensifier Circuit · computed
Diode in vs jumpered, open pot gap: diode in ~85 V pulses with ringing; diode jumpered only ~8 Vpp sawtooth. As wired now, the two windings CANCEL at the gap when joined (series-opposing/paraphase), contrary to the series-aiding drawing. Check with LCR
- Scotchn, Discord #neuralstan20-testing 2026-09-30. Same setup as #3467 (CH1 100X diff across the open pot gap, ~2 MΩ probe load only, C_t out, 20 V/div, trigger CH3), diode jumpered, nothing else changed. CH1 Vpp 8.11 V: a small sawtooth per pulse (a ramp of a few volts during the on-time, back to ~0 in the off-time). With the diode in (#3467): ~85 V pulses with dense ringing.
- KVL — gap = V_BF − V_As = e_B + u (u = V_Bs − V_As). Diode jumpered forces V_Bs = V_AF, so the gap = e_A + e_B. A near-zero gap means e_B ≈ −e_A, i.e. the joined windings are series-OPPOSING (flux-cancelling, paraphase) as currently wired. With the diode open, the windings float separately and the gap shows ~one winding's EMF (~85 V), with u held near 0 by the interwinding capacitance.
- This contradicts the series-aiding assumption in the #3451 drawing (and the retracted LCR series 140.5 mH ≈ 4L). Either the rewire onto the card (new stubs, headers removed) swapped a start/finish pair, or the drawing's B start/finish was reversed.
- Consequence — in this wiring the diode is not blocking a flyback. Its open state is what lets any gap voltage appear, by decoupling a cancelling pair.
- CHECK — power off, diode jumpered, LCR across the gap: aiding ≈ 4 × 33 mH ≈ 130 mH; opposing = leakage only (≪ 33 mH). Re-identify the start/finish of each winding. Relevant to the open question on Meyer's 56/62 winding sense (#3417).
Basis
- Recorded
- Published
- 1 Oct 2026
- Stanbot
- v3
- Source Ref
- Scotchn Rigol capture (diode jumpered), Discord #neuralstan20-testing 2026-09-30; compare #3467; relates to #3417
- Notebook Id
- 3468
scotchn bifilar winding-sense paraphase diode