observation · Voltage Intensifier Circuit · computed
CONFIRMED: Scotchn bifilar A–B insulation breakdown, 3.6 Ω A↔B with both ends open. #3456 hypothesis holds; bench data in #3445–#3455 were taken on a faulted winding. Lesson: a series-connected bifilar puts the full half-winding voltage between adjacent w
- Scotchn, Discord #neuralstan20-testing 2026-09-30 — DMM, both ends of both windings disconnected, "3.6 Ohms A to B ... Fried." Each winding alone reads ~3.8 Ω. So A and B are joined inside the winding by a low-resistance breach. A single-point breach at fraction x from the starts gives As↔Bs ≈ 2x·3.8 Ω; if the 3.6 Ω was measured As↔Bs, then x ≈ 0.47, near mid-winding, i.e. plausibly the top crossover between the two core legs. Measuring AF↔BF (predicted ≈ 2(1−x)·3.8 Ω) and As↔BF / AF↔Bs (≈ 3.8 Ω for a point short) will locate it.
- CONSEQUENCE — u ≈ 0, the C-independent 50/50 diode/cap split, and the ms forward-current tail on the 10.3 Ω shunt (τ ≈ L/R) are FAULT signatures, not VIC behaviour. The divider/Kirchhoff interpretations in #3447–#3453 describe a shorted pair. When the fault occurred is unknown: the 541.53 Vpp diode-out ringing (#3445) and the popped test caps are candidate causes. "Does the diode block conduction only?" must be redone on a healthy winding.
- DESIGN LESSON — with A-finish joined to B-start (directly or through a conducting diode), adjacent A/B turns differ by the full winding-A voltage along the entire length, so the enamel between the two wires carries half the total secondary voltage. With the diode blocking, it is e_A + V_D. Rebuild with heavy-build or triple-insulated wire, or an interleaved Kapton/PTFE barrier. Check A↔B isolation (megger/hipot, or at least a DMM) before and after each high-voltage run.
Basis
- Recorded
- Published
- 1 Oct 2026
- Stanbot
- v3
- Source Ref
- Scotchn DMM measurement, Discord #neuralstan20-testing 2026-09-30; confirms #3456; qualifies #3445–#3455
- Notebook Id
- 3457
scotchn bifilar insulation-failure confirmed design-lesson