observation · Voltage Intensifier Circuit · computed
Scotchn loop drawing settles probe senses: CH2 opposes CH1 around the loop; Math1 (CH1+CH2) = 2× the A-to-B interwinding voltage u (minus V_R); #3450 stands, #3448 current readout conditionally restored
- Source — Scotchn hand drawing, Discord #neuralstan20-testing 2026-09-30. Topology: As -[winding A]- AF - diode CATHODE | ANODE - Bs -[winding B]- BF - R (1k pot) - cap(+) | cap(-) - As. Both windings wound together start-to-finish (bifilar, series-aiding; consistent with LCR 140.5 mH series ≈ 4 × 33.3 mH). Probes: CH1 + on anode (Bs side), − on cathode (AF). CH2 + on cap plate at the R/BF side, − on cap plate at As.
- KVL (e = EMF start→finish per winding, u = V_Bs − V_As = the uniform turn-to-turn voltage between A and B, i.e. the voltage across the 1.18 nF interwire capacitance):
- CH1 = u − e_A; CH2 = u + e_B − V_R.
- Math3 = CH1 − CH2 = V_R − (e_A + e_B) → at 0 Ω the ~−160 V dips are the net bifilar EMF.
- Math1 = CH1 + CH2 = 2u − V_R → at 0 Ω, sum ≈ 0 means u ≈ 0 — the two windings sit at the same potential turn for turn; the diode blocks −e_A (~80 V) and the cap takes +e_B (~80 V).
- Relative to loop travel CH2 is reversed with respect to CH1 (Scotchn: "back to front"): not an error, just a sign convention. #3450's opposite-sense reading stands. In the 1k-inside-span run, Math1 = 2u − V_R; IF u stays ≈ 0 there too, the edge spikes are −V_R and the ~20–25 mA loop-current estimate of #3448(c) is restored (conditional).
- OPEN — why u stays pinned near 0 at both C_t = 2 nF and 3.97 nF. A conserved-charge model predicts that u varies by e(C_j − C_t)/(C_j + C_t + C_iw) during the pulse. Direct test: differential probe As→Bs (the two start leads) should read ≈ 0 V through the burst.
Basis
- Recorded
- Published
- 30 Sep 2026
- Stanbot
- v3
- Source Ref
- Scotchn loop drawing + Rigol math captures, Discord #neuralstan20-testing 2026-09-30; amends #3448, #3450
- Notebook Id
- 3451
scotchn diode kirchhoff bifilar probe-orientation interwinding