observation · Voltage Intensifier Circuit · computed
Scotchn's fixed LCR meter actually tests at 160.4 Hz, not 1 kHz/100 kHz. Every LCR value from it is a 160 Hz, ~2 V small-signal reading; interwinding C (both windings shorted) = 1.122 nF
- Bench — Scotchn, Discord #neuralstan20-testing 2026-10-05. He scoped his fixed-frequency LCR meter's test signal on the Rigol (CH4, 1 V/div): Freq 160.36 Hz. The wave is square-ish with rounded tops, roughly +2.3 V / -1.7 V (read off a photo, and the load during the capture is not confirmed). So every earlier LCR figure from this meter was taken at about 160 Hz, including the C_t 3.97 nF, C_add 1.864 nF and C_jd 1.336 nF in #3447 (those were logged as "100 kHz"), the bifilar L values in #3476, and the earlier 1.162 nF interwinding figure.
- New reading — discharged, diode and cap lifted, As-Af shorted and Bs-Bf shorted, meter REL'd. Between the two shorted pairs: 1.122 nF, twisted-pair extensions included. That is X_C ≈ 884 kΩ at 160.4 Hz, against about 7.1 kΩ at 20 kHz.
- Consequences:
- (1) C_jd 1.336 nF is very likely not junction capacitance. A test wave of about ±2 V forward-biases a SiC Schottky on one half-cycle. Use the datasheet C-V curve, or measure under DC reverse bias.
- (2) #3476 self-resonance explanation for k = 1.16 — at a 160 Hz test, an 8% excess would need f_srf ≈ 594 Hz, i.e. C_eff ≈ 556 nF across 129.3 mH. That is absurd next to 1.1 nF, so self-resonance is ruled out. Still open: µ of the nanocrystalline core depending on amplitude, the cheap meter's handling of a square wave, or a plain meter error. Test: series-opposing reading, then check L_aid + L_opp against 2(L_A + L_B) = 119.8 mH.
- (3) Calibration without a reference cap — RC step from an Arduino pin through a 1% resistor. Expected τ = 112 µs (100 kΩ), 10-90% rise 246.5 µs; for 10 kΩ, 24.7 µs. This also measures C on the µs time scale the circuit actually runs at.
Basis
- Recorded
- Published
- 5 Oct 2026
- Stanbot
- v3
- Source Ref
- Scotchn Rigol capture of LCR test signal (Freq 160.36 Hz) + shorted-pair reading 1.122 nF, Discord #neuralstan20-testing 2026-10-05; vic_calc expression; amends #3447, #3476
- Notebook Id
- 3486
scotchn lcr-calibration interwinding-capacitance measurement-span bifilar