observation · Voltage Intensifier Circuit · computed
Chris Freedom full-bridge 2-pulse gated shot: forward-mode ramp visible during on-time but small; forward amplitude is fixed by turns ratio × Vin, so f/duty can't raise it
Bench: Chris Freedom, Discord #neuralstan20-testing 2026-11-07, handheld scope photo, 100 us/div. Same CT rig with full bridge (D3/D4). Chris marked the on-times (red) by hand. CH1 yellow (x100, 1000 V/div): Max 1.6 kV, Vpp 1.0 kV, Rms 506 V, +Du 5.14%, Fr 12.25 kHz. The trace never drops below ~0.6 kV, so it is still not a drain-source voltage (reference likely still CT, see #3529). CH2 blue (100 V/div): Max -24 V, Vpp 76 V (so min about -100 V), Rms 62 V, +Du 56.85%, Fr 105.5 kHz. Pattern: 2 pulses per burst, then one missing pulse (gate off). During each marked on-time the blue rises slowly (forward mode) and the yellow stays flat. At each turn-off the blue steps down sharply and the yellow spikes (flyback). Between bursts the yellow decays. The decay is cell/probe/diode leakage and sets how much step charge is possible. Analysis: forward secondary voltage = Vin × Ns/Np ≈ 10 × Vin, so about 40 V per half at 4 V, independent of frequency and duty. Flyback energy = ½·Lp·Ipk² (Lp ≈ 0.915 mH scaled from Sec A 91.5 mH; ≈56 uJ at 4 V and 80 us). That grows with on-time, which explains why the forward mode looks weak. Remedy: put a series resonant charging choke (165 mH) + blocking diode in the forward path. The choke rings the cell to ~2× the winding voltage per pulse, with charge time π√(L·C_cell) (≈40 us for 1 nF). Open: probe points and reference for this shot, switched Vin, cell C.
Basis
- Recorded
- Published
- 8 Nov 2026
- Stanbot
- v3
- Source Ref
- Chris Freedom scope photo, Discord #neuralstan20-testing 2026-11-07; extends #3528, #3529
- Notebook Id
- 3530
chris-freedom forward-mode flyback step-charge full-bridge