Presets
t = 0.00 s

⚡ Power & drive

live

🔀 Switching transistor

🧲 Coil & core

live

🛞 Rotor & load

live

Rotor

stopped — press Spin it up

Rotor speed

RPM over sim time

Per-event energy

mJ per pulse

Average power

W at the current pulse rate

Oscilloscope — one magnet event

Three probes across one magnet pass: CH1 trigger-winding EMF (fires the base), CH2 power-coil current ramp, CH3 collector voltage — watch the inductive spike at switch-off. The shaded band is the transistor ON window (dwell).

Driving circuit — Bedini SSG topology

Power coil (“kickback” coil) Trigger coil (same spool) Run battery 12 V R base 2N3055 Flyback diode Charge battery 12 V — drive current — spike →

How the drive circuit works

1
A rotor magnet sweeps past the coil spool. The spool carries two windings: a fat power coil and a thin trigger coil wound together (bifilar), so both see the same passing magnet.
2
The moving magnet induces a small EMF in the trigger coil. Through the base resistor this switches the transistor ON — no commutator, no Hall sensor, the magnet times itself.
3
While ON, the run battery ramps current through the power coil (CH2 on the scope). The energised core pushes on the magnet, torquing the wheel, and energy ½·L·I² is stored in the magnetic field.
4
The trigger EMF reverses as the magnet passes and the transistor snaps OFF. An inductor's current can't stop instantly — the field collapses and the coil voltage spikes high (the famous “kickback”, CH3).
5
The flyback diode steers that spike into the charge battery, recovering most of the stored field energy instead of frying the transistor. Without a recovery path the spike rings up until the transistor avalanches — that's why SSG builders clip a neon bulb (~90 V strike) across the transistor as a tell-tale and protector.

Try it: disconnect the recovery path in Power & drive → Kickback path and watch the spike voltage climb past the transistor's rating. Then compare a slow 60 V 2N3055 against a fast 200 V MOSFET.

The big Ferris wheel: Bedini's 10-foot expo machine wasn't one coil — it was a bank of coils around the rim. One master coil carries the trigger winding; every slave transistor switches in parallel off that single signal, so each magnet pass fires the whole bank at once and every collapse spike is collected. A hub starter motor spins the heavy wheel up (press 🚀 Starter, then switch it off) — once the rim is moving, momentum carries most of the work, and you can drop the bank down to one coil and watch it hold speed on a fraction of the drive power. The spikes can be routed to the charge battery or into ferrite hub-assist coils as extra torque. The ledger stays honest either way: the assist returns at most ~70 % of the spike energy as push, so the run battery is always the ultimate source — what the big wheel really shows off is superb energy recycling on a giant flywheel, not free running.

Physics: saturating-core coil model (Frohlich), variable-reluctance torque T = ½·i²·dL/dθ, RK4 pulse solver with per-event energy quadrature — ported from the PulseMotor Python core and verified against it to <0.1%. Per-event energies are raw, uncapped line items; losses (copper, switch, diode) are explicit. Representative datasheet values for transistors and core materials.