My objective is to devise a TMS unit which makers can reproduce widely, preferably for the cost of a typical Friday night party. I'm guessing that cost normally exceeds $100 dollars. This unit falls woefully short by that measure, and probably on a number of other counts. Early indications are it won't produce Muscle Twitch Threshold (a power level setting used in a recent National Institutes of Health study, results of which we seek to replicate). This unit appears overly complex, and it contains about $1000 dollars worth of hardware. Still it represents a starting point. It appears to be reliable. Peak output power is substantial (measured by the palm of my hand). These are key milestones. Reliability comes first. Next we optimize peak power. Then we simplify, reducing costs while easing reproduction.
Here we see magnetic coupling powering a 120 Volt light bulb, using a 120 turn transformer winding. Inside the applicator coil body, a 35 turn winding is being pulsed fifteen times per second with 2,100 volts, driven by capacitor discharge (20 microfarads), switched by an SCR. That magnetic pulse could just as well be firing neurons from remotely.