Talk-Polywell.org

kunkmiester wrote:My dad mentioned this idea. You get pulsed DC out, which then just needs a transformer. Big challenge is controlling the field collapse on the down side of the DC pulse, which I'm sure people know how to do. A bigger challenge might be that you'd need to pulse the reactor at some arbitrary frequency, and you'd then have to change that down to 60 hz, and depending on how that works, it might not be worth it.

Reactors are NEVER designed for some arbitrary frequency. In power circuits - because of the expense - the circuits are designed to operate over a limited range. For wide ranges (AC motor servos) compensation schemes are designed in - V/Hz for example.

What I meant was that the 50/60 hz might prove impractical enough that you'd need to pulse the reactor at a different rate. If a different pulse width worked well enough it would be favored instead, then you'd need to change your frequency to match the grid. How high could you go and still get a managable pulse? I'd imagine at some point with the way the polywell works, they'd eventually blend together, and you'd not have distinct enough pulses to work with.

kunkmiester wrote:What I meant was that the 50/60 hz might prove impractical enough that you'd need to pulse the reactor at a different rate. If a different pulse width worked well enough it would be favored instead, then you'd need to change your frequency to match the grid. How high could you go and still get a managable pulse? I'd imagine at some point with the way the polywell works, they'd eventually blend together, and you'd not have distinct enough pulses to work with.

With IGBTs the pulsing is done at a 5 KHz rate. If you went to high power MOSFETs the rate could probably be increased to 25 KHz.

However, at this time IGBTs are the low cost solution.