[jeff5may]

oilpan,

The main problem with passive (capacitor) PFC is that the capacitors will continue to draw current when you are not actually welding. Sure, you might use 10 amps less peak current, but you will be continuously drawing more than that 10 amps at idle. Instead of inductive reactance at full load, you will be drawing capacitive reactance at part load and idle.

The other problem is that power factor is a moving target. Depending on how many amps and how much inductance (dig or sharpness control) you are using, the power factor will change. The phase angle will change due to amp draw, arc length, material thickness, grounding method, cable length, etc. ON ITS OWN.

Another thing to consider is that capacitor "padding" the input of the welder only corrects for linear (or passive sinusoidal) mismatching. Anything active in nature (generators, regulators, rectifiers, transistors, etc.) pushed or pulled through causes harmonics, which could interfere with other parts of the system. One bad situation is feedback with the generator due to resonance.

Not trying to put you off your idea, but this is the kind of stuff that electrical engineers have fun with. Of course, they use sophisticated, expensive equipment, and even more sophisticated formulas and methods of figuring out what's actually happening in the circuit. Cuz with AC power, most standard test equipment lies to a certain extent.