|
Not feeling too well at the moment, horrible tummy upset with lots of pain. And this coincides with what I think is the most important revelation in my OU career. I will write this up in detail but meanwhile here is the gist of the approach.
You will note that I recently posted about eddy current heating and also about PM generators having OU and UU regions that are not generally recognized (wrongly posted under the pulse motor thread). Well it turns out that there is a common feature linking these two and this becomes strikingly obvious when you analyze things in the magnetic domain. When you operate a PM generator outside its normal regime by loading it too heavily with a low resistance load you get into the regime where the reactance of the coil becomes greater than the load resistance thus introducing considerable phase shift between induced voltage and current. You don't normally notice this phase shift because you measure only the voltage across the load resistor and that has to be in phase with the current through it, what you can't get at is what is going on magnetically. There is the usual dilemma that the induced voltage comes not just from the changing flux due to movement in a magnetic field, but also from the flux due to the current flow and these add vectorally when considered as phasors. This dilemma is solved when you create the magnetic domain solution where the PM is modeled as a mmf generator in series with a reluctance, and air paths are modeled as reluctances. A coil connected to a load resistor is modeled as a magnetic inductor obeying mmf=-L*dPHi/dt. It then becomes clear that when the coil is in line with the magnetic poles, at the point where torque is zero, there is present this dPhi/dt changing flux that is driving current through the load and that is not coming from the movement, but from the magnet-to coil circuit that involves the mmf of the magnet, the magnetic resistances (actually air reluctances) and the magnetic inductance (the loaded coil). Truly at that point the load power is coming from the magnet's mmf, not from the mechanical drive. You can look upon the coil movement as a form of magnetic switch that switches the coil into the magnetic circuit, then the phase retarded flux flow becomes obvious. You end up with some seemingly weird results, like power is proportional to 1/N^2, IOW you need minimum turns. This explains why the eddy current system works well since you can't get less than a single turn. Lots more to come on this.
Smudge
|
Author
Topic: Dynamic magnetic domain analysis (Read 9211 times)
