I'll take a stab at a non-math based explanation.
Think about what a transformer is, I'm going to use a single phase simple AC core and coil transformer we all use every day. If you took one apart to sell for scrap, you would find it is essentially just two long pieces of wire wrapped in a coil around a piece of steel. How it transforms from one voltage to another is based upon the phenomenon called "induction" which says that if you move a magnetic field across a conductor, it causes current to flow in the conductor. So the transformer, in it's steady state, creates electro-magnetic fields that form around one coil that permeate the steel and make it into a magnet. As the AC current rises and falls, the magnetic flux of the magnet is rising and falling too, so then creating current flow in the other wire coil, which is wrapped around the same steel core. That voltage will be lower or higher than the original one by the ratio of how many times the coils are wrapped around that magnet in relation to each other, called the "turns ratio".
But that is at the STEADY state, and at some point, you must START this process. If you look at the circuit, other than the fact that it is coiled, the wire is just a wire; it is in essence a "short circuit". Once the magnetic fields are created and begin interacting with each other, the current flow in both sides "impedes" the flow of current in the other through a process called "mutual induction". In works kind of like this: Coil A has current flowing, that creates magnetic flux in the steel, which induces current to flow in coil B, which ALSO induces magnetism in the steel, which then opposes the flow of current in coil A, which limits the current in coil B etc. etc. etc. Everyone is happy.
But for a BRIEF moment, BEFORE the flow of current in coil A creates magnetic flux in the steel so that current in B can oppose it, there is NO impedance to the flow of current through coil A, it really IS a "short circuit" for that instant, where the current flow is slowed down ONLY by the natural resistance in the wire itself. The magnitude of that current, which can be anywhere from 1000-2000% of the steady state maximum current, is called the "magnetizing current". It only lasts as long as it takes for the steel to become magnetized and depending on the design, can range from anything between a few cycles (at 60Hz, a cycle is 16.7ms) and a minute, but most fall into that 1-3 cycle range.
The application is a wind turbine. I was talking to some folks on a RE forum I am also on and it seems my existing transformer may not work well for my system redesign where I am increasing battery voltage from 12 to 24. Thought I would post the problem here too as there are some real good theory people here. It puts out three phase AC, variable frequency at 240V nominal (have seen up to 300), 1 KW max. Currently the voltage is stepped down to 12 volts by a wye-delta transformer, then runs thru a three phase bridge rectifier, then to a 12v battery bank. The battery bank will be upgraded to 24V so the current setup will not work. In addition, there is a new maximum power point tracking controller that I would like to incorporate and it can take up to 250VDC input so I really need to at least halve the voltage from the turbine (dont forget the 136% voltage increase from the bridge rectifier). The controller can buck the incoming voltage but not boost. I had the idea to change the primary to delta and the secondary to wye which would give me a 1.73 squared decrease in reduction ratio thus getting me over my battery voltage. It was mentioned that this would increase the magnetizing current which is an issue for a wind turbine because it can have trouble starting connected to a transformer due to the low frequency at cut in. So I was wondering if anyone had any ideas about perhaps unwinding or rewinding the transformer or how I can "derate" one to make it operate with a low magnetizing current. There you go, that should be the most out there problem of the week.
