If you take a standard three-phase stator winding, (such as a squirrel-cage induction motor outer casing) and lie it horizontally on a bench with the motor axis vertical, and supply it with a three-phase voltage, a rotating magnetic field will be set up inside the stator. The strength of the field will be constant but its orientation will rotate smoothly.

The rotating magnetic field moving past the stator windings develops an emf - this is just enough to match the applied external voltage. Electrically, the setup can be represented as a voltage source (the developed emf) connected to an external voltage (the supply) through an impedance which is highly inductive.

If instead of powering the stator and putting a compass needle in the centre, we leave the stator unpowered but put a strong magnet (can be permanent or an electromagnet) in the centre and rotate it, the field of the rotating magnet will develop an emf in exactly the same way as the rotating magnetic field.

With the stator connected to a supply AND a fixed magnetic field rotating inside it, there is interaction between the two fields. To make sense of anything, these must both be rotating at the same speed - or synchronised. In this case, we can look at what happens by analysing the circuit made up of 2 AC voltages of the same frequency but different phases. connected through an inductance.

If the two voltage amplitudes are the same (V) but there is a phase difference of T degrees between them, with a reactance of X ohms connecting the two, a current will flow in the inductance. This current lags the DIFFERENCE in voltage between the two sources by 90 degrees, so has a phase half-way between the two. The result is an electrical power flow from the leading voltage source to the lagging one of (V^2) sin(T)/X watts. As I mentioned in my earlier post, this electrical power flow must be matched by a mechanical power which is proportional to the torque. If we apply more torque to the rotating magnet, the angle between the two fields will increase and sin(T) will increase (but only up to a maximum of 90 degrees - if more torque is applied at this point, pole slipping takes place and the two fields fall out of synchronism.)

As the angle difference between the two fields increases, the voltage difference across the inductive element also increases - so more current flows. Since the angle difference increase the torque and the current at the same time, torque and current are proportional.

Note that my explanation has not so far looked at which way round the two voltages are related. If there is no torque applied, the two fields are in exact step and the angular difference T is 0 - no power is produced or consumed. If the rotating element is loaded in some way, its field will lag behind the stator field and power will flow from the external source into the stator - this then becomes a motor. On the other hand, if the rotating element is driven so that its field moves ahead of the stator field, the torque will oppose the driving torque and power fill flow from the stator to the external source - this is then generator action.

In relation to your question about pole position detection, there is no need for external measurement of pole position - the physics of the interaction between the two fields takes care of that.

I hope this helps - any more detailed explanation will need some pretty heavy maths which this forum isn't really equipped to deal with.