EN

Magnetically Coupled Coils

Introduction

When the behaviour of one coil depends on, or is affected by, the behaviour of another, the coils are said to be magnetically coupled. This is the case with transformers, where at least two coils share the same flux. This flux travels around the magnetic circuit, but along the way part of it leaks through the materials surrounding the laminated core; although they are poorer magnetic conductors, they let part of the flux through, and that flux is wasted.

Taking into account the reluctances of both the laminated core and the surrounding medium, the coupling coefficient of the coils (k) is:

$$ k = \frac{\mathbb{R_p}}{\sqrt{(\mathbb{R_1}+\mathbb{R_p})(\mathbb{R_2}+\mathbb{R_p})}} \leq 1 $$ $$ \mathbb{R_p} = \text{reluctance of the non-magnetic medium} $$ $$ \mathbb{R_1} = \text{reluctance of the primary} $$ $$ \mathbb{R_2} = \text{reluctance of the secondary} $$

If the primary and secondary reluctances are equal, which is quite common, the equation becomes:

$$ k = \frac{\mathbb{R_p}}{(\mathbb{R}+\mathbb{R_p})} \leq 1 $$

This coefficient lets us calculate the mutual inductance (M), which can be described as the ability of one coil to induce a voltage in another when the current in the first one changes.

$$ M = k \cdot \sqrt{L_1 \cdot L_2} $$ $$ M, L_1, L_2 \space \text{in henries} $$

Analysing circuits with mutual inductance

Preliminary considerations

Let's see how mutual inductance affects a circuit. But first, suppose we have two coupled coils whose corresponding terminals we do not know: identifying them is essential before continuing with the analysis.

In fact, if we know how the winding was made, the corresponding terminals are easy to work out. Here we have two coils wound around the same magnetic circuit, so they are coupled. For both AC and DC analysis we need to know the corresponding terminals.

To do this, we arbitrarily choose the direction of the current in both coils. Then we use the Corkscrew rule to find the direction of the flux in the core: if the fluxes of both coils agree, the corresponding terminals are the ones where the current enters. If they disagree for one of the coils, its corresponding terminal is the opposite one.

A core may have more than two coils wound on it; the same criterion applies. As shown with the new coil, choosing the current along an alternative path means the other terminal becomes the corresponding one, so that the corkscrew rule holds for every coil.

In the previous figures we knew how the coils were wound, but often we do not, because the winding is encapsulated and only the connection terminals are accessible. In that case a voltage test is needed. It can be done at rated voltage if basic safety rules are followed and two voltmeters are at hand.

Make the connections shown above; the jumper between terminals V and Y is only for the test. Terminals U and X are corresponding if V2 > V1. Otherwise, that is if V1 > V2, the corresponding terminals are U and Y (or V and X, which is the same for later analysis).

Analysis of basic circuits

Let us first recall that an "ideal" inductor converts electrical energy into magnetic energy, which can be returned to the circuit without losses. In DC it is a component that prevents the current from changing abruptly, and when the current is constant it behaves like a short circuit. It is also assumed to be linear and time-invariant. Similarly, an ideal capacitor stores energy as electric charge, prevents the voltage from changing abruptly, and when subjected to a constant voltage its current is zero, so under those conditions it is equivalent to an open circuit.

$$ \text{DC} $$
$$ \text{AC} $$
$$ u(t) = L \frac{d i}{d t} $$ $$ i(t) = i(t_0) + \frac{1}{L}\int_{t_0}^{t} u(t)\,dt $$
$$ \overline{U} = j\omega L \overline{I} $$ $$ i(t) = \left| I \right|\cos\left(\omega t - \frac{\pi}{2}\right) $$ $$ u(t) = \left| U \right|\cos(\omega t) $$

In DC it follows directly that when \( i = constant\), \( u = 0\)

In AC, \(\overline{U}\) and \(\overline{I}\) are phasors, \(\left| U \right|\) and \(\left| I \right|\) are their magnitudes and \( \omega = 2\pi f \)
Since \(Z = \sqrt{R^2 + X^2}\), and in an ideal inductor \(R = 0\) and \(X = \omega L\), from now on \(\omega L = Z_L\)

Now let's look at a simple coupling: two coils sharing a magnetic core. It does not have to be a conventional transformer; if it were, it would be marked \(1:N\) or \(N:1\) to show the turns ratio. In the following example we already know the corresponding terminals.

First, we arbitrarily choose the directions of the currents and voltages; if we get one wrong, the result will simply come out negative. We have chosen the voltage and current directions treating both coils as "loads". This is not by chance: it simplifies the equations, although it is not mandatory.

Direct current

The voltage \(u_1\) has the value given above, but since this coil is coupled to the other, an additional voltage is induced, proportional to the mutual inductance \(M\) and to the current \(i_2\) in the other coil; the same applies to that coil. The equations take the form:

$$ u_1 = L_1 \frac{d i_1}{d t} \pm M \frac{d i_2}{d t} $$ $$ u_2 = L_2 \frac{d i_2}{d t} \pm M \frac{d i_1}{d t} $$

The \(\pm\) sign of the second term depends on the direction of the currents relative to the corresponding terminals. If both currents enter (or leave) their respective corresponding terminals, the sign of \(M\) is the same as that of the \(L\) term in the same equation; otherwise it has the opposite sign. The figure above therefore follows these equations:

$$ u_1 = L_1 \frac{d i_1}{d t} + M \frac{d i_2}{d t} $$ $$ u_2 = L_2 \frac{d i_2}{d t} + M \frac{d i_1}{d t} $$

Now take the last figure and reverse the direction of \(i_1\). This does not really affect the analysis; the value of this current will simply come out with the opposite sign, and we have to reinterpret what was said above:

On the one hand, in coil 1 the voltage and current arrows oppose each other, which means \(u_1 = -L \frac{d i_1}{d t}\). We also see that \(i_1\) leaves the marked terminal while \(i_2\) enters it, so they disagree. This means the mutual inductance terms in each equation take the opposite sign to the \(L\) term:

$$ u_1 = -L_1 \frac{d i_1}{d t} + M \frac{d i_2}{d t} $$ $$ u_2 = L_2 \frac{d i_2}{d t} - M \frac{d i_1}{d t} $$

The result would have been the same if we had simply changed the sign of \(i_1\).

Now consider a system of three magnetically coupled coils. In this case the couplings are established in pairs: coil 1 couples with coil 2 under certain conditions, coil 2 with coil 3, and coil 3 with coil 1. The number of mutual inductances \(N_m\) for \(n\) coils is \(N_m = \frac{n(n-1)}{2}\). So we have three mutual inductances, whose values may differ. The markings on each coil help us understand this.

$$ u_1 = L_1 \frac{d i_1}{d t} + M_{12} \frac{d i_2}{d t} - M_{13} \frac{d i_3}{d t} $$ $$ u_2 = L_2 \frac{d i_2}{d t} + M_{12} \frac{d i_1}{d t} + M_{23} \frac{d i_3}{d t} $$ $$ u_3 = L_3 \frac{d i_3}{d t} - M_{13} \frac{d i_1}{d t} + M_{23} \frac{d i_2}{d t} $$

The term \(\frac{d}{dt}\) can be replaced by the operator \(D\), and the system can be solved as if \(D\) were a parameter. This yields a differential equation whose solutions depend on the values of all the elements (coefficients) and on the initial conditions of the circuit. After a while the steady state is reached and all the time-dependent variables settle and become constant, which means we can set \(D = 0\) in all the equations. The values obtained remain valid as long as the circuit is not changed.

Alternating current

In alternating current everything said about mutual inductances works the same way, but we will use phasors and complex numbers for the steady-state analysis.

$$ \overline{U_1} = jZ_{L_1} \overline{I_1} + jZ_M \overline{I_2} $$ $$ \overline{U_2} = jZ_{L_2} \overline{I_2} + jZ_M \overline{I_1} $$

$$ \overline{U_1} = jZ_{L_1} \overline{I_1} + jZ_{M_{12}} \overline{I_2} - jZ_{M_{13}} \overline{I_3} $$ $$ \overline{U_2} = jZ_{L_2} \overline{I_2} + jZ_{M_{12}} \overline{I_1} + jZ_{M_{23}} \overline{I_3} $$ $$ \overline{U_3} = jZ_{L_3} \overline{I_3} - jZ_{M_{13}} \overline{I_1} + jZ_{M_{23}} \overline{I_2} $$