Electrical Engineering › Section M · Transformers and polyphase systems › Topic M.1

Mutual inductance

Place one coil near another and a changing current in the first induces a voltage in the second, with no wire between them. That shared flux, measured by mutual inductance, is the principle behind every transformer.

One coil talking to another

In "Faraday's and Lenz's laws" you saw that a changing flux through a coil induces a voltage across it. Now put two coils side by side. A changing current in coil 1 sets up a changing flux, and part of that flux also threads coil 2, so a voltage appears across coil 2 even though nothing connects them electrically. This coupling through a shared magnetic field is mutual induction. By transformer convention, the coil connected to the source is the primary and the coil connected to the load is the secondary. Quantities on each side carry the subscripts p and s: \(N_p\) and \(N_s\) turns, \(L_p\) and \(L_s\) self-inductance.

Flux that links both coils

Let \(\phi_p\) be the flux produced by the primary current and \(\phi_m\) the part of it, the mutual flux, that also passes through the secondary. Faraday's law applied to each winding gives

e_p = N_p\frac{d\phi_p}{dt}, \qquad e_s = N_s\frac{d\phi_m}{dt}

The primary sees all of its own flux; the secondary sees only the shared part. Flux that links one coil but not the other is called leakage flux. It does no transformer work, and later topics model it as a small inductance in series with each winding.

The coefficient of coupling

The fraction of the primary flux that reaches the secondary is the coefficient of coupling:

k = \frac{\phi_m}{\phi_p}

Because \(\phi_m\) can never exceed \(\phi_p\), k lies between 0 and 1. Substituting \(\phi_m = k\phi_p\) gives \(e_s = kN_s\,d\phi_p/dt\). Coils sharing a closed iron core have k very close to 1, because the core gives the flux an easy, low-reluctance path through both windings. Winding one coil directly over the other also raises k. Two air-core coils a few centimetres apart are loosely coupled, with k well below 1, and two coils whose axes are at right angles can have k near zero.

Defining mutual inductance

Self-inductance was defined by how much flux linkage a coil's own current produces. Mutual inductance M is defined the same way across the gap:

M = N_s\frac{d\phi_m}{di_p}

Its unit, like that of L, is the henry. Combining this with Faraday's law gives the working form

e_s = M\frac{di_p}{dt}

The relation also runs backwards: a changing secondary current induces \(e_p = M\,di_s/dt\) in the primary, with the same M. One number describes the coupling in both directions, a property called reciprocity.

M from the two self-inductances

In terms of the coil inductances and the coupling,

M = k\sqrt{L_pL_s}

Larger coils, more turns or tighter coupling all raise M. The formula also sets a ceiling. Since k cannot exceed 1, M can never be larger than \(\sqrt{L_pL_s}\), the geometric mean of the two inductances. A quoted M above that limit signals a measurement or arithmetic error. Rearranged, \(k = M/\sqrt{L_pL_s}\), which is how k is usually found in practice: measure the two self-inductances and M, then divide.

Worked example: current changing in the primary

Coils of \(L_p = 40\,\text{mH}\) and \(L_s = 90\,\text{mH}\) have \(k = 0.5\). The primary current rises steadily by 2 A in 10 ms, and the secondary is open. Find M and both induced voltages. 1. \(M = 0.5\sqrt{(0.040)(0.090)}\), which is \(0.5 \times 0.060 = 30\,\text{mH}\). 2. Rate of change: \(di_p/dt = 2\,\text{A} / 0.010\,\text{s} = 200\,\text{A/s}\). 3. Primary: \(e_p = L_p\,di_p/dt\), which is \(0.040 \times 200 = 8.0\,\text{V}\). 4. Secondary: \(e_s = M\,di_p/dt\), which is \(0.030 \times 200 = 6.0\,\text{V}\). 5. Check: M lies below the ceiling \(\sqrt{L_pL_s} = 60\,\text{mH}\), as it must.

Worked example: flux changing in the core

A primary of 200 turns and a secondary of 600 turns have \(k = 0.8\). The primary flux changes at 25 mWb/s. Find both induced voltages. 1. Primary: \(e_p = N_p\,d\phi_p/dt\), which is \(200 \times 0.025 = 5.0\,\text{V}\). 2. Secondary: \(e_s = kN_s\,d\phi_p/dt\), so \(e_s = 0.8 \times 600 \times 0.025 = 12\,\text{V}\). 3. Check: with perfect coupling the secondary would give \(600 \times 0.025 = 15\,\text{V}\), three times the primary value, matching the 3 : 1 turns. Leakage removes 20% of that, leaving 12 V. The ratio \(e_s/e_p = kN_s/N_p\) shows why tight coupling matters for transformers.

Sinusoidal currents and mutual reactance

With a sinusoidal primary current, the derivative turns into multiplication by \(j\omega\), exactly as for a single inductor in "Response of R, L, and C to sinusoids". With the secondary open, its induced voltage has rms magnitude \(E_s = \omega M I_p\) and sits 90° from \(I_p\). The product \(X_m = \omega M\) is the mutual reactance, in ohms. For \(M = 30\,\text{mH}\) at 60 Hz, \(X_m = 377 \times 0.030 = 11.3\,\Omega\), so a 2.00 A rms primary current induces about 22.6 V rms. Whether that voltage leads or lags depends on which way the coils are wound; the dot convention in "Equivalent circuits, coupled coils, and air-core transformers" settles the sign.

Where mutual inductance shows up

Transformers use it deliberately, with iron cores to push k close to 1. Wireless phone chargers and electric-toothbrush bases use it across an air gap, accepting a lower k. Clamp-on current probes and current transformers sense the alternating current in a conductor through its coupling to a winding around it. Mutual inductance is also a nuisance: neighbouring wires and printed-circuit traces couple unwanted signals into each other, called crosstalk. Designers reduce it by spacing conductors apart, running them at right angles, or twisting signal pairs so the induced voltages largely cancel.

Common mistakes to avoid

First, using \(N_s\,d\phi_p/dt\) for the secondary voltage when the coils are loosely coupled; only the mutual flux \(k\phi_p\) links the secondary. Second, writing \(M = k(L_p + L_s)\) or \(M = kL_pL_s\); the formula takes the square root of the product. Third, accepting k greater than 1 or M greater than \(\sqrt{L_pL_s}\); both are impossible. Fourth, forgetting that the induced voltage depends on the rate of change: a steady dc current, however large, induces nothing in the secondary. Fifth, mixing units, such as millihenries with amperes per millisecond, without converting.

Key takeaways

A changing current in one coil induces a voltage in a nearby coil through their shared flux. The coefficient of coupling k = φm/φp lies between 0 and 1: near 1 on a common iron core, much lower for separated air-core coils. Mutual inductance M, in henries, gives e_s = M di_p/dt, works the same in both directions, and equals k√(L_p L_s), so it can never exceed √(L_p L_s). For sinusoids, the mutual reactance X_m = ωM gives E_s = ωM I_p with the secondary open. Mutual inductance powers transformers and wireless chargers and causes crosstalk between nearby wires.

Flashcards

Flashcard

Mutual induction

A changing current in one coil induces a voltage in a nearby coil

The coils share flux, not wires.

Test your understanding

Question 1 of 5

Two coils have self-inductances of 20 mH and 45 mH and a mutual inductance of 12 mH. What is their coefficient of coupling?

Question 2 of 5

Coils of 50 mH and 200 mH are coupled with k = 0.6. What is their mutual inductance?

Question 3 of 5

Two coils have M = 25 mH. The secondary is open and the primary current changes at a steady 400 A/s. What voltage is induced across the secondary?

Question 4 of 5

Which arrangement of two coils gives a coefficient of coupling closest to 1?

Question 5 of 5

Coils of 10 mH and 40 mH are said to have a mutual inductance of 25 mH. What can you conclude?

← Previous topicNext topic →