12.4 Ampère's law: solenoids and cablesSolenoids, cables and the fields worth knowing

A long solenoid with n turns per meter, carrying current I, has a uniform field inside along its axis and almost no field outside (the fields of the near and far sides of each turn cancel out there). Take a rectangular Amperian loop with one side of length ℓ inside, along the axis, and the opposite side outside. The two short sides are perpendicular to B inside and sit where B ≈ 0 outside, so only the inside side counts: $B\ell = \mu_0(n\ell)I$. So $B = \mu_0 nI$, the same everywhere inside, whatever the distance from the axis.
?Wait, what's the difference between N and n?
I thought about it. Show me
- (a) $n = \frac{600}{0.30} = 2000\ \text{m}^{-1}$. $B = \mu_0 nI = (4\pi \times 10^{-7})(2000)(2.5) = 6.3 \times 10^{-3}$ T.
- (b) $r = \frac{mv}{eB} = \frac{(9.11 \times 10^{-31})(2.0 \times 10^6)}{(1.6 \times 10^{-19})(6.28 \times 10^{-3})} = 1.8 \times 10^{-3}$ m, about 1.8 mm.
Why?
Units 12.2 and 12.4 together: one law gives the field, another gives the motion. Exam questions love that combination. - (c) Moving along the axis means moving parallel to B, so the magnetic force is zero. It goes straight at constant speed.
Answer: (a) $6.3 \times 10^{-3}$ T (b) about 1.8 mm (c) no force: straight line, constant speed.
Stuck? Get a hint
- n = N divided by length, in turns per meter.
- B = μ₀nI, then r = mv/(eB).
- Parallel to B: sin θ = 0.
Coaxial cables and other symmetric currents
A coaxial cable has a central wire carrying I one way and an outer cylindrical shell carrying I back. A circle outside both encloses $I - I = 0$, and by symmetry B is the same all the way around, so $B(2\pi r) = 0$ and B = 0 outside. That's why coax doesn't leak field into its surroundings. Between the wire and the shell, only the center wire is enclosed: $B = \frac{\mu_0 I}{2\pi r}$. Inside a hollow pipe whose wall carries a current, a circle in the hollow encloses nothing, so B = 0 there.
?Wait, why can we say B is the same all around the circle?
| Current | Field | Found with | Direction |
|---|---|---|---|
| Long straight wire | $\frac{\mu_0 I}{2\pi r}$ | Biot-Savart or Ampère | Circles the wire (right-hand grip) |
| Center of a loop, radius R | $\frac{\mu_0 I}{2R}$ | Biot-Savart | Along the axis |
| Center of an arc of angle φ | $\frac{\mu_0 I\varphi}{4\pi R}$ | Biot-Savart | Perpendicular to the arc's plane |
| Inside a thick wire, uniform J | $\frac{\mu_0 Ir}{2\pi R^2}$ | Ampère | Circles the axis |
| Inside a long solenoid | $\mu_0 nI$ | Ampère | Along the axis, uniform |
| Outside a coaxial cable | 0 | Ampère | Net enclosed current is zero |
Beyond the course, as a bonus: a toroid (a solenoid bent into a doughnut of N turns) has, on a circle of radius r inside its core, $B(2\pi r) = \mu_0 NI$, so $B = \frac{\mu_0 NI}{2\pi r}$, and zero outside.
- AZero
- B$\frac{\mu_0(3.0)}{2\pi r}$
- C$\frac{\mu_0(6.0)}{2\pi r}$
- DIt depends on the shell's radius
Show the answer
A. A circle around the whole cable encloses $3.0 - 3.0 = 0$ A, and by symmetry B is the same everywhere on that circle, so B = 0. Choice B counts only the center wire. Choice C adds the currents instead of letting them cancel.

On an Ampère derivation, write the symmetry argument in words before any math: "By symmetry, B is tangent to circles centered on the axis and has the same size everywhere on a given circle." Then $\oint \vec B \cdot d\vec\ell = B(2\pi r)$. That sentence earns a point on its own.
- Long solenoid: $B = \mu_0 nI$ inside (uniform, along the axis), B ≈ 0 outside; $n = N/\ell$.
- Coax: B = 0 outside (net current zero); between, only the center wire counts.
- Hollow current-carrying pipe: B = 0 in the hollow.
- Ampère for symmetric shapes, Biot-Savart for everything else.