ARCHETYPE: Ampère Setups — Cables, Sheets & Cavities

Open a JEE-Advanced magnetism section from almost any year and somewhere in it you will find a cylinder wearing a costume: a solid core inside a hollow return shell, a pipe with current only in its wall, a conductor with an off-centre hole drilled through it. The examiners are testing whether you can run one ritual in several rooms. Ampère's Circuital Law hands you the master key: on any centred circle, $B \times 2\pi s = \mu_0 I_{\text{enc}}$ . Every one of these setups is the same two-line solve; all that changes, region by region, is the bookkeeping of how much current your loop actually encloses. That bookkeeping is this archetype. You will map the complete field profile of a coaxial cable in four counts, silence the inside of a hollow pipe in one, walk a rectangle across an infinite current sheet, and crack the most elegant trick in the family — the off-axis cavity whose field turns out to be perfectly uniform. The profiles echo each other: the coax outside is as field-free as a charged shell's interior was in Gauss's Law at Work: Wire, Sheet & Shell, and for the same deep reason — totals laws count sources, and cancelled sources are invisible. Fair warning about jurisdiction: JEE-Main stops Ampère's law at the wire and the solenoid. These multi-region setups are JEE-Advanced territory.

Key takeaways

Every cylindrical setup obeys one master line: $B \times 2\pi s = \mu_0 I_{\text{enc}}(s)$ — the geometry never changes, only the enclosed-current count per region.
Coaxial cable (core $a$ , shell $b$ to $c$ ): linear rise in the core, $\dfrac{\mu_0 I}{2\pi s}$ in the gap, a progressive kill inside the shell, and exactly zero outside — coax is magnetically invisible.
Hollow cylinder: $B = 0$ everywhere in the bore, full-current $1/s$ field outside — the magnetic twin of the charged shell's field-free interior.
Infinite current sheet ( $K$ amperes per metre of width): $B = \dfrac{\mu_0 K}{2}$ on each side, parallel to the sheet, flipping direction across it — and independent of distance.
Off-axis cavity inside a current-carrying cylinder: superpose the full cylinder with a negative-current cavity cylinder; the cavity field is uniform, $B = \dfrac{\mu_0 J d}{2}$ , set only by the axis-to-axis offset $d$ .
Key idea: pick the loop, identify the region, count the enclosed current honestly — every exotic cylinder question is the straight-wire solve wearing a different $I_{\text{enc}}$ .

The core insight

Multi-region Ampère setups — coaxial cables, hollow pipes, current sheets, off-axis cavities — all collapse to $B \times (\text{loop length}) = \mu_0 I_{\text{enc}}$ with a region-by-region count of the enclosed current; mastering the count (fractions of areas, signed returns, superposed negatives) is mastering the archetype.

Where this lives today

Every BNC lead, antenna feed and cable-TV drop is the four-region profile doing its job: signal out on the core, back on the braid, zero external field — and zero pickup, since the same cancellation hides the world's fields from the cable. Power engineers lay three-phase underground lines so the currents sum toward zero for the same reason; the off-axis cavity trick remains the fastest honest way to field-map drilled and slotted conductors in electric machines. In this course, the sheet returns promoted in the UG deep dive (its $\mu_0 K$ jump is the universal boundary behaviour at current-carrying surfaces), and Displacement Current: Maxwell's Fix (E8) confronts what happens when currents refuse to stay steady.

The setup

The canonical member of the family is the coaxial cable: a solid inner conductor of radius $a$ carrying current $I$ toward you, then a conducting shell from inner radius $b$ to outer radius $c$ carrying the same $I$ back. Four regions, one symmetry: the field is tangent to centred circles with the same magnitude all the way around each one — the legality argument of Ampère's Circuital Law, holding in all four rooms at once.

Figure

The coaxial cable in cross-section — four regions, one recipe, four different counts

The discipline that earns the marks is region identification before formula selection. Ask two questions of every point: which room is $s$ in, and what current does a circle of radius $s$ actually catch? The walk-the-loop meter from Ampère's Circuital Law reads $\mu_0 I_{\text{enc}}$ in every room, and the symmetry sentence legalises dividing by $2\pi s$ in every room too. One inherited result does heavy lifting throughout: inside a uniformly filled conductor the enclosed fraction is $s^2/R^2$ and the field rises linearly — the thick-wire interior derived in Ampère's Circuital Law. We never re-derive it; we install it.

The master pattern: the coaxial cable in four counts

Derivation AThe complete field profile of a coaxial cable

The goal is $B(s)$ in all four regions of a coax — core of radius $a$ carrying $I$ uniformly, gap, return shell from $b$ to $c$ , outside — using one Amperian circle of radius $s$ . The strategy is fixed by the symmetry: in every region the circulation collapses to $B \times 2\pi s$ , so the whole derivation is four honest counts of $I_{\text{enc}}$ . Watch the shell region especially: the return current is subtracted, and only the fraction of the shell's annular area inside your circle has returned so far.

Step 1. Core, $s \le a$ . The circle catches the area fraction $\pi s^2/\pi a^2$ of the uniformly spread current — the same count that built the thick-wire interior in Ampère's Circuital Law. So $I_{\text{enc}} = I s^2/a^2$ and

B \times 2\pi s = \mu_0 I \frac{s^2}{a^2} \;\;\Rightarrow\;\; B = \frac{\mu_0 I s}{2\pi a^2}

rising linearly from zero at the axis to a peak $\dfrac{\mu_0 I}{2\pi a}$ at the core surface.

Step 2. Gap, $a \le s \le b$ . The circle now swallows the whole core and none of the shell: $I_{\text{enc}} = I$ , the plain straight-wire result

B = \frac{\mu_0 I}{2\pi s}

falling as $1/s$ from the peak. The insulation is magnetically irrelevant — only current counts.

Step 3. Shell, $b \le s \le c$ . The new move. The circle encloses $+I$ from the core and minus the fraction of the return current whose annular area lies inside radius $s$ . Annulus areas: the shell's total is $\pi(c^2 - b^2)$ ; the part inside $s$ is $\pi(s^2 - b^2)$ . So

I_{\text{enc}} = I - I\,\frac{s^2 - b^2}{c^2 - b^2} = I\,\frac{c^2 - s^2}{c^2 - b^2}, \qquad B = \frac{\mu_0 I}{2\pi s}\cdot\frac{c^2 - s^2}{c^2 - b^2}

Why the fraction uses annulus areas: the return current is spread uniformly over the shell's ring-shaped cross-section, so "how much has returned by radius s" is a ratio of ring areas, not of full circles — writing $s^2/c^2$ here is the classic harvested error.

Step 4. Outside, $s \ge c$ . The circle encloses $+I$ out and $-I$ back: $I_{\text{enc}} = 0$ , so $B = 0$ exactly — at a millimetre from the shell or a kilometre. Why the exterior zero is exact: the law is blind to where currents sit inside the loop, so equal and opposite threaded currents give zero circulation with no fine tuning — and the symmetry forces the field itself to vanish, not merely average out. *Verified: continuity at every border for $I = 12$ A, $a = 2.0$ mm, $b = 6.0$ mm, $c = 8.0$ mm — Step 1 at $s=a$ gives $1.2$ mT and Step 2 at $s=a$ gives $1.2$ mT ✓; Step 2 at $s=b$ gives $0.40$ mT and Step 3 at $s=b$ (bracket $=1$ ) gives $0.40$ mT ✓; Step 3 at $s=c$ gives zero, matching Step 4 ✓. Spot value: at $s = 7.0$ mm the bracket is $(64-49)/(64-36) = 15/28$ , $I_{\text{enc}} = 6.4$ A, $B = 0.18$ mT, recomputed.*

The profile is worth carrying in your head as a shape — and one landmark transfers straight from the parent module: the half-peak points sit at $s = a/2$ (on the ramp) and $s = 2a$ (on the slide, provided $2a < b$ ) — for our numbers, $0.60$ mT at both $1.0$ mm and $4.0$ mm.

Figure

The coax field profile — ramp, peak, slide, shell dive, then exact zero

The hollow cylinder: silence in the bore

Strip the coax of its core and you have the next costume: a pipe of inner radius $r_1$ , outer radius $r_2$ , current $I$ spread uniformly through the wall. Any centred circle inside the bore encloses no current, so its circulation is zero — and here symmetry does let you conclude the field itself vanishes: $B$ is constant along the circle, and constant times $2\pi s$ equalling zero means $B = 0$ everywhere in the bore. This is the magnetic twin of a result you already own: the charged shell's interior was field-free in Gauss's Law at Work: Wire, Sheet & Shell, by the same totals-law logic. Inside the wall the enclosed fraction is the annulus ratio $\dfrac{s^2 - r_1^2}{r_2^2 - r_1^2}$ ; outside, the pipe is indistinguishable from a plain wire carrying $I$ .

The current sheet: a field that ignores distance

Flatten the geometry. An infinite sheet carries a surface current density $K$ — amperes crossing each metre of width. Pair mirror-image strips of the sheet (each a thin wire whose circular field you know from Fields of Straight Wires & Circular Loops): the perpendicular contributions cancel and the field ends up parallel to the sheet, perpendicular to the current, pointing opposite ways on the two sides.

Go deeper · Derivation B · 2 minThe infinite sheet by a rectangle walk

Derivation BThe infinite sheet by a rectangle walk

The goal is the field magnitude $B$ on either side of an infinite sheet carrying surface current $K$ (A/m), given the direction argument above. The strategy is the walk-the-loop recipe from Ampère's Circuital Law with a rectangular path — the shape that matches straight-line symmetry the way circles matched cylinders. The thing to watch: both long legs collect circulation, because the field flips across the sheet while the walking direction flips too — they add, not cancel.

Step 1. Choose the loop. A rectangle straddling the sheet: two legs of length $L$ parallel to the field, one on each side at equal distances, joined by two crossing legs perpendicular to the sheet.

Step 2. Walk it. Along the top leg the field runs with you: contribution $+BL$ . The two crossing legs are perpendicular to the field: zero. Along the bottom leg, you now walk the opposite way — and the field on that side points the opposite way too, so it is again with you: another $+BL$ . Total circulation: $2BL$ . Why the legs add: this is the flipped-field, flipped-walk double negative. Students who cancel the two legs get $B = 0$ and know something is wrong; the field reverses across the sheet, and so does an honest walker.

Step 3. Count the threaded current. The rectangle's film is pierced by every ampere flowing through the width $L$ of sheet it spans: $I_{\text{enc}} = KL$ .

Step 4. Solve.

2BL = \mu_0 K L \;\;\Rightarrow\;\; B = \frac{\mu_0 K}{2}

where: $K$ — amperes per metre of sheet width. $L$ cancelled, and so did the legs' distance from the sheet: the field is the same at any height. Move away and each strip helps less, but more of the sheet sits at a useful angle — for a $1/s$ elementary field the trade is exactly even, the same conspiracy that made the charged plane's $\sigma/2\varepsilon_0$ distance-blind in Gauss's Law at Work. Verified: for $K = 400$ A/m, $B = \mu_0 \times 200 = 2.5 \times 10^{-4}$ T on each side, recomputed; dimensions — $\mu_0 K$ carries (T·m/A)(A/m) = T ✓; and rolling the sheet into a cylinder of $n$ turns per metre stacks the two sides into $\mu_0 K = \mu_0 n I$ inside and zero outside, exactly the solenoid of Solenoids & Toroids ✓.

The solenoid is this sheet rolled up: inside the tube the two layers reinforce to $\mu_0 nI$ , outside they cancel. If you ever forget the sheet's factor of $\tfrac12$ , rebuild it from the solenoid you trust.

The off-axis cavity: the trick that isn't a trick

The aristocrat of the family. A long cylinder of radius $R$ carries current density $J$ (A/m²) uniformly — except that an off-centre cylindrical cavity of radius $r$ runs through it, parallel to the axis, its axis a distance $d$ from the main axis. Find the field inside the cavity. No single Amperian loop has the symmetry to crack this — the drilled conductor is lopsided. The move is superposition: the drilled cylinder equals a full cylinder of density $J$ plus a cavity-sized cylinder of density $-J$ in the hole. The fictitious currents cancel exactly where the real conductor carries nothing, and each piece separately has the full cylindrical symmetry Ampère's law loves. Inside a full uniform cylinder the field at distance $s$ from its axis has magnitude $\dfrac{\mu_0 J s}{2}$ (the linear-rise interior result of Ampère's Circuital Law, rewritten through $I = J\pi R^2$ ), wrapping around that cylinder's own axis. At a cavity point, the full cylinder's field is set by the distance from the main axis, the negative cylinder's by the distance from the cavity axis, oppositely wrapped. Add them: the position-dependent parts cancel, and what survives depends only on the offset between the axes:

B_{\text{cavity}} = \frac{\mu_0 J d}{2} \qquad (1)

where: $J$ — the current density of the intact material, $d$ — the axis-to-axis distance. Read the astonishing part: $s$ has vanished. Every point of the cavity feels the same field — same magnitude, same direction (perpendicular to the line joining the axes) — a uniform field manufactured by drilling a hole. If the hole is centred ( $d = 0$ ) the cavity field is zero, as the hollow-cylinder section already said.

The pick-your-loop triage table

setup	loop	the count $I_{\text{enc}}$	field
coax, core $s\le a$	centred circle	$I s^2/a^2$	$\dfrac{\mu_0 I s}{2\pi a^2}$ , linear rise
coax, gap $a\le s\le b$	centred circle	$I$	$\dfrac{\mu_0 I}{2\pi s}$
coax, shell $b\le s\le c$	centred circle	$I\,\dfrac{c^2-s^2}{c^2-b^2}$	dying to zero at $c$
coax, outside $s\ge c$	centred circle	$0$	$0$ exactly — invisible cable
hollow pipe, bore	centred circle	$0$	$0$ — the quiet bore
hollow pipe, outside	centred circle	$I$	plain wire, $\dfrac{\mu_0 I}{2\pi s}$
infinite sheet	straddling rectangle	$KL$	$\dfrac{\mu_0 K}{2}$ each side, distance-blind
off-axis cavity	superpose $J$ and $-J$	(two symmetric pieces)	uniform $\dfrac{\mu_0 J d}{2}$

Common trapThe return shell shields the whole cable, so the field between core and shell must be zero too

The most tempting wrong sentence in the archetype — and exactly backwards. The shell's cancellation works only for loops that enclose the return current: outside the cable, $+I$ and $-I$ thread together and the field dies. In the gap, your circle encloses the core's $+I$ and none of the return — the shell might as well not exist, and the field is the full $\mu_0 I/2\pi s$ . A conductor does not "block" magnetic fields by sitting there; only enclosed, cancelling current erases circulation. The same misread flips the hollow pipe: the bore is field-free because a bore loop encloses nothing, not because the wall shields it — run a separate wire down the middle and its full field fills the bore unimpeded.

Worked examples

Board level — the hollow cylinder, all three rooms

A long hollow copper pipe has inner radius $3.0$ mm and outer radius $5.0$ mm and carries $10$ A distributed uniformly through the wall. Find the magnetic field at (a) $s = 2.0$ mm, (b) $s = 4.0$ mm, (c) $s = 8.0$ mm from the axis. ( $\mu_0/2\pi = 2\times10^{-7}$ T·m/A.)

Step 1. Region identification, before any formula: (a) is in the bore, (b) is inside the wall, (c) is outside.

Step 2. Bore: a circle of radius $2.0$ mm encloses no current, and symmetry lets the zero circulation speak for the field itself: $\boxed{B = 0}$ .

Step 3. Wall: the annulus fraction enclosed is $\dfrac{s^2 - r_1^2}{r_2^2 - r_1^2} = \dfrac{16 - 9}{25 - 9} = \dfrac{7}{16}$ , so $I_{\text{enc}} = 10 \times \dfrac{7}{16} = 4.4$ A. Then $B = \dfrac{2\times10^{-7} \times 4.375}{4.0\times10^{-3}} = \boxed{2.2 \times 10^{-4}\ \mathrm{T}}$ .

Step 4. Outside: all $10$ A enclosed, plain wire: $B = \dfrac{2\times10^{-7} \times 10}{8.0\times10^{-3}} = \boxed{2.5 \times 10^{-4}\ \mathrm{T}}$ . Sanity: the wall value sits between the bore's zero and the surface peak $4.0\times10^{-4}$ T ✓.

Board level — reading a coax at three radii

A coaxial cable has a solid core of radius $2.0$ mm carrying $12$ A out of the page and a shell from $6.0$ mm to $8.0$ mm returning it. Find $B$ at (a) $1.0$ mm, (b) $3.0$ mm, (c) $9.0$ mm, and (d) state the peak field and where it occurs.

Step 1. Rooms: (a) core, (b) gap, (c) outside.

Step 2. Core, linear rise: $B = \dfrac{\mu_0 I s}{2\pi a^2} = \dfrac{2\times10^{-7} \times 12 \times 1.0\times10^{-3}}{(2.0\times10^{-3})^2} = \boxed{6.0 \times 10^{-4}\ \mathrm{T}}$ .

Step 3. Gap, full current: $B = \dfrac{2\times10^{-7} \times 12}{3.0\times10^{-3}} = \boxed{8.0 \times 10^{-4}\ \mathrm{T}}$ .

Step 4. Outside: enclosed current $12 - 12 = 0$ , so $\boxed{B = 0}$ — write the reason: equal and opposite currents thread the loop, so the circulation and, by symmetry, the field vanish.

Step 5. Peak: at the core surface, $B_{\max} = \dfrac{2\times10^{-7} \times 12}{2.0\times10^{-3}} = \boxed{1.2 \times 10^{-3}\ \mathrm{T}}$ .

Competitive level — inside the return shell (JEE-Advanced pattern, ~4 minutes)

Same cable ( $a = 2.0$ mm, $b = 6.0$ mm, $c = 8.0$ mm, $I = 12$ A each way). Find (a) the field at $s = 7.0$ mm, and (b) the two radii at which the field is exactly half its maximum value.

Step 1. Part (a) is a shell-room count. Fraction of the return already enclosed: $\dfrac{s^2 - b^2}{c^2 - b^2} = \dfrac{49 - 36}{64 - 36} = \dfrac{13}{28}$ . Net enclosed: $I_{\text{enc}} = 12\left(1 - \dfrac{13}{28}\right) = 12 \times \dfrac{15}{28} = 6.4$ A.

Step 2. Solve: $B = \dfrac{2\times10^{-7} \times 6.43}{7.0\times10^{-3}} = \boxed{1.8 \times 10^{-4}\ \mathrm{T}}$ — already down to $15\%$ of the peak, en route to zero at $c$ .

Step 3. Part (b): the peak is $1.2$ mT at $s = a$ . Half-peak on the ramp: $s = a/2 = \boxed{1.0\ \mathrm{mm}}$ ; on the $1/s$ slide: $s = 2a = \boxed{4.0\ \mathrm{mm}}$ — legal because $4.0$ mm is still inside the gap.

Step 4. Sanity: a third half-peak radius inside the shell? No — the shell field starts at $0.40$ mT ( $< 0.60$ mT) at $s=b$ and only falls. State it; setters plant a shell-region distractor for exactly this check. Timing note: four minutes, of which the physics is one — rooms, counts, solve. The annulus fraction $13/28$ is the only place to lose it; write areas, not radii.

Competitive level — the drilled conductor (JEE-Advanced pattern, ~4 minutes)

A long cylindrical conductor of radius $R = 8.0$ mm carries a uniform current density $J = 4.0 \times 10^{6}$ A/m² parallel to its axis. A cylindrical cavity of radius $2.0$ mm runs through it, parallel to the axis, its centre $d = 1.0$ mm from the conductor's axis. Find the magnetic field (magnitude and character) inside the cavity.

Step 1. Superpose: (full cylinder, density $J$ ) + (cavity cylinder, density $-J$ ). In the cavity the two cancel to the real zero current; each piece alone is perfectly symmetric.

Step 2. Quote the interior result (from Ampère's Circuital Law, via $I = J\pi R^2$ ): inside a uniform cylinder, $B = \dfrac{\mu_0 J s}{2}$ at distance $s$ from its own axis, wrapping around that axis.

Step 3. Add the two wrapped fields at any cavity point. The position-dependent parts cancel and only the axis offset survives:

B = \frac{\mu_0 J d}{2} = \frac{(4\pi\times10^{-7})(4.0\times10^{6})(1.0\times10^{-3})}{2} = \boxed{2.5 \times 10^{-3}\ \mathrm{T}}

Step 4. Character — where the real marks hide: the field is uniform across the entire cavity, perpendicular to the line joining the axes. Neither radius appears — options decorated with $r$ or $R$ are dead on arrival. Limit check: $d \to 0$ gives $B \to 0$ , the centred bore ✓. Timing note: four minutes if you must reconstruct the superposition; ninety seconds if $\mu_0 Jd/2$ is a landmark you carry. If the stem gives total current $I$ instead of $J$ , convert with the metal's actual area: $J = I/[\pi(R^2 - r^2)]$ — using $\pi R^2$ is the planted error.

Exam traps

Formula before region. The single biggest mark-loser: applying the gap formula inside the shell, or the full-current formula inside the core. Write the room first — " $s = 7$ mm: shell region" — then pick the count.

Circle areas where annulus areas belong. Inside a shell or pipe wall, the enclosed fraction is $\dfrac{s^2 - b^2}{c^2 - b^2}$ (ring inside $s$ over whole ring). Writing $s^2/c^2$ feels symmetric and is wrong.

Distance from the surface instead of the axis. Every $s$ in this archetype is measured from the axis. Stems love "at $1.0$ mm from the cable's outer surface" — convert before touching a formula.

The shielding misread. Zero outside a coax does not mean the shell blocks fields; it means enclosed currents cancel. In the gap the shell contributes nothing and the full field lives on.

Cavity problems fed total current. $J = I/(\pi R^2 - \pi r^2)$ — the current squeezes through the remaining metal. Feeding $\pi R^2$ to the formula understates every downstream number.

Losing the factor of 2 on the sheet. One sheet: $\mu_0 K/2$ each side. Solenoid (rolled sheet, layers reinforcing inside): $\mu_0 nI$ .

Jurisdiction and constants. These setups are JEE-Advanced culture; JEE-Main stops Ampère applications at wire + solenoid, and CBSE/ISC/NEET stay with the parent module's cases. Constants: $\mu_0 = 4\pi\times10^{-7}$ T·m/A, workhorse $\mu_0/2\pi = 2\times10^{-7}$ exactly.

Origins · 1 minOliver Heaviside and how this was discovered

Oliver Heaviside

1850–1925 · London and Devon

Oliver Heaviside left school at sixteen, taught himself Maxwell's Treatise as a telegraph operator, and in 1880 patented the coaxial cable — one conductor wrapped inside another so that signal currents run out and back in the same line, cancelling their external field and shutting interference out. The four-region field map of this archetype is the physics of his invention: all the magnetic action confined between core and shell, silence outside. (How signals rise and fall in such cables — his transient story — belongs to RC Circuits: Charging & Discharging; this module is his geometry.) Working alone in a Devon cottage, he compressed Maxwell's twenty equations into the four every physicist now writes; the establishment he bypassed elected the self-taught telegrapher a Fellow of the Royal Society in 1891.

History

1826

— Ampère's Mémoire completes the circulation law this archetype runs on.

1880

— Heaviside patents the coaxial cable: the four-region field profile becomes a communications technology.

1934

— The first commercial coaxial telephone line (New York–Philadelphia) carries hundreds of calls through a magnetically invisible pipe.

1962

— Coax carries the first live transatlantic television, the archetype at ocean scale.

Going further

Texts. HC Verma, Concepts of Physics Vol. 2, Ch. 35 (Ampère's law exercises — the pipe and cable problems) · Irodov, Problems in General Physics, §3.6 (the cavity problem in its classic form) · NCERT Physics XII Ch. 4 for the parent law.

Provenance. Fresh archetype module; the coax profile downshifted from the UG Ampère's-law canon, all numbers re-derived. A verifier should recompute: coax ( $I=12$ A, $a=2$ , $b=6$ , $c=8$ mm) — $0.60$ mT at $1$ mm, $1.2$ mT peak at $a$ , $0.18$ mT at $7$ mm (fraction $15/28$ ), zero outside, half-peaks at $a/2$ and $2a$ ; pipe ( $r_1=3$ , $r_2=5$ mm, $10$ A) — zero bore, $2.2\times10^{-4}$ T at $4$ mm ( $7/16$ ); sheet $K=400$ A/m → $2.5\times10^{-4}$ T; cavity $J=4.0\times10^6$ A/m², $d=1.0$ mm → $2.5$ mT.

Prerequisites. Ampère's Circuital Law · Solenoids & Toroids · Gauss's Law at Work: Wire, Sheet & Shell (the totals-law twin). Leads to. Direction Triage — Hands, Thumbs & Cross Products (closing this chapter) · Displacement Current: Maxwell's Fix (E8, when currents stop being steady) · Ampère's Law (UG deep dive, where the sheet's jump becomes a boundary condition).

Why is the field in the coax gap unaffected by the shell right next to it?

Because Ampère's law only ever counts current that threads your loop. A circle in the gap encloses the core's current and none of the shell's — so the shell contributes nothing to the circulation, and by symmetry nothing to the field there either. It is exact, the same statement as the electrostatic shell theorem of Gauss's Law at Work. The shell's moment comes only when your loop grows past $c$ and swallows the return current — then it erases everything.

How can the field outside the coax be exactly zero when the core and shell are at different distances from me?

Distance is precisely what the circulation law ignores. Any outside loop threads $+I$ and $-I$ : net zero. Symmetry then upgrades zero circulation to zero field — the field cannot vary around a centred loop, so a zero total forces a zero value. For a non-coaxial pair (two separate parallel wires) the net threaded current is also zero, but no symmetry legalises the collapse — and indeed the field outside two separated wires is not zero, as the census of Net Fields of Wire Systems showed. Coaxiality is what turns cancellation into invisibility.

In the shell region, why do I subtract the return current gradually instead of all at once?

Because your circle at radius $s$ has only caught the part of the return flowing through the ring between $b$ and $s$ ; uniform spreading makes that share $(s^2-b^2)/(c^2-b^2)$ . At $s = b$ none has returned; at $s = c$ all of it has, and the field dies. Subtracting the whole $I$ the moment you enter the shell would make the field jump discontinuously — and fields of finite current distributions never jump like that.

Where does the cavity result μ0Jd/2 actually come from, if no loop has the right symmetry?

From superposition doing the symmetry's job. The drilled conductor equals a full cylinder of density $J$ plus a cavity-shaped one of density $-J$ — together they reproduce the truth, and each piece separately has perfect cylindrical symmetry, with interior field $\mu_0 J s/2$ wrapped around its own axis. Adding a field wrapping one axis to an opposite field wrapping a shifted axis, the position-dependent parts cancel and a constant leftover survives — proportional to the shift $d$ . The law never failed; we fed it a decomposition it could digest.

Is the current sheet result of any use at HS level, or is it a UG import?

It earns its place three ways. It is a legal two-line Ampère solve — JEE-Advanced has dressed it as "a large flat strip carrying current" more than once. It is the fastest sanity check on the solenoid: roll the sheet and $\mu_0 K/2 + \mu_0 K/2 = \mu_0 nI$ inside, cancellation outside. And its distance-blindness is a favourite MCQ hook: options that decay with $z$ die instantly.

For the cavity problem, does the direction of B inside the hole matter for exams?

Yes, and it is cheap insurance: the field is perpendicular to the line joining the two axes and is the same arrow everywhere in the cavity. Setters test the uniformity more often — options like "stronger near the cavity wall" or "depends on r" are the planted kills. If a stem asks the direction, place the point on the axis line, run the right-hand grip rule from Fields of Straight Wires & Circular Loops on both pieces and subtract; that read becomes a ten-second drill in Direction Triage — Hands, Thumbs & Cross Products.

Why does JEE-Main not ask these if they are just the same law again?

Syllabus scope, stated verbatim: JEE-Main lists "Ampère's law and its applications to infinitely long current carrying straight wire and solenoid" — nothing multi-region. JEE-Advanced grants the law without that fence, and its papers use the freedom: cables, pipes and cavities are where the recipe's judgement — region identification, honest counting, superposition — can actually be examined. Sitting only Main/NEET? Own the parent module cold and treat this gym as insurance. Sitting Advanced? This archetype is core training.