The Hall Effect

Every ammeter you have ever used answers "how much charge flows?" None of them can answer a more basic question: what is flowing? Positive charge drifting left and negative charge drifting right produce identical currents, identical heating, identical everything a circuit meter can see. For most of the nineteenth century, the sign of the moving charge in a wire was simply unknowable — Maxwell himself wrote that the magnetic force acts on the wire, not on whatever moves inside it, and that seemed to close the case. In 1879 a 24-year-old graduate student, Edwin Hall, decided the great man's footnote deserved a test. He passed a current along a thin gold leaf, placed it face-on to a strong magnet, and looked for a voltage across the strip — sideways, where no battery was pushing. He found it: a few microvolts, steady, reversing when either the current or the field reversed. The magnetic force was acting on the moving charges themselves, sweeping them toward one edge until the piled-up charge pushed back. And the polarity of that tiny voltage revealed, eighteen years before the electron was discovered, that the movers in a metal are negative. The Hall effect is A-Level territory — CIE 9702 asks you to derive and use $V_H = \dfrac{BI}{ntq}$ — and the derivation is a beautiful two-line collision of things you already own: the magnetic force on a moving charge and the microscopic picture of current. The payoff is a formula that counts charge carriers per cubic metre from a bench measurement, explains why sensors are made of semiconductors rather than copper, and puts a magnetic-field meter inside every phone.

Key takeaways

Pass current through a flat slice with a magnetic field through its face: the force $F = Bqv$ pushes drifting charges toward one edge, which charges up until its electric field exactly cancels the push.
The balance $qE = Bqv$ leaves a steady sideways potential difference — the Hall voltage — across the slice: $V_H = \dfrac{BI}{ntq}$ .
$t$ is the slice's dimension along the field; the width across which you measure cancels out entirely.
The polarity of $V_H$ tells you the sign of the mobile carriers — the experiment that proved metal currents are carried by negative charges, and today's bench test for n-type vs p-type semiconductors.
Because $V_H \propto 1/n$ , sparse-carrier semiconductors give millivolts where copper gives microvolts — the entire Hall-sensor industry rests on that ratio.
Key idea: a current-carrying slice in a magnetic field becomes its own sideways voltmeter, reporting the sign and the number density of whatever carries its current.

The core insight

In a current-carrying conductor with a magnetic field through its face, drifting carriers are pushed sideways by $F = Bqv$ until accumulated edge charge builds an electric force that balances it; the resulting Hall voltage $V_H = BI/(ntq)$ measures the carriers' sign and number density, and — once calibrated — the field itself.

Where this lives today

Your phone's compass is three Hall slices reading the Earth's field along three axes. Every brushless motor — drones, e-bikes, electric cars, laptop fans — carries Hall switches that tell the controller where the spinning magnets are (the electronic commutation that The DC Motor & Its Cousins previewed). Clamp meters measure a wire's current without touching the circuit by funnelling the wire's own magnetic field through a Hall element — the formula run backwards, with $B$ as the unknown. And in the fab, a Hall measurement of $n$ on each wafer batch is still the fastest honest check that the doping went in as designed. The deepest afterlife is metrological. Cooled to a few kelvin in a strong field, a two-dimensional electron sheet shows Hall resistance locked to exact quantized plateaus, reproducible to parts per billion regardless of dirt or shape — and since 2019 the SI's electrical units rest on constants that make this quantized Hall resistance exact. A graduate student's microvolt of 1879, promoted to the bedrock of the ohm.

The setup

Take a thin rectangular slice of conducting material. Give its three dimensions jobs: current $I$ flows along its length; the magnetic field $B$ passes through its face, along the thickness $t$ ; and the width $d$ runs across, from one long edge to the other. A voltmeter connects the two long edges. That's the whole apparatus — the physics does the rest.

Figure

The Hall slice — current along, field through the face, and the sideways pile-up that becomes a voltage

Now run the film from the moment the current switches on. Each carrier drifts along the slice at the drift velocity $v$ — the slow shuffle Drift Velocity & the Microscopic Picture quantified as $I = nAvq$ . Because it moves through a magnetic field, each feels the sideways force Force on a Moving Charge: qv×B established — magnitude $Bqv$ , since drift and field are perpendicular. Carriers start landing on one long edge. But the pile-up is self-defeating — and that is the elegance. Charge accumulating on one edge (and the matching charge deficit on the other) builds an electric field $E$ across the slice, which pushes later carriers back toward the middle. Every arriving carrier strengthens the push-back on the next. Within a vanishingly small fraction of a second the system finds its truce: the electric force exactly cancels the magnetic force, carriers resume drifting dead straight, and the edge charge stops growing. What remains is a permanent, steady sideways potential difference — the Hall voltage — quietly reporting the balance inside.

Deriving the Hall voltage

This is the derivation CIE 9702 names explicitly ("derive and use $V_H = BI/(ntq)$ "): two known results collide, and every unmeasurable microscopic quantity cancels.

Derivation AFrom the force balance to the Hall voltage formula

We derive $V_H = BI/(ntq)$ for a slice of width $d$ and thickness $t$ carrying current $I$ with field $B$ through its face. The strategy: write the equilibrium condition (electric force = magnetic force on one drifting carrier), convert the electric field to the measured voltage across the width, then eliminate the unmeasurable drift velocity using the microscopic current equation. Watch two things: which dimension survives to the final formula, and how the drift velocity — the one quantity no bench instrument reads — cancels completely.

Step 1. Balance the two forces on a drifting carrier at equilibrium. Magnetic: $F = Bqv$ (drift perpendicular to field). Electric, from the edge-charge field $E$ : $F = qE$ . In the steady state they cancel:

qE = Bqv \quad\Rightarrow\quad E = Bv

Why equality and not merely similarity: while the magnetic force is even slightly ahead, carriers keep landing on the edge and E keeps growing; the moment E overshoots, the surplus pushes carriers back. The system self-corrects to exact balance — negative feedback in a slab of metal.

Step 2. Convert field to voltage. The Hall field is uniform across the slice, and field times distance gives potential difference across the width $d$ :

V_H = E\,d = Bvd

where: $d$ — the width between the voltmeter's contact edges. So far the formula contains $v$ , which no meter reads — we trade it away next.

Step 3. Bring in the microscopic current equation from Drift Velocity & the Microscopic Picture: $I = nAvq$ , where $n$ is the number of carriers per cubic metre and $A$ is the cross-section the current flows through. That cross-section is width times thickness, $A = dt$ , so:

v = \frac{I}{n\,d\,t\,q}

Step 4. Substitute into $V_H = Bvd$ and watch the width cancel:

V_H = \frac{B\,I\,d}{n\,d\,t\,q} = \frac{BI}{ntq}

where: $B$ — flux density through the face; $I$ — the current; $n$ — carrier number density; $t$ — the slice's thickness measured along $B$ ; $q$ — the charge of one carrier ( $q = e = 1.6\times10^{-19}\ \mathrm{C}$ for electrons). The width $d$ has vanished: making the slice wider spreads the same current over more, slower carriers, weakening the field by exactly the factor the longer measuring distance regains. Only the dimension parallel to the field matters — and it sits in the denominator, so thin is loud: halve $t$ and you double the signal. Hall used gold leaf in 1879 for precisely this reason. Verified: semiconductor slice, $n = 1.5\times10^{23}\ \mathrm{m^{-3}}$ , $t = 0.10\ \mathrm{mm}$ , $I = 5.0\ \mathrm{mA}$ , $B = 0.12\ \mathrm{T}$ : $V_H = (0.12\times5.0\times10^{-3})/(1.5\times10^{23}\times1.0\times10^{-4}\times1.6\times10^{-19}) = 2.5\times10^{-4}\ \mathrm{V}$ . Cross-check by the long route: $v = I/(ndtq) = 0.52\ \mathrm{m/s}$ (with $d = 4.0\ \mathrm{mm}$ ), $E = Bv = 0.0625\ \mathrm{V/m}$ , $V_H = Ed = 2.5\times10^{-4}\ \mathrm{V}$ — identical, as the cancellation demands.

Pause on what the equation is made of. $B$ , $I$ , $t$ : things you set or measure with a ruler. $q$ : a fundamental constant. That leaves one unknown, $n$ — a Hall measurement counts the charge carriers per cubic metre. Current, magnet, voltmeter: no microscope needed. This is why the Hall effect graduated from curiosity to the semiconductor industry's standard census tool.

The sign verdict — what made 1879 historic

Here is the subtlety that turns a small voltage into a deep measurement. You might expect that swapping positive carriers for negative ones would just move the pile-up to the opposite edge and change nothing measurable. Work it through carefully and the opposite happens. Fix the conventional current direction (left to right, say) and the field (into the face). Case one: the carriers are positive, drifting left-to-right with the current. The force $Bqv$ on them points — apply the perpendicular-force rule from Force on a Moving Charge: qv×B — toward, say, the top edge. Positive charge accumulates on top; the top edge sits at higher potential. Case two: the carriers are negative (electrons), which must drift right-to-left to make the same left-to-right conventional current. Two things flipped: the charge sign and the velocity direction. In the force $Bqv$ the two flips cancel — the force pushes electrons toward the same top edge. But now the charge that lands there is negative: the top edge sits at lower potential. Same slab, same current, same field, same deflected edge — opposite voltmeter reading.

Figure

Same current, same field, same deflection edge — but the polarity betrays the carrier

So the deflection side is fixed by $I$ and $B$ , but the sign of what piles up there is the carriers' own. The polarity of the Hall voltage is a bulk, room-temperature measurement of the sign of the mobile charge — the only simple one there is. Hall's polarity came out negative for metals: the movers are negative, eighteen years before J.J. Thomson identified them as electrons. And when some materials (zinc among the metals, p-type semiconductors by design) gave the positive polarity, the Hall effect had flagged the existence of positive "holes" — the concept half of semiconductor electronics is built on.

Metal or semiconductor: the 1/n economics

The formula's sharpest practical lesson sits in the denominator: $V_H \propto 1/n$ . Copper packs $n = 8.5\times10^{28}$ free electrons per cubic metre; a typical doped semiconductor might have $n = 1.0\times10^{22}$ — nearly seven million times fewer. Same current, same field, same geometry: the semiconductor's Hall voltage is millions of times larger. Physically, sparse carriers must drift fast to carry the same current, and faster drift means a stronger magnetic push needing a bigger balancing field. Run the honest numbers. A copper strip, $t = 0.20\ \mathrm{mm}$ , carrying a hefty $8.0\ \mathrm{A}$ in a strong $0.50\ \mathrm{T}$ field: $V_H = BI/(ntq) = (0.50\times8.0)/(8.5\times10^{28}\times2.0\times10^{-4}\times1.6\times10^{-19}) \approx 1.5\ \mathrm{\mu V}$ . Microvolts, from eight amps and half a tesla — invisible without careful instrumentation, which is why nobody trips over the Hall effect by accident. The semiconductor probe in the derivation above: a modest $5.0\ \mathrm{mA}$ and $0.12\ \mathrm{T}$ already gave $0.25\ \mathrm{mV}$ — a hundred times larger from a thousandth of the current. Every commercial Hall sensor is a semiconductor for this one arithmetical reason.

case	consequence	reading
$B \to 0$	$V_H \to 0$ in proportion	calibrated slice = magnetic field meter
thinner slice ( $t\downarrow$ )	$V_H \uparrow$ as $1/t$	thin films and gold leaf; "thin is loud"
metal ( $n \sim 10^{29}$ )	microvolt signals	great wires, terrible sensors
doped semiconductor ( $n \sim 10^{22}$ )	millivolt signals	the sensor material of choice
positive carriers	$V_H$ polarity reverses	n-type vs p-type bench test
field parallel to the face	no sideways force, $V_H = 0$	probe must face the field — rotate for maximum

Worked examples

Board level — derive, then read a probe (CIE 9702 structured pattern)

A Hall probe uses a slice of doped semiconductor with carrier density $n = 1.5\times10^{23}\ \mathrm{m^{-3}}$ , thickness $t = 0.10\ \mathrm{mm}$ along the field, and width $4.0\ \mathrm{mm}$ . The probe current is $5.0\ \mathrm{mA}$ . (a) Show that the Hall voltage is given by $V_H = BI/(ntq)$ . (b) Calculate $V_H$ when the probe sits in a field of $0.12\ \mathrm{T}$ . (c) State and explain what happens to the reading if the probe is rotated so its face is parallel to the field.

Step 1. (a) Reproduce the derivation's spine — this is a "show that," so every link earns a mark: force balance $qE = Bqv$ ; voltage across the width $V_H = Ed = Bvd$ ; drift velocity from $I = n(dt)vq$ ; substitute and cancel $d$ : $V_H = BI/(ntq)$ . Four lines, four marks.

Step 2. (b) Substitute with everything in SI units ( $t = 1.0\times10^{-4}\ \mathrm{m}$ ): $V_H = \dfrac{0.12 \times 5.0\times10^{-3}}{1.5\times10^{23} \times 1.0\times10^{-4} \times 1.6\times10^{-19}} = \dfrac{6.0\times10^{-4}}{2.4\times10^{0}}$ . So $\boxed{V_H = 2.5\times10^{-4}\ \mathrm{V} = 0.25\ \mathrm{mV}}$ — comfortably measurable, because the material is a semiconductor.

Step 3. (c) Rotated face-parallel to $B$ , the carriers drift along the slice while the field lies in the plane of the face: the component of field perpendicular to the face is zero, no sideways force acts, and $V_H$ falls to zero. In practice you rotate a Hall probe until the reading is a maximum — that orientation puts the face perpendicular to the field and reads the full flux density.

Competitive level — the mystery slab, plus the copper contrast (9702 A2 hardest-paper pattern)

(a) A copper strip ( $n = 8.5\times10^{28}\ \mathrm{m^{-3}}$ ) of thickness $0.20\ \mathrm{mm}$ carries $8.0\ \mathrm{A}$ with $B = 0.50\ \mathrm{T}$ through its face. Find $V_H$ and comment on why copper makes a poor Hall sensor. (b) A slab of unknown semiconductor, $t = 50\ \mathrm{\mu m}$ , carries $2.5\ \mathrm{mA}$ in $B = 0.60\ \mathrm{T}$ and shows $V_H = 3.9\ \mathrm{mV}$ , with the polarity indicating positive carriers. Determine the carrier type and number density.

Step 1. (a) Copper: $V_H = \dfrac{0.50 \times 8.0}{8.5\times10^{28} \times 2.0\times10^{-4} \times 1.6\times10^{-19}} = \dfrac{4.0}{2.72\times10^{6}}$ . So $\boxed{V_H \approx 1.5\ \mathrm{\mu V}}$ — despite eight amps and half a tesla. Copper's enormous $n$ buries the signal; a meter reading microvolts reliably needs shielding and patience. Comment sentence: the carrier density of a metal is so large that the drift velocity, and hence the Hall voltage, is tiny; low- $n$ semiconductors give signals a million times larger.

Step 2. (b) Carrier type first — it costs no algebra: positive polarity ⇒ the mobile carriers behave as positive charges ⇒ p-type material (conduction by holes).

Step 3. Rearrange the Hall formula for the census: $n = \dfrac{BI}{V_H\,t\,q} = \dfrac{0.60 \times 2.5\times10^{-3}}{3.9\times10^{-3} \times 50\times10^{-6} \times 1.6\times10^{-19}}$ .

Step 4. Evaluate: numerator $1.5\times10^{-3}$ ; denominator $3.12\times10^{-26}$ . So $\boxed{n \approx 4.8\times10^{22}\ \mathrm{m^{-3}}}$ — six orders sparser than copper, exactly the regime that makes millivolt signals possible. Sanity check by rebuilding: $V_H = BI/(ntq) = 1.5\times10^{-3}/(4.8\times10^{22}\times5.0\times10^{-5}\times1.6\times10^{-19}) = 3.9\ \mathrm{mV}$ ✓. Timing note: 5–6 minutes as a full structured question. The arithmetic is powers-of-ten bookkeeping — write numerator and denominator separately before dividing, and carry the units. The free marks are the two sentences: why copper fails, and polarity ⇒ carrier sign.

Exam traps

The $t$ mix-up. In $V_H = BI/(ntq)$ , $t$ is the dimension along the field — through the face — not the width you measure the voltage across. The width cancelled in the derivation. Questions deliberately quote all three dimensions; pick $t$ by asking "which way does $B$ point?"

Unit slips in the powers-of-ten storm. Millimetres and micrometres must become metres, milliamps become amps, before substitution. With $10^{22}$ -scale densities and $10^{-19}$ -scale charges in one fraction, a single stray $10^{-3}$ is invisible in the working and fatal to the answer. Compute numerator and denominator separately.

Deriving with the wrong cross-section. In $I = nAvq$ , the area is the one the current flows through: $A = dt$ (width × thickness), never the face area (length × width). Wrong $A$ , no cancellation, no marks for the "show that."

Polarity reasoning half-done. "Electrons deflect to the other edge" is the classic wrong turn. Both carrier signs deflect to the same edge (two sign flips cancel in $Bqv$ ); what changes is the sign of the charge that accumulates there. Practise the two-case argument until it is mechanical.

Probe orientation. A Hall probe reads the component of $B$ perpendicular to its face. Rotate until the reading peaks — that maximum is the field magnitude. Forgetting this costs the practical-paper mark, and "reading less than expected" questions usually mean a tilted probe.

Which boards ask this. The Hall effect is a CIE 9702 syllabus item (20.3.3: derive and use $V_H = BI/(ntq)$ ). JEE, NEET, CBSE and IB do not list it — if you sit those, this module is enrichment that deepens Force on a Moving Charge rather than exam-required material. A-Level candidates: it appears in both theory and practical-context questions.

Constants: $e = 1.6\times10^{-19}\ \mathrm{C}$ ; carrier densities are given in questions (copper's $8.5\times10^{28}\ \mathrm{m^{-3}}$ is worth recognising on sight).

Origins · 1 minEdwin Herbert Hall and how this was discovered

Edwin Herbert Hall

1855–1938 · Johns Hopkins, then Harvard — the student who out-measured Maxwell's footnote

Hall was a graduate student in Henry Rowland's Baltimore laboratory when he read, in Maxwell's Treatise, that the force on a wire in a magnetic field acts on the wire itself, "not on the electric current." If that were true, the moving charges inside would feel nothing, and no experiment inside the metal could detect the field. Rowland doubted it; Hall built the test. Thick wires showed nothing — the effect hides as $1/(nt)$ , and bulk metal buries it twice over. The experimental insight was to go thin: gold leaf, micrometres thick, where the signal is thousands of times larger. In October 1879 his galvanometer registered a steady transverse voltage, flipping with current and with field — the magnetic force acts on the moving charges, and their pile-up is measurable. The polarity said the carriers were negative, before anyone knew what an electron was; the anomalous metals that read positive seeded the idea of holes. Hall spent his career at Harvard on conduction physics, and lived to see his thesis measurement become the semiconductor age's standard instrument — though not to see its quantum version define the ohm itself.

History

1879

— Edwin Hall, testing Maxwell's claim in Rowland's lab, measures a steady transverse voltage in gold leaf

1897

— J.J. Thomson identifies the electron; Hall's negative polarity acquires its particle

1929

— Band theory explains the "anomalous" positive metals: nearly full bands conduct as positive holes

1948

— The transistor arrives; Hall measurements become the standard carrier census of the semiconductor fab

1980

— von Klitzing discovers the Hall resistance of a cold 2D electron gas is quantized — the effect becomes a resistance standard exact enough to define the ohm

Going further

Texts. CIE 9702 syllabus §20.3 (derive and use $V_H = BI/(ntq)$ — the exact assessed wording) · Sang, Jones, Chadha & Woodside, Cambridge International AS & A Level Physics, magnetic fields chapter · Duncan, Advanced Physics, for the classic sign-of-carriers discussion.

Provenance. All numbers recomputed: probe $V_H = (0.12\times5.0\times10^{-3})/(1.5\times10^{23}\times1.0\times10^{-4}\times1.6\times10^{-19}) = 0.25\ \mathrm{mV}$ , cross-checked via $v = 0.52\ \mathrm{m/s}$ , $E = 0.0625\ \mathrm{V/m}$ ; copper $V_H = 4.0/(2.72\times10^{6}) = 1.5\ \mathrm{\mu V}$ ; mystery slab $n = 1.5\times10^{-3}/3.12\times10^{-26} = 4.8\times10^{22}\ \mathrm{m^{-3}}$ with rebuild check $3.9\ \mathrm{mV}$ ; quiz value $V_H = (0.30\times2.0\times10^{-3})/(5.0\times10^{22}\times2.5\times10^{-5}\times1.6\times10^{-19}) = 3.0\ \mathrm{mV}$ .

Prerequisites. Force on a Moving Charge: qv×B (the sideways push) · Drift Velocity & the Microscopic Picture (the $I = nAvq$ census). Leads to. The Hall Effect (UG deep dive) — the Hall coefficient, carrier mobility and the quantum Hall story · Semiconductors (future track) — holes, doping and band theory.

Why does the width of the slice cancel out? It feels like it should matter — the voltage is measured across it.

Two opposing effects tie exactly. Widen the slice at fixed current: the cross-section $dt$ grows, so the drift velocity $v = I/(n\,dt\,q)$ drops in proportion — each carrier moves slower, feels a weaker magnetic push, and the balancing field $E = Bv$ is weaker. But you now measure that weaker field across a proportionally longer span, $V_H = Ed$ . The product $Ed$ is unchanged. Physically: the same total current spread thinner produces the same edge-to-edge tally. Only the thickness along $B$ escapes this cancellation, which is why it alone appears in the formula.

Is the Hall voltage a source of EMF? Could I power something from the side terminals?

No — it is an electrostatic potential difference maintained by a fixed arrangement of edge charge, not a seat of energy conversion. Try to draw a sideways current and you remove the very charges whose field maintains the balance: the voltage sags immediately, and the trickle of energy you extract is stolen from the battery driving the main current (the magnetic force itself does no work — it only redirects). This is why Hall probes are read with high-resistance voltmeters: the measurement is honest only while essentially no sideways current flows.

How does a Hall probe actually measure B? The formula needs n, which nobody knows at the bench.

By calibration, which buries $n$ inside one constant. For a given probe at fixed current, everything except $B$ is fixed: $V_H = (\text{const})\times B$ . Place the probe in a field of accurately known strength (a standard magnet or calculated solenoid field — Solenoids & Toroids explains why a solenoid's interior is the calibration field of choice), record the reading, and the constant is known forever after. Commercial gaussmeters do this at the factory. Exam questions occasionally reverse this: given a calibration reading, find the unknown field by simple proportion.

Do both electrons AND holes exist in a semiconductor? What does the probe read then?

In doped material one species vastly outnumbers the other — that is what doping is for — so the Hall voltage reports the majority carrier cleanly: n-type reads negative-carrier polarity, p-type reads positive. In nearly pure (intrinsic) material, both species conduct at once and their Hall contributions partially cancel; the measured voltage then depends on which carriers are more mobile, not just more numerous. A-Level questions stay in the cleanly doped regime, but the two-carrier tug-of-war is a good reminder that $V_H = BI/(ntq)$ assumes a single carrier type.

Why did it take until 1879 to notice this effect?

Because in metals it is genuinely tiny. The Hall voltage scales as $1/(nt)$ : metals have colossal $n$ , and pre-1879 experimenters used thick wires — microvolts at best, below casual apparatus noise. Hall's contribution was experimental strategy as much as theory: suspect the effect, then engineer the geometry (gold leaf — minimal $t$ ) to amplify it into measurability. A standing lesson: a null result sometimes means the effect is real but your sample shape is fighting you.

Which way do the electrons deflect — I keep getting the edge wrong.

Use conventional current plus the perpendicular-force rule, then patch the sign at the end. Find the force direction for a positive carrier moving with the conventional current (Fleming's left hand or the qv×B rule from Force on a Moving Charge). Electrons — moving oppositely with opposite charge — feel the force toward the same edge (the two reversals cancel). So: the deflection edge is whatever your positive-carrier analysis says, always. The only thing electrons change is the sign of the charge arriving there, hence the polarity of the voltage. Never re-run the geometry for electrons; just flip the final polarity.