The Electrostatic Potential

A pigeon lands on a bare 100,000-volt transmission line and calmly preens its feathers. A lineworker in an insulated bucket touches the same cable, bare-handed, and is equally fine. Yet a fallen wire brushing wet grass can kill from metres away. The quantity that sorts out these three fates is not the field and not the charge — it is the electric potential, and specifically the fact that only potential differences move charge and deliver energy. Both of the bird's feet sit at the same potential, so no current crosses its body; the grass, at nearly zero volts, offers a lethal difference. You arrive at this idea equipped with the electric field — a vector at every point of space, three numbers each, directions to track everywhere. The potential is your reward for having learned that the electrostatic field is curl-free: the work it does between two points doesn't care about the path. That single fact lets you compress the entire vector field into one scalar landscape $V(\mathbf{r})$ — a height map for charge — and recover the field whenever you want it by asking how steeply the landscape tilts. Mastering the potential buys you the working language of real electricity. Batteries, voltmeters, EKGs, neurons, and every circuit you will ever analyze speak in volts, not newtons per coulomb — and the machinery you build here becomes Poisson's equation, the master equation of electrostatics and the subject of the next two articles.

Key takeaways

The potential at a point is the work per unit charge to bring a test charge there from a reference point: $V(\mathbf{r}) = -\int_{\mathcal{O}}^{\mathbf{r}} \mathbf{E} \cdot \mathrm{d}\mathbf{l}$ , measured in volts ( $1\ \mathrm{V} = 1\ \mathrm{J/C}$ ).
It exists because $\nabla\times\mathbf{E}=0$ — the line integral is path-independent, so "the potential at a point" is a single well-defined number.
The field is recovered by $\mathbf{E} = -\nabla V$ : the field points downhill, from high potential to low, and a positive charge rolls that way.
Potential is a scalar, so contributions simply add as numbers: $V = \frac{1}{4\pi\varepsilon_0}\sum_i q_i / r_i$ — no components, no directions.
Only differences matter; the zero of potential is yours to choose (infinity for localized charges, "ground" in circuits) — but a source that itself extends to infinity forces the reference to a finite point.
Key idea: one scalar landscape replaces three field components — compute $V$ by easy scalar addition, then differentiate once to get $\mathbf{E}$ back.

The core insight

The curl-free electrostatic field admits a scalar potential $V$ defined by $V(\mathbf{r}) = -\int_{\mathcal{O}}^{\mathbf{r}}\mathbf{E}\cdot\mathrm{d}\mathbf{l}$ , from which the field is recovered as $\mathbf{E}=-\nabla V$ — working with one scalar instead of three field components, superposed by ordinary addition, makes most electrostatic calculations far simpler.

Where this lives today

The potential is the native language of every measurement of living tissue. An EKG is nothing but a map of millivolt-scale potential differences on the skin, sampled at a dozen electrodes and driven by the heart's wandering dipole; an EEG reads microvolts through the skull; and modern brain–computer interfaces stake everything on resolving those tiny $\Delta V$ 's against noise. Every neuron in your head is a battery maintaining about $-70\ \mathrm{mV}$ across its membrane, and thought itself is the propagation of controlled breakdowns in that potential. At the other end of the scale, electric-vehicle platforms migrated from 400 V to 800 V architectures precisely because of equation (6): doubling $\Delta V$ delivers the same power at half the current, cutting resistive losses and charging times. And in accelerator physics the potential has become a unit — particle energies are quoted in electron-volts, from the 500 eV beam of a desktop electron gun to the tera-electron-volts of the LHC: the humble volt, multiplied by thirteen orders of magnitude.

The physical picture

Think of $V$ as altitude. Around every charge distribution, space is a landscape of electrical hills and valleys: positive charges sit on peaks they have built, negative charges dig hollows. A positive test charge behaves exactly like a ball on this terrain — it rolls downhill, from high potential toward low, picking up kinetic energy as the landscape drops. The electric field is the slope: $\mathbf{E} = -\nabla V$ points in the direction of steepest descent, and its magnitude is how steep the descent is. Equipotential surfaces are the contour lines of the map. Walking along one costs no work, just as strolling around a hillside at constant altitude costs no climbing — which is why field lines must pierce equipotentials at right angles: any parallel component would mean the "level" path wasn't level. Where the contours crowd together the terrain is steep and the field is strong; where they spread out, the field is weak. A conductor in equilibrium is a perfectly flat plateau — one value of $V$ throughout, zero field inside, which is precisely why the charges in it have stopped moving.

Figure

The potential landscape of a point charge — contours, field lines, and the 1/r hill

One more reading of the landscape: since only slopes are physical, the whole terrain can be raised or lowered by any constant without changing a single force. That freedom — choose your own sea level — is why the pigeon survives. Its perch may be at 100 kV relative to the ground, but relative to itself the wire is flat. And one more picture before any calculus, because it is the simplest landscape there is. Between two charged plates the terrain is a perfect ramp: equipotentials evenly spaced like the contour lines of a constant grade, and the field the same everywhere — just the voltage across divided by the distance between.

Figure

Uniform field between plates — evenly spaced equipotentials, constant downhill slope

Hold these two terrains side by side — the $1/r$ hill of a point charge, the constant ramp of a capacitor gap — and you already own the intuition the algebra below formalizes: the field is never the value of the potential, always its slope.

The definition, and how to use it

Start with the question the landscape metaphor begs: how do you assign an altitude to every point of space, when what you actually know is the field? Take a unit positive test charge, carry it from an agreed reference point $\mathcal{O}$ to the point $\mathbf{r}$ you care about, and total up the work you do against the field along the way. Derivation A below shows that the answer cannot depend on the route you took — so the number belongs to the destination alone, and deserves a name:

V(\mathbf{r}) = -\int_{\mathcal{O}}^{\mathbf{r}} \mathbf{E} \cdot \mathrm{d}\mathbf{l} \qquad (1)

where: $V$ — electrostatic potential (volts, J/C); $\mathbf{E}$ — the electric field; $\mathcal{O}$ — the reference point where $V \equiv 0$ ; $\mathrm{d}\mathbf{l}$ — a step along any path from $\mathcal{O}$ to $\mathbf{r}$ . The minus sign encodes "work done against the field": marching against $\mathbf{E}$ costs energy and raises $V$ , which is why positive charges sit on peaks. Note also what the definition leaves free — the choice of $\mathcal{O}$ . Shift the reference and every $V$ in the problem shifts by the same constant; nothing physical moves, because only differences will ever matter. Equation (1) compresses three field components into one scalar, so there must be an inverse operation that recovers the field when a force is what you actually need. Differentiation undoes integration, and the three-dimensional derivative of a scalar is the gradient: ask at each point which way the landscape climbs fastest and how steeply, then flip the sign to point downhill:

\mathbf{E} = -\nabla V \qquad (2)

where: $\nabla V$ — the gradient, pointing toward increasing $V$ ; the minus sign turns $\mathbf{E}$ downhill, from high potential to low. This is the slope-reading of the contour map: field lines pierce equipotentials at right angles because the gradient is always perpendicular to the level curves. Its one-dimensional shadow, $E_x = -\mathrm{d}V/\mathrm{d}x$ , gives the capacitor-gap rule $E = \Delta V/d$ — and it says volts per metre and newtons per coulomb are the same unit, a conversion you get for free. Now feed the simplest possible source through definition (1). A point charge $Q$ makes the Coulomb field, and integrating that field in from infinity — the natural reference for a localized charge, and a four-line calculation done honestly in Derivation B — gives the hill the charge builds around itself:

V(r) = \frac{1}{4\pi\varepsilon_0}\frac{Q}{r} = \frac{k_e Q}{r} \qquad (3)

where: $r$ — distance from the charge; $k_e = 1/4\pi\varepsilon_0 = 8.99\times10^9\ \mathrm{N\,m^2/C^2}$ ; note the first power of $r$ — energies integrate one power up from forces. Compare it with Coulomb's $1/r^2$ and read the difference physically: the potential dies off more slowly than the field, because it remembers all the work accumulated coming in from infinity. Positive $Q$ builds the $k_eQ/r$ peak drawn in the hero figure; negative $Q$ digs the mirror-image hollow. The real prize appears when there are many charges. Their fields superpose as vectors — components, angles, partial cancellations, three equations per point. Their potentials superpose as plain signed numbers, because work is a scalar and works simply add:

V(\mathbf{r}) = \frac{1}{4\pi\varepsilon_0}\sum_i \frac{q_i}{r_i} \qquad (4)

where: $r_i$ — distance from charge $q_i$ to the field point. No unit vectors, no components — this is why the potential is the lazy physicist's route to the field: add the easy scalars first, then take one gradient at the end, instead of wrestling $N$ vectors from the start. And the sum begs to be promoted: smear the charges into a continuous distribution $\rho(\mathbf{r}')$ and the sum becomes an integral over charge elements,

V(\mathbf{r}) = \frac{1}{4\pi\varepsilon_0}\int \frac{\rho(\mathbf{r}')}{|\mathbf{r}-\mathbf{r}'|}\,\mathrm{d}\tau' \qquad (5)

where: $\mathbf{r}'$ — the source point being integrated over; $\mathbf{r}$ — the field point where you want $V$ ; $\mathrm{d}\tau'$ — a source volume element. This master formula — the workhorse that handles rings, disks, spheres, and every distribution in the problem sets — gets an article of its own next (Potential of a Localized Charge Distribution), where it is built honestly and put to work. Here, register the shape: one scalar integral, no vectors anywhere. Finally, the payoff line — the reason all this machinery earns its keep. Multiply a potential difference by the charge that crosses it and you have an energy; this is the equation batteries are built to enforce, the one an EKG reads off the skin, and the one every circuit you will ever meet runs on. For a charge $q$ carried from $a$ to $b$ , the work done by an external agent is:

W = q\,[V(b) - V(a)] = q\,\Delta V \qquad (6)

where: $\Delta V$ — the potential difference, the thing a voltmeter measures and a battery maintains. Notice that the reference point $\mathcal{O}$ has dropped out entirely — only the difference survives, which is the pigeon's salvation made quantitative. The sign carries the physics: positive $W$ means you paid energy to move the charge; negative means the field did the pushing, as it does for the electron in worked example 2. Set $q$ to one electron charge and $\Delta V$ to one volt, and (6) mints the electron-volt, the currency of atomic and accelerator physics. One table to organize the toolkit — each equation is a different bridge between source, landscape, field, and energy:

equation	bridge	reach for it when
$V = -\int \mathbf{E}\cdot \mathrm{d}\mathbf{l}$	field → potential	you already know $\mathbf{E}$ (say, from Gauss's law) and want the landscape
$\mathbf{E} = -\nabla V$	potential → field	you have $V$ and need a force, a field, or a direction of motion
$V = \frac{1}{4\pi\varepsilon_0}\sum_i q_i/r_i$	point sources → potential	you know where the charges are — the fast scalar route
$V = \frac{1}{4\pi\varepsilon_0}\int \rho\,\mathrm{d}\tau'/\|\mathbf{r}-\mathbf{r}'\|$	spread-out sources → potential	continuous distributions — the next article's subject
$W = q\,\Delta V$	potential → energy	the question mentions work, speed, or energy gained

Derivation AWhy the potential at a point is well defined

The goal is the license behind equation (1): to show that the line integral from $\mathcal{O}$ to $\mathbf{r}$ gives the same number along every path, so that " $V$ at a point" is a genuine function of position and not a diary of the route taken. The strategy is to convert path-independence into a statement about closed loops, and then spend the one deep fact we own about electrostatics — $\nabla\times\mathbf{E}=0$ , established honestly (point charge plus superposition plus Stokes) in the previous article, The Curl of the Electrostatic Field. Watch where that input enters: everything else is bookkeeping.

Step 1. Start from the loop statement. For any electrostatic field, $\oint \mathbf{E}\cdot \mathrm{d}\mathbf{l} = 0$ around every closed loop. Why: this is $\nabla\times\mathbf{E}=0$ in integral clothing via Stokes' theorem — the previous article's headline result, recalled here, not re-proved.

Step 2. Take two paths. Any two routes, path 1 and path 2, from $\mathcal{O}$ to $\mathbf{r}$ , run one forward and one backward, together form a closed loop. Why: out along path 1 and back along path 2 returns to the start — that is all a loop is.

Step 3. Subtract. Zero circulation around that loop means $\int_{(1)} \mathbf{E}\cdot \mathrm{d}\mathbf{l} - \int_{(2)} \mathbf{E}\cdot \mathrm{d}\mathbf{l} = 0$ , so both paths give the same integral. Why: reversing a path flips the sign of its line integral, so the loop integral is exactly the difference of the two path integrals.

Step 4. Conclude. The integral in (1) depends only on the endpoints, so $V(\mathbf{r})$ is a single-valued function of position — a genuine landscape, not a path-dependent bookkeeping trick. Had the value depended on the route, "height" would be meaningless: you could gain energy walking in circles, and no consistent energy bookkeeping would survive. Verified: against Griffiths §2.3.1; the chain curl-free ⟺ path-independent ⟺ potential exists is the standard equivalence, and it is exactly where time-varying fields will later break the construction ( $\nabla\times\mathbf{E}=-\partial\mathbf{B}/\partial t\neq0$ ).

Go deeper · Derivation B · 1 minThe point charge potential from its field

Derivation BThe point charge potential from its field

The goal is equation (3): the potential landscape of a single point charge $Q$ , computed from the definition (1) with the reference at infinity. The strategy exploits the freedom Derivation A just purchased — since every path gives the same answer, choose the laziest one, a straight radial line coming in from infinity, so that the dot product collapses to an ordinary one-dimensional integral. The only care needed is with signs and limits; the check at the end is that differentiating must hand back Coulomb's field.

Step 1. Write the field: $\mathbf{E} = \dfrac{k_e Q}{r'^2}\,\hat{\mathbf{r}}$ for a point charge at the origin, with $r'$ the running distance. Why: Coulomb's law per unit test charge — the one input from the sources side.

Step 2. Choose the easiest path: integrate straight in along a radius from infinity to $r$ , so that $\mathbf{E}\cdot \mathrm{d}\mathbf{l} = E\,\mathrm{d}r'$ with no angular terms surviving. Why: path-independence licenses the laziest possible route; any spiral or detour must give the same answer, so take the one that makes the algebra trivial.

Step 3. Integrate:

V(r) = -\int_{\infty}^{r} \frac{k_e Q}{r'^2}\,\mathrm{d}r' = k_e Q\left[\frac{1}{r'}\right]_{\infty}^{r} = \frac{k_e Q}{r}

Why: the antiderivative of $1/r'^2$ is $-1/r'$ , and the boundary term at infinity vanishes — which is precisely what fixes $V(\infty) = 0$ and makes infinity a legal reference for a localized source.

Step 4. Check by differentiating: $-\dfrac{\partial}{\partial r}\left(\dfrac{k_e Q}{r}\right) = \dfrac{k_e Q}{r^2}$ — the Coulomb field returns, radially outward for positive $Q$ . Verified: matches Griffiths §2.3.1 and Jackson §1.5; the gradient check confirms equations (1) and (2) are inverses of each other on the simplest possible landscape.

When infinity fails — reference-point etiquette

The reference at infinity worked for the point charge because the boundary term died: the field falls fast enough that the work accumulated coming in from infinitely far away is finite. That is a property of localized sources, not of the definition — and the standard counterexample is worth meeting now, because it ambushes everyone once. Take an infinite line of charge, density $\lambda$ . Its field, $\mathbf{E} = \lambda/(2\pi\varepsilon_0 s)\,\hat{\mathbf{s}}$ in the cylindrical radius $s$ (a two-line Gauss's-law result — see Applications of Gauss's Law and Symmetry), falls only as $1/s$ , and integrating $1/s$ out to infinity diverges logarithmically. Infinity is simply not an admissible zero. The repair is to declare the zero at any finite radius $s_0$ :

V(s) = -\int_{s_0}^{s} \frac{\lambda}{2\pi\varepsilon_0 s'}\,\mathrm{d}s' = -\frac{\lambda}{2\pi\varepsilon_0}\ln\frac{s}{s_0} \qquad (7)

where: $s_0$ — the declared reference radius, entirely yours to choose. The logarithm diverges at both $s\to0$ and $s\to\infty$ , so no choice makes $V$ vanish at infinity — but none is needed: every measurable answer is a difference, $V(s_1)-V(s_2) = (\lambda/2\pi\varepsilon_0)\ln(s_2/s_1)$ , from which $s_0$ cancels entirely. This is why coaxial-cable formulas always carry $\ln(b/a)$ with both radii physical: the geometry supplies its own reference pair. The divergence is the idealization talking — real cables end — but within the model, a finite reference is mandatory. There is a deeper lesson hiding in this bookkeeping. The freedom to shift $V$ by a constant — to move the zero from $s_0$ to anywhere else — changes no field and no force: it is physics' first and mildest example of gauge freedom, the idea that the mathematical potentials carry a redundancy the physical fields never see. In electrodynamics proper the redundancy grows richer (the vector potential of E9 has a whole function's worth), and in the Standard Model it becomes the organizing principle of the forces themselves. Here it is one humble constant — but it is the same idea.

Go deeper · Derivation C · 1 minFrom the potential to Poisson equation

Derivation CFrom the potential to Poisson equation

The goal is to close the loop between the potential and the sources: equations (1)–(2) connect $V$ to the field, but Gauss's law connects the field to the charge density, and chaining the two must produce a differential equation that $V$ itself obeys. The strategy is pure substitution — write Gauss's law in differential form, express $\mathbf{E}$ through the potential, and recognize the operator that appears. The result compresses all of electrostatics into a single scalar equation, and it is the launch point for the next two articles in this chapter.

Step 1. Take Gauss's law in differential form: $\nabla\cdot\mathbf{E} = \rho/\varepsilon_0$ . Why: this is the local statement of how charge sources the field — the divergence half of the electrostatic field's specification, just as curl-free was the circulation half.

Step 2. Substitute the gradient: $\nabla\cdot(-\nabla V) = \rho/\varepsilon_0$ . Why: equation (2) expresses the same field through the potential; the substitution trades a vector unknown for a scalar one.

Step 3. Recognize the Laplacian:

\nabla^2 V = -\frac{\rho}{\varepsilon_0} \qquad (8)

Why: the divergence of a gradient is the Laplacian — and all of electrostatics is now one second-order scalar differential equation. Where $\rho = 0$ it becomes Laplace's equation $\nabla^2 V = 0$ , whose solutions are uniquely pinned down by boundary conditions. *Verified: standard route, Griffiths §2.3.3. Poisson's equation and Laplace's equation each get a full article next in this chapter; the boundary-value machinery they power is E2's entire subject.*

Limiting cases

Case	Result	Meaning
Far from any net charge $Q$	$V \to k_e Q/r$	every bounded distribution looks like a point charge from far away
Far from a neutral distribution	$V \sim k_e\,\mathbf{p}\cdot\hat{\mathbf{r}}/r^2$	monopole gone; the dipole term leads, falling one power faster
Inside a conductor in equilibrium	$V = \text{const}$ , $\mathbf{E} = 0$	a flat plateau — the whole conductor is one equipotential
Uniform field (capacitor gap)	$V = V_0 - Ex$	constant slope: evenly spaced equipotentials, $\Delta V = Ed$
Charge extending to infinity	$V(s) = -\dfrac{\lambda}{2\pi\varepsilon_0}\ln(s/s_0)$	infinity is an illegal reference; a finite anchor $s_0$ must be chosen
Time-varying fields	$\nabla\times\mathbf{E} = -\partial\mathbf{B}/\partial t \neq 0$	the construction breaks; $V$ alone no longer encodes $\mathbf{E}$ (E9 repairs it)

Common trap“Volts measure energy” — no: potential is not potential energy

A volt is a joule per coulomb. The potential $V$ is a property of a location — it exists whether or not anything sits there. The potential energy $U = qV$ belongs to a charge placed at that location, and it scales with the charge: doubling $q$ doubles $U$ but leaves $V$ untouched. The pairing has a treacherous twin: high potential does not mean strong field. Inside a charged hollow sphere the potential is high and perfectly uniform while the field is exactly zero — a high plateau is still flat. The field tracks the slope of $V$ , never its value; and conversely $V = 0$ at a point says nothing about $\mathbf{E}$ there (midway between $+q$ and $-q$ , the potential vanishes while the field is at full strength).

Worked examples

1 · Two charges, one scalar sum (confidence builder)

$Q_1 = +3.0\ \mu\mathrm{C}$ sits at the origin and $Q_2 = -2.0\ \mu\mathrm{C}$ at $x = 0.40\ \mathrm{m}$ . Find the potential at the midpoint $P$ , $x = 0.20\ \mathrm{m}$ .

Step 1. Distances: $r_1 = r_2 = 0.20\ \mathrm{m}$ . Both charges are equidistant from $P$ .

Step 2. Superpose as numbers, signs included — equation (4): $V = k_e\left(\dfrac{3.0\times10^{-6}}{0.20} + \dfrac{-2.0\times10^{-6}}{0.20}\right)$ .

Step 3. The bracket collapses to $\dfrac{1.0\times10^{-6}}{0.20} = 5.0\times10^{-6}$ .

Step 4. $V = (8.99\times10^9)(5.0\times10^{-6}) = \boxed{+4.5\times10^4\ \mathrm{V} = 45\ \mathrm{kV}}$ . Notice what you did not do: no components, no angles, no unit vectors. The negative charge simply subtracted its share. Computing the field at $P$ would have required vector care; the potential asked only for arithmetic.

2 · The electron gun (exam level)

An electron ( $q = -1.60\times10^{-19}\ \mathrm{C}$ , $m = 9.11\times10^{-31}\ \mathrm{kg}$ ) starts from rest and is accelerated through a potential difference of $500\ \mathrm{V}$ . Find its final kinetic energy in joules and electron-volts, and its final speed. Is a classical treatment safe?

Step 1. Energy bookkeeping via equation (6): the electron, being negative, spontaneously moves toward higher potential, and its potential energy falls by $|q|\,\Delta V$ .

Step 2. Kinetic energy gained: $KE = (1.60\times10^{-19})(500) = 8.0\times10^{-17}\ \mathrm{J}$ .

Step 3. In accelerator units: one electron through one volt defines one electron-volt, so $KE = \boxed{500\ \mathrm{eV}}$ — read straight off the voltage, no arithmetic at all.

Step 4. Speed: $v = \sqrt{2\,KE/m} = \sqrt{2(8.0\times10^{-17})/(9.11\times10^{-31})} = \sqrt{1.76\times10^{14}}$ .

Step 5. $v = \boxed{1.3\times10^7\ \mathrm{m/s}}$ — about 4% of the speed of light, so relativistic corrections are at the $10^{-3}$ level: classical mechanics is safe here, though a 50 kV electron microscope would need the relativistic formula.

Step 6. Cross-check with the field picture: whatever the geometry of the electrodes, however curved the path, the answer is the same — that is path-independence earning its keep. (For the plate geometry itself, look back at the uniform-field landscape in the physical picture: the electron simply rolls down that constant ramp.)

Origins · 1 minAlessandro Volta and how this was discovered

Alessandro Volta

1745–1827 · Como and Pavia, Lombardy

Volta turned a dispute into a device. When Luigi Galvani claimed that twitching frog legs revealed "animal electricity," Volta suspected the metals, not the frog — and proved it by stacking alternating discs of zinc and copper separated by brine-soaked cloth. His 1800 "pile" was the first battery: the first machine in history to maintain a steady potential difference and drive a continuous current. Within weeks of his letter to the Royal Society, chemists were using piles to decompose water. The rigorous mathematics of potential came later, from Poisson and Green; Volta's contribution was making potential difference something you could build, hold, and measure — which is why the unit bears his name.

History

1745

— The Leyden jar stores charge at high potential for the first time; electricity becomes bottled.

1800

— Volta's pile delivers a steady potential difference — the first battery, and the birth of current electricity.

1813

— Siméon Denis Poisson publishes

\nabla^2 V = -\rho/\varepsilon_0

, the differential equation the potential obeys.

1828

— George Green's essay names the "potential function" and founds potential theory, including the theorems behind boundary-value methods.

1881

— The volt is adopted as an international unit at the first International Electrical Congress in Paris.

Going further

Texts. Griffiths, Introduction to Electrodynamics, §2.3 · Purcell & Morin, Electricity and Magnetism, Ch. 2 · Jackson, Classical Electrodynamics, §1.5 · Zangwill, Modern Electrodynamics, §3.3.

Provenance. All equations standard and verifiable in Griffiths §2.3, Jackson §1.5, Purcell Ch. 2, Zangwill §3.3. A verifier should recompute: (1) the point-charge potential by integrating Coulomb's field from infinity, recovering $V=q/4\pi\varepsilon_0 r$ ; (2) $\mathbf{E}=-\nabla V$ reproduces the Coulomb field; (3) substituting into Gauss's law yields $\nabla^2 V=-\rho/\varepsilon_0$ ; (4) the infinite-line logarithmic potential and the parallel-plate $\Delta V=\sigma d/\varepsilon_0$ . Constants: $\varepsilon_0=8.854\times10^{-12}\,\mathrm{F/m}$ , $1\,\mathrm{eV}=1.602\times10^{-19}\,\mathrm{J}$ , $1\,\mathrm{V}=1\,\mathrm{J/C}$ — all CODATA/textbook standard. Worked numbers rechecked for this edition: example 1 gives $V = 8.99\times10^9 \times 5.0\times10^{-6} = +4.5\times10^4$ V; example 2 gives $KE = 8.0\times10^{-17}$ J $= 500$ eV and $v = \sqrt{1.76\times10^{14}} = 1.3\times10^7$ m/s.

Builds on: The Electric Field · Coulomb's Law · The Curl of the Electrostatic Field · Superposition Principle for Electric Fields Leads to: Poisson's Equation · Potential of a Localized Charge Distribution · Applications of Gauss's Law and Symmetry

What exactly is the difference between potential, potential energy, and voltage?

Three distinct objects, best separated by units. The potential $V$ is a property of a location — work per unit charge, in volts (J/C) — set up by the sources whether or not anything sits there. The potential energy $U = qV$ belongs to a specific charge $q$ placed at that location, in joules; flip the sign of $q$ and the energy landscape inverts, which is why electrons "fall upward" in $V$ . Voltage is engineering shorthand for a potential difference between two identified points — battery terminals, circuit nodes — and because it is always a difference, it dodges the reference-point question entirely. A "9-volt battery" promises nine joules per coulomb moved between its terminals and says nothing about any absolute zero.

Why is there a minus sign in the definition, and what breaks if I drop it?

The minus sign makes $V$ count the work done against the field, so that the field points from high potential toward low — downhill — and a positive charge released at rest loses potential energy as it gains kinetic, exactly like a ball on a hillside. Drop the sign and every force in your bookkeeping reverses: charges would "roll uphill," $U = qV$ would rise as particles accelerate, and energy conservation would look violated at each step. It is a convention, but a load-bearing one, chosen so electrostatic energy behaves like the mechanics you already know — and every voltmeter and textbook on Earth has voted for it.

Can the potential itself ever be measured, or only differences?

Only differences — and this is physics, not instrument limitation. The reference point in the definition is a free choice; moving it adds one constant to $V$ everywhere, and no experiment can detect a constant added to the potential of the whole universe, because every force and every energy exchange involves $\nabla V$ or $\Delta V$ . Even the "absolute" potentials quoted for power lines or Van de Graaff domes are tacitly differences from ground. This humble fact is your first meeting with gauge freedom — an unmeasurable redundancy in the potentials underlying a fully measurable theory — which grows, by E9 and beyond, into one of the organizing principles of modern physics.

Why can a scalar potential be defined for the electrostatic field but not for magnetic or time-varying fields?

Because the construction consumes exactly one input: $\nabla\times\mathbf{E} = 0$ . Zero curl means zero circulation around every loop (Stokes), which means path-independent line integrals, which is what makes " $V$ at a point" a single number. A magnetostatic field has $\nabla\times\mathbf{B} = \mu_0\mathbf{J} \neq 0$ wherever current flows, so no single-valued scalar potential can represent it globally. And once fields vary in time, Faraday's law gives $\nabla\times\mathbf{E} = -\partial\mathbf{B}/\partial t$ : the electric field itself stops being conservative, "voltage between two points" becomes path-dependent (a fact that regularly startles people probing fast circuits), and the repair requires the vector potential: $\mathbf{E} = -\nabla V - \partial\mathbf{A}/\partial t$ , the story of E9.

Where should I put the reference point, and when is infinity illegal?

Wherever it makes your formulas cleanest — the physics is indifferent. For localized distributions the polite default is infinity, which makes $V = \frac{1}{4\pi\varepsilon_0}\sum q_i/r_i$ come out reference-free. Infinity fails whenever the idealized source itself extends to infinity: the infinite line gives $V \propto \ln s$ (divergent at both ends of the range — pick a finite $s_0$ ), the infinite plane gives $V \propto -|z|$ (same disease). These are artifacts of idealization — real wires end — but within the model the finite reference is mandatory, and only differences appear in physical answers. In circuit practice the chosen reference is called ground.

If the potential is zero at a point, is the field zero there too — or vice versa?

Neither implication holds, and both directions are exam traps. Midway between $+q$ and $-q$ the potential is exactly zero while the field is at full strength, pointing from the positive charge toward the negative one: the landscape crosses sea level while sloping. Conversely, inside a hollow charged conductor the field is exactly zero while the potential is high and constant: a plateau far above sea level is still flat. The field is the slope of the potential, $\mathbf{E} = -\nabla V$ — so $\mathbf{E}$ probes how $V$ changes, never what $V$ is. Whenever a problem hands you one of the two and asks about the other, differentiate or integrate; never eyeball.

When should I compute V first and when should I go straight for E?

Default to $V$ first — that is the article's whole economy argument. Scalars superpose by signed arithmetic, so for a collection of charges or a continuous distribution, computing $V$ and then taking one gradient is almost always less work than assembling vector components from the start. Go straight for $\mathbf{E}$ in two situations: when the problem has enough symmetry that Gauss's law hands you the field in two lines (sphere, cylinder, plane — see Applications of Gauss's Law and Symmetry), or when you only need the field at a point of special symmetry where the direction is obvious. And if you already have $\mathbf{E}$ everywhere, remember the bridge runs both ways: $V$ follows by a line integral along the laziest path you can find.