Electronics Guide

The Principle of Superposition

In vacuum and in linear materials, the field of several charges is the vector sum of the fields each would produce alone. Distributed charge is described by a volume density ρ_v (C/m³), a surface density ρ_S (C/m²), or a line density ρ_L (C/m). Each element, such as ρ_v dv, contributes a point-charge field, and the total is the integral

E = ∫ ρ_v dv / (4πε₀R²)

where R is the distance from each element to the field point, and each contribution points along that line. The integral is simple in principle and laborious in practice, which is why Gauss's law and numerical methods matter.

The Electric Dipole

Equal and opposite charges +Q and −Q a small distance d apart form an electric dipole with moment p = Qd, directed from −Q to +Q. At distances r much larger than d the two fields nearly cancel, and on the dipole's axis the field is

E = 2p / (4πε₀r³)

This field points along p. On the perpendicular bisector the field has half this magnitude and points opposite to p. Either way, it falls as the inverse cube of distance rather than the inverse square. Dipoles also describe the polarization of insulating materials, treated below.

Electric Flux and the Integral Form

The electric flux through a surface is Φ_E = ∫_S D · dS, where dS is a small patch of surface oriented along its outward normal. Physics texts often integrate E instead, which in vacuum divides this flux by ε₀. Gauss's law states that the net flux out of any closed surface S equals the free charge enclosed:

∫_S D · dS = Q_enc

The law follows from the inverse-square law. A point charge's flux density weakens as 1/r² while the area of a surrounding sphere grows as r², so the total flux through a concentric sphere, D × 4πr² = Q, does not depend on its radius. The flux through any small patch depends only on the solid angle the patch subtends at the charge, so every closed surface around the charge passes the same total. Superposition extends the result to any distribution of charge. Gauss's law always holds, but it yields the field directly only when symmetry provides a surface over which D is either normal to the surface with constant magnitude or tangent to it, contributing no flux.

Three Useful Symmetries

Point symmetry. For spherically symmetric charge, a concentric sphere of radius r gives D × 4πr² = Q_enc. Outside the charge, the field is that of a point charge at the center. Inside a thin charged shell it is zero, and inside a uniformly charged sphere of radius R it rises linearly from the center, E = Qr/(4πε₀R³).

Line symmetry. For a long line with density ρ_L, a coaxial cylinder of radius r and length L gives D × 2πrL = ρ_LL, so E = ρ_L/(2πε₀r). Between the conductors of a coaxial cable this field, divided by the dielectric's ε_r, applies. Outside a shield that carries the equal and opposite charge, the enclosed charge is zero, and so is the field.

Plane symmetry. For a large sheet with density ρ_S, a short cylinder, or pillbox, piercing the sheet gives 2DA = ρ_SA, so E = ρ_S/(2ε₀) on each side, independent of distance. Sheets carrying +ρ_S and −ρ_S produce fields that add between them and cancel outside, leaving a uniform field ρ_S/ε₀ in the gap: the ideal parallel-plate capacitor.

The Differential Form

Applied to a vanishingly small volume, Gauss's law becomes

div D = ρ_v

where div D = ∂D_x/∂x + ∂D_y/∂y + ∂D_z/∂z in Cartesian coordinates. Flux originates only where free charge exists. The divergence theorem links the two forms, and both belong to Maxwell's equations, so Gauss's law also holds for fields that vary in time.

Field Concentration at Edges and Points

Charge crowds onto regions of small radius of curvature. Two conducting spheres of radii r₁ and r₂, far apart but joined by a thin wire, show why. At equal potential, Q₁/(4πε₀r₁) = Q₂/(4πε₀r₂), so charge is proportional to radius, and the surface field, Q/(4πε₀r²), is inversely proportional to it:

E₁ / E₂ = r₂ / r₁

An isolated sphere at potential V has a surface field of V/r, so a 1 mm radius at 3 kV reaches 3 MV/m, about the uniform-field breakdown strength of air. A surface this sharply curved tolerates a higher surface field before corona starts, because an electron avalanche needs distance to grow, but the trend holds. Points, burrs, solder spikes, and the corners of copper traces concentrate fields the same way, which is why discharges begin at edges and why high-voltage hardware uses large radii.

In a coaxial geometry with voltage V between an inner conductor of radius a and a shield of inner radius b, the peak field is at the inner surface:

E_max = V / (a ln(b/a))

For a fixed shield, the peak field is lowest when ln(b/a) = 1, that is, when b/a equals Euler's number, about 2.718. With V = 10 kV and b = 10 mm, inner radii of 1 mm, 3.7 mm, and 5 mm give peak fields of 4.3, 2.7, and 2.9 kV/mm.

Electrostatic Shielding

A closed conducting shell with no charge inside keeps its interior free of field, whatever charges lie outside, because external field lines end on induced surface charge. This is the Faraday cage. A charge inside a shell, however, induces an equal charge on its outer surface, which produces an external field unless the shell is grounded.

Circuits apply the principle at small scale. Grounded copper pours shield high-impedance nodes, a Faraday shield between transformer windings diverts capacitively coupled noise to ground, and a driven guard held at a sensitive input's potential leaves no voltage across the leakage paths it intercepts. A static field penetrates an opening only to a depth comparable to the opening's size, so a mesh shields well when its holes are small. An electrostatic shield does not, by itself, stop low-frequency magnetic fields.

Bound Charge and Permittivity

Polarization leaves uncanceled bound charge at a material's surfaces, with density equal to the outward normal component of P, and inside it wherever P varies, with density −div P. This bound charge opposes the applied field inside the material. The flux density includes its effect,

D = ε₀E + P

so that div D = ρ_v counts only free charge, the charge an engineer controls. In a linear, isotropic material, P = ε₀χ_eE, where χ_e is the electric susceptibility, and

D = ε₀(1 + χ_e)E = ε₀ε_rE = εE

The relative permittivity ε_r = 1 + χ_e, or dielectric constant, is the factor by which a material filling the space between a capacitor's electrodes raises the charge stored at a given voltage.

The values come from the table in Sears, Zemansky, and Young's University Physics, as reproduced by Georgia State University's HyperPhysics. Permittivity also varies with frequency and temperature, so circuit-material datasheets state the measurement conditions. Rutile's two values show that permittivity can depend on direction; in such anisotropic materials, D and E are generally not parallel.

Frequency, Loss, and Nonlinearity

Each polarization mechanism needs time to respond. Dipole rotation, the slowest, stops following the field in many polar materials somewhere in the radio or microwave range; ionic displacement follows up to infrared frequencies, and electronic displacement through visible light into the ultraviolet. As each mechanism drops out, ε_r falls, and energy is dissipated near each transition.

Dielectric loss is described by a complex permittivity, ε = ε′ − jε″, and a loss tangent, tan δ = ε″/ε′. In sinusoidal steady state the average power dissipated per unit volume is ωε″E_rms². For a capacitor whose only loss is in its dielectric, the dissipation factor equals tan δ, and the equivalent series resistance is tan δ / (ωC).

Linearity is an approximation. Class 2 ceramic capacitor dielectrics based on barium titanate are ferroelectric, with permittivities in the thousands that fall as DC bias rises and vary strongly with temperature, so capacitance at rated voltage can be a fraction of the nominal value; manufacturers publish DC-bias curves for this reason. Class 1 ceramics and polymer films trade lower permittivity for stability.

Voids and Layers in Insulation

When insulating layers are stacked across a field, D is the same in each, so each layer's field is inversely proportional to its permittivity, and a thin air gap carries ε_r times the field in the surrounding solid. Take 10 kV across 2 mm of insulation with ε_r = 4, a uniform 5 kV/mm. A 0.1 mm air-filled delamination lying across the field divides the voltage as 10 kV = E_solid(1.9 mm) + 4E_solid(0.1 mm), leaving 4.3 kV/mm in the solid and 17.4 kV/mm, about 1.7 kV, across the gap.

Whether the gap discharges depends on its thickness and pressure through the Paschen curve, described below, but overstressed voids are the classic site of partial discharge. A spherical void is milder: its internal field is 3ε_r/(2ε_r + 1) times the distant field, or 1.33 times for ε_r = 4. This is why potting processes remove air and why high-voltage insulation is tested for partial discharge.

Plates, Cylinders, and Spheres

Plates of area A a distance d apart have E = Q/(εA) and V = Qd/(εA), so

C = ε₀ε_rA / d

provided the plates are much wider than their spacing. Power and ground planes 10 cm square and 100 µm apart in a laminate with ε_r = 4 have 3.5 nF. A multilayer ceramic capacitor with N interleaved electrodes stacks N − 1 such layers in parallel, so C = ε₀ε_r(N − 1)A/d, which is why thin layers and high permittivity drive capacitor miniaturization.

For a coaxial line with inner radius a and shield radius b, V = (ρ_L/(2πε)) ln(b/a), so the capacitance per unit length is

C′ = 2πε₀ε_r / ln(b/a)

With a = 0.45 mm, b = 1.5 mm, and solid polyethylene (ε_r = 2.25), C′ is about 104 pF/m. Only the ratio b/a matters. Concentric spheres of radii a and b have C = 4πεab/(b − a), and letting b grow without limit gives an isolated sphere, C = 4πε₀a, about 11 pF for a 10 cm radius. The formulas for wires follow from the method of images, described later.

A wire 1 mm in diameter with its axis 10 mm above a metal chassis in air has about 15.1 pF/m. Doubling the height lowers that only to about 12.7 pF/m, because capacitance depends on the logarithm of a dimension ratio.

Fringing Fields

Real plates have edges, and the field bulging beyond them adds capacitance. The Palmer formula, based on H. B. Palmer's 1937 Schwarz–Christoffel conformal mapping of the edge field, approximates thin rectangular plates of width W and length L separated by a gap G in air with one correction factor for each pair of edges:

C = (ε₀WL/G) × [1 + G/(πW) + (G/(πW)) ln(2πW/G)] × [1 + G/(πL) + (G/(πL)) ln(2πL/G)]

For square plates whose side is 10 times the gap, the correction raises capacitance about 35 percent above εA/d; at 100 times the gap, about 5 percent. In a 2006 study of MEMS actuators at École Polytechnique de Montréal, Mehran Hosseini, Guchuan Zhu, and Yves-Alain Peter compared the formula with finite-element simulation of 300 by 600 µm electrodes 305 µm apart. At that gap, εA/d gave only about a third of the simulated capacitance, while the Palmer formula came within about 6 percent, and closer as the gap narrowed. Fringing matters most for small electrodes: sensors, touch pads, MEMS structures, and narrow circuit-board traces. Precision fixtures avoid it with a guard ring held at the measuring electrode's potential, which moves the fringe field onto the guard.

Systems of Many Conductors

Many coupled conductors are described by a capacitance matrix, Q_i = ∑_j C_ijV_j, whose diagonal entries are positive and whose off-diagonal entries are negative or zero. The mutual capacitance a circuit simulator places between nodes i and j is −C_ij, and each conductor's capacitance to the reference is the sum of its row. Field solvers compute this matrix for crosstalk analysis and parasitic extraction.

Electrostatic Forces

Differentiating stored energy with respect to position gives force. Charged parallel plates attract with

F = Q²/(2εA) = ½εE²A

a pressure of ½εE². At 3 MV/m in air that pressure is about 40 Pa, tiny beside the atmosphere's 101 kPa. At micrometer scale the balance changes. Paschen's law, described below, predicts that a 7.5 µm air gap holds about 327 V, a field near 44 MV/m that exerts about 8 kPa, so electrostatic actuation is widely used in MEMS devices such as comb drives and micromirror arrays.

Such actuators have a characteristic instability. A plate on a spring of stiffness k, at initial gap g₀, moves stably only until it has traveled one-third of the gap; beyond that, the electrostatic force outgrows the spring force and the plate snaps shut. For a rigid, parallel plate with fringing ignored, this pull-in occurs at

V_PI = √[8kg₀³ / (27εA)]

Analytical Methods

• Direct integration handles one-dimensional cases. Between parallel plates d²V/dx² = 0, so V varies linearly across the gap.
• Separation of variables expands solutions in series suited to rectangular, cylindrical, or spherical boundaries.
• The method of images replaces a grounded plane with a mirror charge. Above the plane, a charge Q at height h produces the field of Q plus an image −Q at depth h, and it is pulled toward the plane with force Q²/(16πε₀h²). The wire-over-plane capacitance follows the same way.
• Conformal mapping transforms two-dimensional geometries into simpler ones; Palmer's fringing formula came from it.

Poisson's Equation in a Semiconductor Junction

In a one-sided abrupt junction, such as a heavily doped p region against a lightly doped n region, nearly all of the depletion region lies on the lightly doped side. There, ionized donors of density N_D give a charge density ρ_v ≈ qN_D, where q is the elementary charge. Integrating Poisson's equation once gives a field that rises linearly to qN_DW/ε_s at the junction, where W is the depletion width and ε_s the semiconductor's permittivity. Integrating again gives a parabolic potential with a drop of qN_DW²/(2ε_s). Equating that drop to the built-in potential plus any reverse bias gives W, and the junction capacitance per unit area, ε_s/W, is a parallel-plate result whose gap changes with voltage.

Finite Differences

On a square grid in a charge-free region of uniform permittivity, the two-dimensional Laplace equation becomes a simple rule: each node's potential is the average of its four neighbors.

V_i,j = (V_i+1,j + V_i−1,j + V_i,j+1 + V_i,j−1) / 4

In three dimensions six neighbors are averaged, and for Poisson's equation a term h²ρ_v/ε, with h the grid spacing, is added to the sum before dividing. Relaxation solves the resulting equations:

1. Fix the potential of every node on a conductor or outer boundary, and give the other nodes an initial guess.
2. Sweep through the free nodes, replacing each value with the average of its neighbors. Using updated values at once, as the Gauss–Seidel method does, and overcorrecting slightly, as successive over-relaxation does, both speed convergence.
3. Repeat until the largest change in a sweep falls below a chosen tolerance.
4. Find the field from differences between neighboring nodes, each conductor's charge from Gauss's law around it, and capacitance from C = Q/V.

A square trough with its top side at 100 V and the other sides at 0 V makes a good test. Its four rotations together put 100 V on every side and so make the whole interior 100 V; by superposition, the center of the original problem sits at exactly 25 V. A grid of 5 by 5 nodes converges to the values below and reproduces that 25 V, as a correct solver must.

The other nodes carry discretization error. The exact solution of the continuous problem, a Fourier series, gives 54.05 V beside the middle of the 100 V side, against 52.68 V on this grid, and the difference shrinks as the grid is refined.

Finite Elements and Boundary Elements

The finite element method divides the region into triangles or tetrahedra, approximates the potential simply within each, and picks the nodal values that minimize the total field energy for the given conductor potentials. Elements follow curved boundaries and material interfaces, and they shrink where fields change rapidly, as at conductor edges. FEMM, a finite element program by David Meeker, solves linear electrostatic problems this way on two-dimensional planar and axisymmetric domains.

Boundary element methods, usually called the method of moments in electromagnetics, mesh only surfaces: the conductors and any interfaces between different dielectrics. Each panel carries an unknown charge, and a dense linear system finds the charges that give every conductor its known potential. Because open space needs no mesh, these methods suit capacitance extraction for interconnect and packages.

Checking a Solver

• Reproduce an analytical case, such as a coaxial line, before trusting a model.
• Refine the mesh until capacitance stops changing within the needed tolerance.
• Move open boundaries outward and confirm that the answer holds.
• Expect the field at a perfectly sharp corner never to converge, because the ideal value there is infinite. Capacitance converges regardless; for peak fields, model the real edge radius.

Breakdown in Gases

A gas always holds a few free electrons, released by cosmic rays and natural radioactivity. A strong field accelerates them enough to ionize molecules on collision, and the freed electrons ionize further in an avalanche, which becomes self-sustaining with the help of electrons that ion impacts release from the cathode. Friedrich Paschen found empirically, and Townsend's avalanche theory explains, that the breakdown voltage of a uniform gap depends on the product of pressure p and gap d:

V_b = Bpd / [ln(Apd) − ln(ln(1 + 1/γ))]

Here A and B are empirical constants of the gas and γ is the secondary-emission coefficient of the cathode. The curve has a minimum: at large pd, electrons collide too often to gain ionizing energy; at small pd, too rarely to multiply. For air between flat electrodes, the minimum is commonly given as about 327 V at a pd of 5.7 Torr·mm, which at atmospheric pressure is a gap of about 7.5 µm. Gaps near 10 µm and smaller can depart from the curve once field emission from the cathode becomes important. Ordinary gaps lie on the high-pd side of the minimum, so reduced pressure is hazardous: at altitude or in a partial vacuum, pd falls toward the minimum, and a gap that was safe at sea level can arc.

The course notes for MIT's introductory electromagnetism course, 8.02, give the breakdown field of dry air as 3 × 10⁶ V/m, or 3 kV/mm, a representative value for a uniform field at room temperature and atmospheric pressure across gaps of roughly a centimeter. Because the breakdown field rises as a gap narrows toward the Paschen minimum, a uniform gap a fraction of a millimeter wide withstands a higher field; contamination, nearby insulating surfaces, and nonuniform fields all lower practical values. Near a sharp electrode the local field reaches breakdown first and produces corona, a partial discharge confined to the high-field region that erodes insulation, generates ozone and radio noise, and can grow into complete breakdown.

Breakdown in Solids and Liquids

Solid insulation fails by electronic breakdown, when the field frees and accelerates charge carriers; by thermal breakdown, when dielectric loss heats the material into a runaway; and by electromechanical breakdown, when electrostatic pressure thins a soft insulator. Over time, partial discharges in voids grow branching channels called electrical trees, and thin gate oxides wear out through time-dependent dielectric breakdown. Standardized short-time tests such as ASTM D149 and IEC 60243-1 measure dielectric strength, but the results depend on thickness, electrodes, voltage ramp rate, and temperature, and thin films generally show higher strength per unit thickness than thick sections. In liquids such as transformer oil, moisture and particles lower the strength.

Clearance and Creepage

IEC 60664-1, the basic safety publication on insulation coordination for equipment within low-voltage supply systems, turns this physics into minimum spacings. Clearance, the shortest distance through air between conductive parts, must withstand transient overvoltages. Creepage, the shortest distance along the surface of solid insulation, must resist tracking, the gradual growth of conductive paths through surface contamination, at the working voltage.

The figures that follow come from the second edition, IEC 60664-1:2007, as reproduced by Wei Zhang and Thomas LaBella in a Texas Instruments Power Supply Design Seminar paper on clearance and creepage. Clearance follows from the rated impulse voltage, tested with a 1.2/50 µs waveform. For overvoltage category II, equipment energized from a building's fixed installation, Table F.1 assigns 2,500 V on supplies up to 300 V line-to-neutral, such as 230 V, and 1,500 V on supplies up to 150 V, such as 120 V. Table F.2 converts that impulse voltage to a clearance.

The 1.5 mm required at 2.5 kV is an average field of about 1.7 kV/mm, well below air's uniform-field strength, because the inhomogeneous-field case allows for field concentration around real conductors. Pollution degree 2, the usual case for household and office equipment, needs 0.2 mm at 0.5 kV where pollution degree 1 needs 0.04 mm, since deposits and condensation can bridge small gaps. Reinforced insulation uses the next higher impulse row. Above 2,000 m, clearances are multiplied by the correction factors of Table A.2: 1.14 at 3,000 m, 1.48 at 5,000 m, and 3.02 at 10,000 m.

Creepage comes from Table F.4, by working voltage, pollution degree, and material group, which ranks insulators by comparative tracking index. At 400 V rms and pollution degree 2, it gives 2.0 mm for material group I, 2.8 mm for group II, and 4.0 mm for group III, and reinforced insulation needs twice the creepage of basic insulation.

The third edition, IEC 60664-1:2020, keeps clearances in Table F.2 but moves creepage to Table F.5, adds 1,500 V DC to the tables in Annexes B and F, and updates the altitude correction in a new Table F.10. Take design figures from the current edition and from the product standard that governs the equipment, such as IEC 62368-1 for audio, video, information, and communication technology equipment, which builds on this framework and sets its own values.

Capacitor Dielectrics

Capacitance per unit volume rises with permittivity and falls with dielectric thickness, while the voltage rating demands enough thickness to keep the working field well below the dielectric strength. Since energy density is ½εE², the best energy-storage dielectrics combine high permittivity with high strength. Film capacitors use polymers such as polyester (ε_r about 3.1) and polypropylene, the latter with very low loss. Ceramic capacitors use stable class 1 dielectrics or class 2 barium titanate formulations with permittivities in the thousands. Electrolytic capacitors use a very thin anodic oxide of aluminum or tantalum, grown to a thickness set by the rated voltage, for very large capacitance per volume.

Electrostatic Discharge

Contact and separation transfer charge between materials. Because a charged body's voltage is Q/C, reducing its capacitance at constant charge, as when a person lifts a foot from the floor, raises its voltage, while humidity helps by letting charge leak away through surface moisture. The ESD Association gives typical voltages of 35,000 V for walking across carpet at 10 to 25 percent relative humidity, against 1,500 V at 65 to 90 percent, and 12,000 V against 250 V for walking across vinyl tile.

The human body model of ANSI/ESDA/JEDEC JS-001 treats a person as 100 pF discharging through 1,500 Ω. At 4 kV that stores 400 nC and only 0.8 mJ, yet the ESD Association describes the current as rising in 2 to 10 ns to 0.67 A per kilovolt, about 2.7 A at 4 kV, over roughly 200 ns, enough to damage a susceptible device. A device can also acquire charge by induction and then discharge through its own pins, the event tested by the charged device model of ANSI/ESDA/JEDEC JS-002, which the ESD Association notes often lasts less than a nanosecond with peaks of several tens of amperes. System-level tests to IEC 61000-4-2 use 150 pF and 330 Ω, which produce a much higher peak current than the human body model.

Capacitive Touch Sensing

In self-capacitance sensing, a controller measures an electrode's capacitance to ground, and a finger, coupled to its surroundings through the body, adds capacitance. In mutual-capacitance sensing, the controller drives one electrode and measures the charge coupled into a neighbor; a finger diverts some of the field lines, so the mutual capacitance falls. By the parallel-plate formula, a finger fully covering a 1 cm² pad through 1 mm of acrylic (ε_r = 3.40) forms at most about 3.0 pF with it, and less in practice. In this simple model, doubling the cover thickness halves the signal, and a higher-permittivity cover such as glass raises it. Water, with ε_r near 80, couples strongly, which is why droplets can mimic touches. The article on Touch Sensing Systems covers controllers and panel construction.

High-Voltage Layout

• Keep radii large. Round copper corners, avoid pointed pads, trim leads, and remove solder spikes.
• Meet clearance and creepage from the governing product standard, with any altitude correction. The smallest distance anywhere along a barrier governs.
• Lengthen surface paths with slots milled between high-voltage and low-voltage regions, and use a qualified conformal coating or potting process to permit a lower pollution degree.
• Eliminate voids in potting, under coatings, and between laminated layers, where the field is elevated and partial discharge can start.
• Watch triple junctions, where conductor, solid insulator, and air meet; tiny air gaps there concentrate the field and often start surface flashover.
• Guard high-impedance inputs with a ring held at the input's potential to intercept surface leakage from high-voltage conductors.