The Capacitance

Field notes are where theory meets the bench. They are not textbooks. They are records of what happened when you stopped thinking about capacitance as an equation and started thinking about it as a physical reality — a thing you can hold, measure, and occasionally set on fire.

What Is a Capacitor, Really?

Textbooks will tell you that capacitance is C = Q/V. They'll show you two parallel plates separated by a dielectric and derive C = εA/d. All correct. All abstract. None of them tell you what happens when you charge a 4700µF capacitor to 50 volts and then short it with a piece of wire.

What happens is this: the wire glows white-hot. The capacitor hisses. The wire may melt. And all of that energy came from nothing more than an electric field trapped between two metal surfaces.

A capacitor is, at bottom, a charge separator. Put positive charge on one plate and negative on the other, and the attraction between them creates an electric field in the space between. That field stores energy. Not in the plates. Not in the wire. In the gap. In the dielectric. In the vacuum itself, if that's what fills the space.

The Geometry of Storage

The formula C = εA/d tells you everything and nothing. Yes, larger area means more capacitance — more surface for charge to gather on. Yes, smaller separation means more capacitance — the plates attract more strongly, and you can pack more charge before the voltage rises to a given level. Yes, a higher permittivity dielectric increases capacitance — the material polarizes, its molecules aligning with the field, which reduces the internal field and allows more charge to accumulate at the same voltage.

But the formula hides the complications. Edge effects at the boundary of plates. The fact that real capacitors have series inductance and resistance. The fact that at high frequencies, the capacitor stops behaving like a capacitor and starts behaving like an inductor. The parasitics are not corrections to the ideal model. They are the real device.

I once measured a 100nF ceramic capacitor and found that at 10 MHz, its impedance was rising — inductive behavior dominated. The datasheet showed a self-resonant frequency of 30 MHz, but my measurement showed 12 MHz. The difference? PCB trace inductance. The capacitor was fine. The circuit around it was not.

Dielectric Absorption

Here is a phenomenon that will not appear in your introductory textbook but will cost you hours of debugging if you encounter it: dielectric absorption. When you charge a capacitor, then discharge it, then measure its voltage — it slowly creeps back up. A discharged capacitor, sitting on your bench, will develop a measurable voltage over minutes or hours.

This happens because the dielectric material absorbs charge the way a sponge absorbs water. Polar molecules reorient over time, releasing stored charge even after the external circuit has been connected to ground. Tantalum capacitors are notorious for this. Polypropylene is better. Air is the best — it doesn't absorb anything because there's nothing to absorb.

This matters when you're building sample-and-hold circuits, or precision integrators, or anything that depends on a capacitor holding a charge without leaking. If you've ever wondered why your voltage holds for a while and then slowly drifts, dielectric absorption is often the culprit.

Energy in the Field

The energy stored in a capacitor is E = ½CV². Note the V². Double the voltage, and you quadruple the stored energy. This is why high-voltage capacitors are so much more dangerous than low-voltage ones of the same capacitance. A 100µF capacitor at 10 volts stores 5 millijoules — noticeable but harmless. At 400 volts, it stores 8 joules — enough to weld metal, start fires, or kill you.

The energy is in the field. The field strength is E = V/d. Break down the dielectric — push the field so hard that the material can no longer insulate — and the capacitor discharges violently. Air breaks down at about 3 kV/mm. Paper, about 16 kV/mm. Air is cheap and reliable but low-energy-density. Ceramics and plastics can hold much more but are more expensive and more fragile.

Practical Wisdom

A few things I've learned from years of working with capacitors:

1. Capacitors are not ideal. They have ESR (equivalent series resistance), ESL (equivalent series inductance), and leakage current. Your schematic symbol is a lie.
2. Decoupling capacitors work best when they're small and close to the IC they're decoupling. A 0.1µF capacitor at 1 cm of trace has more effective inductance than you'd expect.
3. Polarized capacitors (electrolytic, tantalum) will fail catastrophically if reverse-biased. Not "fail safe" — catastrophically. The capacitor remembers its polarity.
4. Capacitors age. Electrolytic capacitors dry out. Their capacitance drops and their ESR rises. A power supply that works on day one may fail after three years because a $0.05 capacitor died.
5. Unused capacitors are not harmless. They hold charge. Always discharge before handling. A capacitor on a high-voltage board can hold a lethal charge for hours.

The Quiet Power

Capacitance is quiet. There's no moving part, no audible click, no indicator light. You charge a capacitor and nothing happens. You discharge it and nothing happens — except that something else happens, somewhere else, in the circuit that receives the energy.

But the energy was there all along, sitting in the electric field between two metal surfaces, waiting for the moment when you close the circuit and let the universe remember that charge wants to return to equilibrium.