Dynamo self-excitation

A locust in isolation triggers nothing. Add enough others to the same patch of ground and a threshold trips: serotonin floods the system, behavior shifts from solitary to gregarious, and a self-amplifying positive feedback loop begins. What follows is a swarm of millions built from a tiny seed — residual excitation bootstrapping itself to continental scale. Inside every piece of iron is the same principle waiting to run.

Inside every piece of iron is a ghost. When iron is magnetized and then demagnetized, a small fraction of the magnetic alignment survives — residual magnetism too weak to do useful work, a ghost of previous fields. For most of electrical engineering's early history, this ghost was a nuisance. For dynamo self-excitation, it was the seed.

The early dynamo was a prisoner of its own design. To generate electricity, the machine required a magnetic field. To create a magnetic field strong enough to be useful, the field coils required electric current. That current had to come from somewhere outside the machine — typically a battery or a smaller companion magneto. Early dynamos were never truly autonomous. They depended on an external source to start and often to sustain their operation.

Ányos Jedlik, a Benedictine monk and professor of physics at the University of Pest, understood the paradox before anyone published a solution. In his laboratory notebooks of the mid-1850s — later confirmed to date from around 1856 — Jedlik described a generator whose field magnets were powered by the machine's own output. The principle: if the iron cores retained even a trace of residual magnetism from previous use, that faint field would induce a tiny current as the rotor turned. Feed that current back through the field coils and the field strengthens marginally. The stronger field induces more current. More current strengthens the field further. The system climbs its own positive feedback loop until it reaches equilibrium at full power output. The machine bootstraps itself from ghost magnetism to full generation.

Jedlik never patented it. He apparently believed he had not been first. He had.

A decade later, the same principle was independently discovered by at least four researchers within eighteen months of each other. Henry Wilde identified residual magnetism as a self-starting seed in Manchester in 1866. Samuel Alfred Varley filed a patent on December 24, 1866. Werner von Siemens — who grasped the commercial significance more clearly than the others — presented the principle to the Berlin Academy of Sciences in January 1867, calling his paper "On the Conversion of Mechanical Energy into Electric Current Without the Use of Permanent Magnets." Charles Wheatstone announced the same principle to the Royal Society of London the same month.

The simultaneous convergence was not coincidence. All the components — electromagnetic theory, iron-core construction, efficient rotor design — had matured to the point where self-excitation was the only gap remaining. The adjacent possible had exactly one door, and four people walked through it in parallel.

The commercial consequences were immediate. Siemens used self-excited dynamos to power electric arc furnaces in the early 1870s — the first large-scale industrial application of electrical power. The principle is embedded in every subsequent generator design: Edison's Pearl Street Station generators, Westinghouse alternators, modern power plant turbines. Self-excitation removed the last external dependency from electrical power generation.

The biological mechanism is exact. A locust in isolation shows no swarming behavior. Add enough locusts to a patch of ground — the density threshold triggers serotonin release, which shifts individuals toward gregarious behavior. Gregarious individuals aggregate. Aggregation increases density further. More serotonin. More gregariousness. The system escalates from a handful of solitary insects to a swarm of millions through a pure positive feedback loop seeded by initial concentration. The locust swarm is a phase transition from solitary to gregarious behavior; the self-excited dynamo is a phase transition from dormant to powered state. Both transform a tiny perturbation into a self-sustaining state through positive feedback. Path dependence then locks the result: every generator design after Siemens inherited the self-excitation bootstrap, because no external-dependency alternative could compete with a machine that needed nothing to start.

The business insight travels equally well. Network effects, viral growth, and brand momentum are all self-excitation phenomena: the initial signal (residual magnetism, seed users, early reputation) triggers a feedback loop that amplifies to equilibrium. The critical question in each case is not whether the loop will run — positive feedback loops always run — but whether the seed is large enough to survive the initial noise before the loop takes hold.