Transformers: The Grid's Voltage Gearbox

The war that voltage won

In the 1880s electricity was a neighbourhood luxury. Edison's DC dynamos could only push power a kilometre or so before the voltage sagged to nothing, so every few city blocks needed its own coal-fired plant. The villain was simple physics: pushing power through a wire wastes energy as heat at a rate of I²R — proportional to the square of the current. Double the current and you quadruple the loss. To send real power a real distance, you had to get the current *down*.

But you cannot just lower the current — your toaster still needs its watts. Power is P = V·I, so the only way to carry the same watts at lower current is to raise the voltage in exact proportion. Send power at 100× the voltage and you carry it at 1/100 the current, which cuts the I²R loss by a factor of 10,000. The whole game of the grid is: transmit at terrifying voltage, then tame it back down at the doorstep. The device that does both — cheaply, silently, with no moving parts — is the transformer. It is why George Westinghouse and Nikola Tesla's AC system beat Edison's DC, and why your home runs on alternating current today.

Two coils, one iron core, no contact

Open a transformer and you find something almost insultingly simple: a closed loop of laminated iron with two coils of insulated copper wound around it. The primary coil takes the incoming AC; the secondary delivers the outgoing AC. The two coils never touch — they are electrically isolated. They talk to each other entirely through the magnetic field shared by the iron core. That iron is not decoration: it channels the magnetic flux so that almost every field line made by the primary threads through the secondary.

The physics is Faraday's law of induction. The AC in the primary makes the magnetic flux in the core swing up and down. A changing flux through a coil induces a voltage in that coil — and crucially, it induces the *same volts-per-turn* in every coil sharing the core. So a coil with twice as many turns sees twice the voltage. That is the entire secret: voltage divides up among the turns, and the ratio of turns sets the ratio of voltages.

          PRIMARY              SECONDARY
         (Np turns)            (Ns turns)

  AC  o──┐  ╔═══════════════════╗  ┌──o  AC
  in     │  ║  ███ iron core ███ ║  │     out
         ▓  ║  ███             ███║  ▓
  Vp     ▓  ║                     ║  ▓   Vs
  Ip     ▓  ║   <── magnetic ──>  ║  ▓   Is
         ▓  ║       flux  Φ       ║  ▓
  AC  o──┘  ╚═══════════════════╝  └──o
            no electrical contact
            — coupled only by Φ

   Vp/Vs = Np/Ns        (volts follow turns)
   Ip/Is = Ns/Np        (amps go the other way)
   Vp·Ip = Vs·Is        (power in = power out)

The ideal transformer, worked by hand

Strip away the imperfections and a transformer obeys three tidy relations. Let Np and Ns be the turns on the primary and secondary. The turns ratio n = Np/Ns governs everything. Voltage follows the turns directly; current follows them inversely; and because an ideal transformer stores no energy, the power flowing in must equal the power flowing out.

1. Step up or down? Compare the turns. More turns on the secondary (Ns > Np) means higher output voltage → a step-up. Fewer means a step-down. The grid steps up at the power plant and steps down near you.
2. Find the output voltage: Vs = Vp × (Ns/Np). The voltage scales by the turns ratio.
3. Find the output current from power conservation: Is = Ip × (Np/Ns). As voltage goes up, current comes down by the same factor.
4. Sanity check: multiply Vs × Is. It should equal Vp × Ip. If watts in ≠ watts out, you have made an arithmetic slip (or modelled a real, lossy transformer).

WORKED EXAMPLE — a generator step-up transformer
=================================================
Given:  Generator output  Vp = 20 kV  (typical alternator)
        Primary turns     Np = 100
        Secondary turns   Ns = 1900
        Real power drawn   P = 38 MW

Turns ratio  n = Ns/Np = 1900/100 = 19    (STEP-UP, ×19)

Secondary voltage:
   Vs = Vp × (Ns/Np) = 20 kV × 19 = 380 kV

Primary current (from P = Vp·Ip):
   Ip = P / Vp = 38e6 W / 20e3 V = 1900 A

Secondary current (power conserved):
   Is = Ip × (Np/Ns) = 1900 A / 19 = 100 A

Check the power:
   Vp·Ip = 20 kV × 1900 A = 38 MW   ✓
   Vs·Is = 380 kV × 100 A = 38 MW   ✓

The payoff — line loss in a 50 Ω round-trip line:
   At 20 kV / 1900 A:  I²R = 1900² × 50 = 180 MW  (!! more than the power itself)
   At 380 kV / 100 A:  I²R = 100²  × 50 = 0.5 MW  (1.3% of the load)

Real transformers leak, heat, and hum

The ideal transformer is a beautiful lie. A real one wastes a little of the power passing through, and the losses fall into two camps that behave very differently — a distinction every power engineer carries in their head.

1. Copper losses (I²R, load losses): the windings have real resistance, so current heats them. These scale with the *square of the load current* — heavy at full load, near zero at no load. They are why big transformers are oil-cooled with radiator fins.
2. Core losses (iron losses): the iron is magnetised back and forth 50 or 60 times a second. Hysteresis (re-flipping the magnetic domains) and eddy currents (loops of current induced in the iron itself) both dissipate heat. These depend on *voltage and frequency, not load* — they run 24/7 the moment the transformer is energised, even with nothing plugged in.
3. Leakage flux: not every field line stays in the core; a little escapes and couples nothing. This shows up as a series *leakage reactance* that makes the output voltage droop under load — useful, actually, for limiting fault current.

The engineering moves follow straight from the loss list. To kill eddy currents, the core is built from thin laminations — sheets of silicon steel varnished apart so no big current loop can form. To cut copper loss, you use fatter conductor and accept a heavier, costlier machine. The trade-off between copper and iron is literally a cost-optimisation problem solved for every transformer's expected duty cycle. That faint 100/120 Hz hum you hear from a substation? That is the core physically flexing twice per AC cycle — a phenomenon called magnetostriction.

The voltage staircase: plant to plug

A transformer rarely works alone. Electricity descends from the power plant to your wall in a staircase of voltage levels, each step a transformer in a substation. The grid is essentially a chain of these gearboxes, shifting up for distance and down for safety. Modern grids do this with three-phase power, so each 'transformer' is really three windings (or a three-limb core) handling all three phases at once.

  THE GRID VOLTAGE STAIRCASE  (typical values)
  =============================================

  ⚡ Power plant generator ............  ~20 kV
        │  STEP-UP transformer  ▲
        ▼
  ═══ Transmission lines ═════════  138–765 kV   ← long haul, low current
        │  STEP-DOWN (bulk substation) ▼
        ▼
  ─── Sub-transmission ───────────   34.5–138 kV
        │  STEP-DOWN (distribution substation) ▼
        ▼
  ─── Primary distribution ───────    4–35 kV     ← the lines down your street
        │  STEP-DOWN (pole / pad transformer) ▼
        ▼
  🏠 Service to home ..............  120/240 V    ← what your socket sees

  Each ▼/▲ is a transformer. ~5-8 voltage changes
  separate the generator from your toaster.

Why so many steps instead of one giant transformer? Each tier serves a different job. The 765 kV transmission backbone is for moving gigawatts across hundreds of kilometres with minimum loss. Sub-transmission feeds regional substations. The medium-voltage distribution lines on your street balance reach against the cost of insulation and the danger of high voltage near people. The final pole-top or pad-mounted transformer — the green box on the lawn or the grey can on the pole — does the last drop to the 120/240 V your house actually uses.