From breadboard to board
You have come a long way: dividers, transistors, op-amps, filters, oscillators, power supplies, sensors, converters. Almost all of it has lived on a breadboard or inside a simulator — wonderful for learning, but a dead end for anything real. A breadboard's springy contacts add resistance, its long parallel rows add stray capacitance, and its loops add inductance, so anything fast or precise stops working the moment you build it for keeps. The printed circuit board (PCB) is where a design finally grows up: a rigid, flat, repeatable object you can make a thousand identical copies of.
A PCB is, at its simplest, a thin insulating board with a pattern of copper printed onto it. The copper is the wiring — every connection your schematic drew as a clean line becomes a real strip of metal etched onto the board. The component legs poke through holes or sit on copper landing spots, you melt solder to join them, and the whole circuit becomes one solid, vibration-proof, mass-producible piece. This rung — five guides — walks the road from a finished schematic to a board the factory can build.
The sandwich: FR-4 and copper
Pick up almost any circuit board and you are holding a sandwich. The bread is FR-4: a stiff, flame-retardant composite of woven fiberglass cloth soaked in epoxy resin. It is an excellent insulator and gives the board its mechanical backbone — that familiar greenish-tan slab. Bonded to the FR-4 is the filling: thin sheets of copper, an excellent conductor, etched away everywhere except where you want wires to run. Think of the FR-4 as the wall and the copper as the plumbing pinned flat against it.
Copper is measured not in width but in thickness, quoted in ounces: one ounce of copper spread over one square foot is about 35 micrometres thick (the everyday default). Thickness matters because copper, though a great conductor, is not perfect — every trace has a little resistance, and current through resistance makes heat (P = I^2 times R). One-ounce copper has a sheet resistance of about 0.5 milliohm per square, where a 'square' is any length equal to its width. So a trace 0.5 mm wide and 50 mm long is 100 squares, giving 100 times 0.5 = 50 milliohm. At 1 A of current that is just a 50 mV drop and 50 mW of heat — fine for a signal, something to widen for power. Wider and thicker copper carries more current with less sag and less warming.
cross-section of a 4-layer board (not to scale)
silkscreen white ink: part names, polarity, logos
solder mask green lacquer (openings expose pads)
--- TOP copper ---- signal traces + pads 1 oz / 35 um
prepreg glue + insulation
=== inner 1 ======= GROUND plane (solid copper sheet)
FR-4 core fiberglass-epoxy (the thick middle)
=== inner 2 ======= POWER plane (solid copper sheet)
prepreg glue + insulation
--- BOTTOM copper -- signal traces + pads 1 oz / 35 um
solder mask green lacquer
silkscreen white inkOne layer, two layers, many
The cheapest board has copper on one side only — fine for a simple light or a power strip, but you quickly run out of room because no two wires may cross. Putting copper on both faces (a double-sided board) lets a trace dive to the other side to dodge a crossing, and a plated hole called a via carries the signal from one layer to the next, like a tiny elevator between floors. When even two faces are not enough, the fab presses several copper layers together with insulation between them: a multilayer PCB of 4, 6, 8 or more layers. Most real products are 4-layer or more. (Vias get a guide of their own — guide 3.)
More layers buy more than just routing room. They let you dedicate whole inner layers to solid sheets of copper: a ground plane and a power plane (made by flooding the layer with copper, a copper pour). A solid plane gives every part a short, low-resistance path back to ground and a steady supply rail, and it lets a decoupling capacitor — that local water tank beside each thirsty chip — do its job, because the path from cap to chip is tiny. Why the ground plane is so transformative is the whole of guide 4; for now, just know that having one is the single biggest reason to go multilayer.
Mask and silkscreen: the finishing coats
The copper does not stay bare. Over it goes the solder mask — that signature green (or red, or black, or purple) lacquer. It covers all the copper except the spots where you actually want to solder, which it leaves open as little windows. The mask does three quiet jobs: it stops solder from accidentally bridging neighbouring pins into a short, it shields the copper from oxidation and scratches, and it insulates traces from stray contact. Picture painting a whole wall but carefully masking off the electrical outlets so they stay usable.
On top of the mask comes the silkscreen: printed ink (usually white) carrying the labels a human needs — component names like R12 and U3, the + mark on a polarized capacitor, the band on a diode, the board's name and version, warning symbols. The silkscreen makes a bare board assemblable and serviceable. Finally, the exposed copper windows get a surface finish — a thin coat of tin (HASL) or gold over nickel (ENIG) — so the copper stays solderable and does not corrode before assembly.
How parts attach, and the road ahead
Components meet the board in one of two styles. Through-hole parts have wire legs that poke through drilled, plated holes and get soldered on the far side; they are big, sturdy, and forgiving to hand-solder — the way the breadboard taught you to think. Surface-mount parts have tiny metal ends that simply sit on pads on the surface, with no hole at all; they are far smaller and denser, placed by machine and reflow-soldered all at once, and they dominate modern boards. The honest trade: through-hole is rugged and easy to rework but bulky and slow to assemble; surface-mount is compact and cheap at volume but unforgiving at the bench. Most real designs mix both.
So how does a schematic become this layered copper object? Through a two-step EDA (electronic design automation) flow: first schematic capture, where you draw the circuit as symbols and connections; then PCB layout, where you place the real footprints and route the copper. The link between them is the netlist — a plain list of every node and which pins connect to it — so the layout tool knows exactly what must be wired, even as you decide where. That hand-off is guide 2.
From there the rung finishes: guide 3 is the physical vocabulary in depth (traces, vias, pads, footprints); guide 4 is the all-important ground plane and the idea that layout is itself a circuit; and guide 5 is getting it made — a design rule check to catch errors before they cost money, then the Gerber and drill files that tell the fab where every patch of copper, mask, and hole goes, plus a bill of materials listing every part to buy. By the end you will be able to look at a board, name every feature, and understand why it was laid out the way it was.