What Is a PCB? Printed Circuit Boards Explained

A PCB — a printed circuit board — is the flat, usually green board inside almost every electronic device you own. It does two jobs at once: it physically holds the components, and it wires them together. Instead of running individual wires between parts by hand, a printed circuit board carries thin copper conductors bonded to a sheet of insulating material, so the wiring is built into the board itself. Open a phone, a router, a washing machine, an electric toothbrush or a satellite and you will find one.

That is the short answer. The rest of this page covers what a PCB is made of, what every part is called, the types you will meet and how a bare board is manufactured — in plain English, with the real engineering numbers alongside.

What Does PCB Stand For?

PCB stands for printed circuit board. You will also see printed wiring board (PWB) in older and aerospace documentation, and PCBA — printed circuit board assembly — for a board that has had its components soldered on. In everyday use, "PCB" means the bare board, and "board" means whichever one the speaker is holding.

The word "printed" is literal. The copper pattern is created photographically: a light-sensitive film is imaged with the artwork and the unwanted copper is chemically etched away, leaving conductors in exactly the shape the designer drew. Before this became standard in the 1950s, electronics were point-to-point wired by hand — fine for a radio, hopeless for a laptop.

What a Printed Circuit Board Is Physically Made Of

A basic two-sided board is a sandwich of four materials. From the middle outward:

• The substrate (core). Almost always FR-4: woven fiberglass cloth impregnated with flame-retardant epoxy resin, cured into a rigid sheet. It is the mechanical body of the board and the electrical insulator between copper layers. The default thickness is 1.6 mm (0.063 in), which is why almost every board feels the same; 0.8 mm and 1.2 mm are common where space is tight. Standard FR-4 has a dielectric constant near 4.3 and a glass transition temperature around 130–150 °C.
• Copper foil. Sheets of copper laminated to the substrate and then etched into the wiring pattern. Thickness is quoted in ounces per square foot: 1 oz copper is about 35 µm (1.4 mil), the industry default. Half-ounce foil is used on inner layers of dense boards; 2 oz and heavier for power electronics. As a rough guide, a 0.25 mm (10 mil) trace in 1 oz copper carries around 1 A with a modest temperature rise — IPC-2152 has the real curves.
• Solder mask. The green coating, and the part people most often misunderstand. It is a photoimageable epoxy lacquer applied over the copper everywhere except the pads. Its job is not decoration: it stops molten solder bridging between adjacent pads during soldering, insulates traces against accidental shorts and keeps the copper from oxidizing. Green is a convention with good inspection contrast; other colors cost the same at most fabs and behave identically.
• Silkscreen (legend). The white ink printed on top of the solder mask: component outlines, reference designators like R1 and U3, polarity marks, pin-1 dots, connector labels and revision numbers. It carries no current and exists purely so humans can assemble, test and repair the board.

A fifth material sits on the exposed pads: the surface finish, a thin protective coating that keeps bare copper solderable. More on that below.

The Anatomy of a PCB: Traces, Pads, Vias and Planes

Once you can name the features, board photos stop looking like abstract art. These are the parts you will meet on every design.

Vias come in more flavors than the table suggests. A through via runs from top to bottom; a blind via connects an outer layer to an inner one; a buried via lives entirely inside the stack; and a laser-drilled microvia (typically under 0.15 mm) is what makes the dense HDI boards in phones possible. Each step up adds real cost, so most designs stick to through vias.

Types of PCB by Layer Count

The single biggest classification is how many copper layers the board has. More layers mean more routing space and better electrical behavior, at higher cost.

The honest rule of thumb: stay on two layers if the design is simple enough to keep an unbroken ground pour on the bottom side. Move to four the moment you have a fine-pitch BGA to escape, a controlled-impedance interface (50 Ω single-ended, 90 Ω USB, 100 Ω Ethernet), an RF section or an EMC test to pass. At prototype quantities the price gap is small enough that many engineers default to four.

Rigid, Flex and Rigid-Flex

Layer count is not the only axis. A standard FR-4 board is rigid. A flex circuit replaces the fiberglass with a thin polyimide film, so the board bends — used inside cameras, laptop hinges, wearables and anywhere a cable would otherwise be. Rigid-flex laminates the two into one part: rigid islands for the components, flexible ribbons connecting them, with no connectors to fail. There are also specialist families: aluminum metal-core boards that pull heat out of high-power LEDs, and PTFE or ceramic-filled laminates such as Rogers for microwave and RF work above a few gigahertz.

How a Printed Circuit Board Is Manufactured

Bare-board fabrication is a chemical process, not a machining one. Boards are made many at a time on a large panel, roughly like this:

1. Imaging and etching. Copper-clad laminate is coated with photoresist, exposed with the layer artwork (usually by laser direct imaging) and developed. An etchant dissolves the unprotected copper, leaving traces and pads. On multilayer boards the inner layers go first and are optically inspected.
2. Lamination. Inner cores are stacked with sheets of resin-rich prepreg and pressed under heat and pressure into one solid block.
3. Drilling. CNC machines drill every hole with carbide bits — commonly down to 0.2–0.3 mm — while microvias are drilled by laser.
4. Plating. Drilled holes are insulating inside, so a conductive seed layer is deposited chemically and built up by copper electroplating. This is the step that turns a hole into a plated through-hole and connects the layers.
5. Solder mask and legend. Liquid photoimageable mask is coated over the panel, exposed and developed so it clears the pads, then cured. The white silkscreen legend is printed or inkjetted on top.
6. Surface finish. The exposed pads get a protective coating. HASL (hot air solder leveling) dips the panel in molten solder and blows the excess off — cheap and robust, but slightly domed, so it is a poor fit for fine-pitch parts. ENIG (electroless nickel immersion gold) plates a few microns of nickel with a thin gold flash — dead flat, long shelf life, better for BGAs, more expensive. OSP and immersion silver sit in between.
7. Electrical test. Flying probes or a bed-of-nails fixture check every net for continuity and isolation against the netlist extracted from your files.
8. Profiling. A router cuts each board's outline out of the panel, or a V-groove is scored so boards can be snapped apart. Then final inspection and packing.

From Idea to Physical Board

The design side is shorter to describe than to do. You draw a schematic — the logical circuit, which parts connect to which. You turn that into a layout: real footprints placed on a real board outline, with every connection routed as copper. You export manufacturing files — Gerbers or IPC-2581, an NC drill file, a bill of materials and a pick-and-place file. A fab house builds the bare boards, and an assembly house (or your own soldering iron) populates them. The BuildPCB step-by-step guide to making a PCB walks the whole loop, and the manufacturing files guide explains what each output file is for.

That workflow is the reason PCB design has a reputation for being hard: it demands a schematic editor, a layout editor, footprint libraries and a lot of accumulated judgment. BuildPCB collapses the first two steps by generating a schematic and a routed layout from a plain-English description, then handing you standard KiCad files and Gerbers. If you would rather learn the tools by hand, start with PCB design software for beginners.

Why PCBs Exist at All

It is worth asking why an industry standardized on etched copper instead of wires. Four reasons:

• Reliability. A trace bonded to fiberglass cannot vibrate loose, chafe or be reconnected to the wrong pin. Hand-wired assemblies fail at the joints; printed ones largely do not.
• Repeatability. Every board from a panel is electrically identical to a fraction of a millimeter — which is what makes automated assembly, automated test and warranty math possible.
• Density. A 0.4 mm pitch BGA with 400 balls cannot be wired by hand. Multilayer boards route in three dimensions, which is why a phone mainboard is the size of a stick of gum.
• Cost at volume. The first board is expensive because of setup; the ten-thousandth costs a few dollars. Hand-wiring costs the same every time, and labor never gets cheaper.

There is an electrical argument too. Copper planes give return currents a short, predictable path home, which is what keeps fast digital signals clean and radiated emissions low. That control simply does not exist in a bundle of wires. If you want to go deeper on any of this, the full guide library covers layout, design rules, files and fabrication, including how AI is changing PCB design.

Frequently asked questions

What does PCB stand for?

PCB stands for printed circuit board. In older and aerospace documents you may also see "printed wiring board" (PWB), and "PCBA" refers to a printed circuit board assembly — a board with its components already soldered on.

What is a PCB made of?

A typical board is an FR-4 fiberglass-epoxy substrate, usually 1.6 mm thick, with copper foil (commonly 1 oz, about 35 µm) etched into traces and pads, a solder mask coating over the copper, a silkscreen legend on top, and a surface finish such as HASL or ENIG on the exposed pads.

Why are most PCBs green?

The green is the solder mask pigment. Green became the default because it gives good contrast for optical and human inspection, and because early mask chemistries were green. Other colors are available at most fabs for the same or a slightly higher price, with no electrical difference.

What is the difference between a PCB and a PCBA?

A PCB is the bare board — substrate, copper, mask and silkscreen, with no components on it. A PCBA is that same board after assembly, with resistors, chips, connectors and everything else soldered into place.

How many layers does my PCB need?

Two layers are enough if you can keep a mostly unbroken ground pour on the bottom side. Move to four layers when you have a fine-pitch BGA, a controlled-impedance interface like USB or Ethernet, an RF section, or an EMC requirement — the extra ground plane solves problems that are very hard to fix on two layers.