2.1 PCB materials and layers
The
The substrate: FR-4 and the “Tg” factor
Section titled “The substrate: FR-4 and the “Tg” factor”The vast majority of modern electronics are built on a substrate called FR-4 (Flame Retardant 4), which is a rigid composite made of woven fiberglass cloth saturated with epoxy resin. This composite acts as the non-conductive mechanical skeleton of the board. However, while “FR-4” is the generic industry category, the most critical explicitly defined specification is the material’s Tg (Glass Transition Temperature).
Understanding Tg (the softening point)
Section titled “Understanding Tg (the softening point)”It is crucial to understand that Tg is not the melting point. Rather, it is the specific temperature threshold at which the normally rigid epoxy resin becomes soft and rubbery. When the material crosses this thermal threshold, it expands rapidly in the Z-axis (thickness).
- The Engineering Reality: Modern
lead-free soldering processes requirereflow oven temperatures to reach up to 260˚C. - Standard Tg (130˚C – 140˚C): This is generally acceptable for simple, low-cost consumer electronics that only have a few layers.
- High Tg (170˚C+): This is highly recommended for industrial equipment, automotive electronics, or any high-density board with 8 or more layers.
- The Trap: When the resin expands too much during
reflow because a Low Tg material was specified for a complex board, the tiny copper barrels plated inside the vias can stretch and fracture (Via Cracking). This causes intermittent open circuits that are notoriously difficult to debug.
The stackup: layers as a sandwich
Section titled “The stackup: layers as a sandwich”A bare PCB is manufactured by pressing and laminating alternating layers of conductive copper and insulative substrate under intense heat. The “Stackup” is the highly specific and engineered arrangement of these individual layers.
Core vs. prepreg
Section titled “Core vs. prepreg”- Core: This acts as the foundation. It is a rigid, fully cured piece of FR-4 with solid copper foil already permanently bonded to both sides.
- Prepreg: These are uncured, “sticky” fiberglass sheets used as the binding layer placed between the rigid cores. Inside the lamination press, extreme heat and pressure melt the Prepreg, permanently fusing all the layers together into a solid board.
2-Layer board
Section titled “2-Layer board”- Structure: Top Copper – Solid Core – Bottom Copper.
- Use Case: Simple, heavily cost-optimized, low-speed circuits (like toys or basic breakout boards).
- Limitation: A 2-layer board offers zero electromagnetic shielding. Fast signals running on the top layer are highly likely to bleed through and interfere with signals running on the bottom layer.
4-Layer board (the industry standard)
Section titled “4-Layer board (the industry standard)”- Structure: Top Signal – Continuous Ground Plane – Continuous Power Plane – Bottom Signal.
- Advantage: The internal, solid sheets of copper act as effective electromagnetic shields. They provide a stable voltage supply to all components and significantly reduce emitted
electromagnetic interference (EMI). - Actionable Rule: When a design includes any high-speed digital signals (like USB, WiFi, or fast memory), a 4-layer board becomes the minimum requirement to ensure signal integrity.
Multi-layer (6, 8, 12+ layers)
Section titled “Multi-layer (6, 8, 12+ layers)”- Use Case: Extremely high-density routing (Smartphones, Motherboards, dense computing modules).
- Cost Driver: Every additional pair of layers requires an entirely new, time-consuming lamination cycle in the factory. Moving a design from 4 layers to 6 layers will often increase the cost of the
bare board by 30% to 50%.
Copper weight: current vs. definition
Section titled “Copper weight: current vs. definition”In the PCB industry, copper thickness is traditionally measured in ounces per square foot (oz/ft²), usually just called “ounce copper.” This physical thickness directly dictates how much electrical current the routed traces can safely carry before overheating.
- 1 oz (35 µm): This is the standard. It provides a balance between decent current-carrying capacity and high-precision chemical etching.
- 2 oz (70 µm): Used for high-power applications. This should be explicitly specified for heavy motor controllers or main power supplies to handle high-amp loads without overheating the board.
- 0.5 oz (18 µm): Used for fine-pitch designs. This ultra-thin copper is often required for highly dense digital boards where the traces must be incredibly narrow.
The engineering trade-off
Section titled “The engineering trade-off”Heavier, thicker copper is exponentially harder for the factory to etch accurately.
- The Trap: Specifying thick 2 oz copper to handle power precludes using tiny 4 mil trace widths for data lines on that same layer. The acid in the factory baths cannot etch straight down through that much thick copper without also eating sideways into the trace walls (a defect known as “undercutting”).
- The Result: Using thick copper forces a significant increase in trace spacing, which drastically reduces the overall routing density of the board.
Final Checkout: PCB materials and layers
Section titled “Final Checkout: PCB materials and layers”| Feature | Standard Spec | High Performance | The Engineering Rule |
|---|---|---|---|
| Material | Standard Tg (135˚C) | High Tg (170˚C) | Use High Tg for >6 layers, automotive, or harsh thermal environments. |
| Copper | 1 oz (35 µm) | 2 oz (70 µm) | Thicker copper requires wider spacing between traces. |
| Layers | 4 Layers | 6+ Layers | Always use continuous internal Ground/Power planes for complex logic to prevent EMI. |
| Structure | Balanced | Balanced | Symmetry is a physical law. The stackup must be symmetrical from the center out to prevent the board from warping (potato-chipping) during oven heating. |