Printed Circuit Board (PCB) Basics

Layers and their Stackup

In this class we will be building boards with four conductive layers --- so-called 4-layer boards. These boards are a lamination of various conductive and non-conductive layers. Essentially, PCBs are constructed from multiple layers of thin fiberglass. Some layers (known as "cores") are coated on one or both sides with copper foil and others are bare fiberglass (known as "prepreg" -- preimpregnated fiberclass cloth -- essentially a woven glass matt of given weave properties impregnated with a polyester epoxy resin). The thickness of each layer lies in the range from several mils to tens of mils. Layer thickness and stackup is chosen so as to have the required overall electrical properties and dimensions. Electrical connections are formed by etching patterns into the copper layers. The patterns are formed using masks and photographic techniques. The fabrication of 4-layer boards for this class requires, at a minimum, the seven masks listed below:

For this class, the boards will be assembled by hand, so-called custom hand assembly. For this we need just the seven masks listed above rather than nine. If our designs were going to be mass-produced then at least two additional masks would be required:

These additional masks specify the shapes and locations where solder paste would be dispensed in support of automated assembly. Since we are making a very small number of copies of our boards and since the soldering is all being done by hand these solderpaste masks are not needed.

Holes in PCBs

Most printed circuit boards have holes drilled in them. There are several types of holes described below. The older PCB technology, primarily supporting components with leads requiring a hole for each pin, is known as "through hole" or "PTH" (plated through hole) technology whereas the more modern style that utilizes denser surface-mounted components takes the name "SMT" or "surface-mount technology". The primary difference between these styles is the number of, size of, and usage of holes. Note that PTH and SMT components can be freely intermixed on a board. The types of holes and the terminology follows:

Holes have an implicit interaction with each of the PCB layers they pass through. As a plated-through hole carrying a signal passes through a ground plane, a (negative) clearance pad is used to ensure that the copper ground plane is kept well separated from the signal that is passing through. Similar clearance pads are also used on power plane layers. Clearance pads must be larger in diameter than the drilled hole (by approximately 15 mils).

But what do we do when we want to make a connection between a trace on either the topside or bottomside and one of the inner planes? It would be sufficient to merely omit the clearance pad; that would result in a solid connection between the plated wall of the through hole and the copper plane. Since inner planes are intended to be used for power supply and ground connections, however, we expect relatively large currents to flow in these locations. Usually, these planes are made thicker, so that they can handle higher currents. And, since copper is an excellent conductor of heat, the power and ground planes also provide a heat-spreading or heat-sinking effect. This can make it extra difficult to solder to pins that connect to planes .... they can take a great deal of heat. To improve the solderability of pins that connect to planes, we normally provide a tiny "expansion joint" at these locations. The implementation of such a joint is known as a thermal pad, or more simply, just a thermal. The image below shows a region of an inner plane with some thermals of various sizes. You can recognize them by their "X"-shaped structure. In this image, the green material represents copper and the black areas represent no copper. The hole in the middle of each thermal will be drilled out, leaving a plated-through attachment between the voltage on the plane and whatever is connected to the through-hole on the topside and/or bottomside. The structures that look like a donut (with no "X" across) are a type of clearance pads. They isolate the plated-through hole from the plane to ensure that no connection is made.

The two lines with via holes on each end are traces with isolation around them. Once a plane region has been isolated by surrounding it with non-copper space, it can be used like a regular outer plane, including routed traces. It is normally better to use the inner planes for power and ground connections --- i.e. as planes --- but there are sometimes good reasons to include routing on these planes.

Packages and Component Footprints

Printed circuit boards are all about circuits, i.e. interconnections among electronic components. Most components are soldered to PCBs. In order to provide the right environment for soldering to be effective, the leads of a component must make reliable contact (both electrical and mechanical) to the etched copper traces. Each component must have an appearance on the board; this appearance is referred to as the component's land pattern. A component's land pattern is a set of etched copper features that directly corresponds to the leads of the component. These features are normally made a bit larger than the component leads so there is space for the solder. In the Mentor Graphics tools, these land patterns arise as "cells". There is a tutorial within the Library Manager section about how to make a "cell" as well as how to use them in a design.

Below is a list of packages (incomplete) currently used in the electronics industry.

Forbidden Packages for ECE189 Projects

For cost, handling, and debugging reasons we do not allow use in this class of some of the more aggressive electronic packaging that is available on the market. A partial list of forbidden packages follows:

Solder Masks

As component packaging shrinks in size, the effort required to solder components on boards increases dramatically. The soldermask layer is a tough semi-transparent material coated on both sides of a PCB. It is what makes the PCB green (actually soldermasks come in other colors as well but the boards for this class will always be green). The purpose of the soldermask is to define the regions where components can be soldered to pads or traces. The soldermask has openings directly over each of the spots where a component lead touches its land pattern on the board. When there are multiple component leads densely packed in a row it is the soldermask that helps to prevent solder bridges (unwanted extra connections between adjacent solder points).

Silkscreens

The last layers that go onto your PCB during the manufacturing process are the silkscreens (sometimes referred to as legends). There may be slikscreens for the topside, the bottomside, or none at all. The silkscreen layers specify any painted graphics (normally white in color) that are to appear. In an effort to keep our budgets under control we normally request only a topside silkscreen though if your project has components that need to be soldered on the bottomside it will be necessary to request a bottomside silkscreen as well.

Silkscreens may have just about any graphical content --- e.g. logos, project name, designer name(s) --- though there are certain expectations associated with the electronic assembly process. One expectation is that each and every component should have its unique reference designator (aka "refdes") printed nearby the component footprint and there must be some sort of outline or clear indication on the silkscreen to indicate how the component is to be oriented (e.g. where its pin #1 is located). For polarized parts such as diodes, LEDs, electrolytic capacitors, etc., it is imperative that the polarity is marked on the board (via the silkscreen). It is customary to place a small "+" sign and/or a "-" sign near the appropriate pin of polarized parts. Often these outlines and refdes positions are a built-in part of each "cell" and if you make your own cells you must be sure to include them.

The assembler relies on the reference designators to find each part in its little bag that is in the parts kit we provide him. This single marking on the board is the only link between the parts kits, the bill of materials, and the bare printed circuit boards. If your board has one or more components that are not identified by their reference designators on the silkscreen, then the assembler will simply skip them. The refdes field is a critical link for assembly. The assembler will simply ignore any "extra" parts kit components.

How printed circuit boards are made and assembled

Here is a short video (running just under 5 minutes) that shows how printed circuit boards are made and assembled. If you have trouble with the embedded viewer, try an external viewer here.