[ Team LiB ] Previous Section Next Section

Expansion Boards

You know an expansion board as soon as you see one. It has an edge connector that fits only one place: on the motherboard made to accommodate it. But the reality of the expansion board is confused by issues of nomenclature and standards.

Nomenclature

Even the name expansion board holds a bit of confusion. Some people call them expansion cards. Is it a card or a board? In fact, the computer industry uses the two terms indiscriminately. This confused usage has a good precedent—long before computers, the electronics industry used the same terms for printed circuit assemblies. If there was any distinction, board was the generalized term (as in printed circuit board). Cards typically were smaller and usually plugged into a connector. Boards were usually screwed or bolted in place. No one cared, because both terms were generally understood.

However, the expansion boards used by notebook computers truly are cards, even though their printed circuit cards (or boards) are safely sealed inside their metal cases. They are called cards because of their similarity to credit cards. The names of the standards for them explicitly make them cards: PC Card and CardBus. These names and the similarity to credit cards were not accidental—the little expansion boards were designed to be a familiar and friendly credit-card size.

In the computer industry, several terms describe the boards used for expanding computers, sometimes with subtle differences in design or technology. Some of these terms include the following:

  • Expansion boards. The smaller printed circuit boards that plug into your computer's motherboard are most often termed expansion boards because they provide you with the means of expanding the capabilities of your computer. As noted before, the expansion board is distinct from the expansion slot, the space inside the computer chassis the board occupies (or potentially occupies), and the expansion connector into which you plug the board.

    Expansion boards are often distinguished by the standard followed by their interface or the connector at the bottom of the board. For example, an ISA board follows the Industry Standard Architecture bus standard, and a PCI board follows the Peripheral Component Interconnect standard. We'll discuss these standards a bit later.

  • Option boards. Some computer-makers prefer to describe expansion boards as option boards. You plug them into your system to add an optional feature. Strictly speaking, then, a standard equipment expansion board—for example, a graphics adapter—would not be an option board, but for consistency's sake (or maybe inconsistency), most manufacturers include such standard equipment among their options boards, perhaps to give you the idea you're getting options for free—just like that lunch.

  • Daughter boards. Strictly speaking, any board that plugs into a motherboard should be a daughter board, but in the realm of the computer, the family relationship is not so straightforward. Many boards that plug into the motherboard of a computer have special names of their own—memory modules, microprocessor cartridges, and expansion boards are all daughter boards. However, most computer hardware–makers reserve the term daughter board for add-on circuit boards that attach as a second layer to their expansion boards.

    This two-story form of packaging was prevalent when all the circuitry needed to build an expansion board just wouldn't fit in the space available in a single slot. The daughter board bought added square inches for circuitry. Today's circuits are so compact that this form of construction is rarely used. Most manufacturers don't even use the entire allowable size for their expansion board products.

  • Riser boards. As noted previously, low-profile computers reduce their size by providing horizontal slots for their expansion boards. To connect these boards to the motherboard, most use a special board called a riser board. As with an ordinary expansion board, the riser board plugs into the motherboard, but its circuit endowment comprises little more than a set of connectors to accommodate your expansion boards.

    Although computer expansion boards can all be considered daughter boards, not all daughter boards are expansion boards. For example, some computer expansion boards can themselves be expanded by plugging a daughter board onto them. Because such boards plug only into their host board, they are not true computer expansion boards. Most people call the circuit boards that plug into the motherboard the system's expansion boards. Circuit boards that plug into expansion boards are daughter boards. That convention at least relieves us of adding another generation and creating the granddaughter board.

Construction

As a concept, an expansion board might be almost anything as long as it can fulfill its purpose of enhancing the capabilities of the motherboard. As a practical matter, however, the expansion board has evolved to become a printed circuit board that fits into a connector in the motherboard and a space inside the computer chassis that makes up an expansion slot. This size and shape of an expansion board is entirely arbitrary. The expansion board could be as large as all creation or as small as a single chip.

Of course, expansion boards are not. Standards have been set on the size of expansion boards to make them interchangeable so that boards from different manufacturers will fit in as many computers as possible.

These standards are not entirely arbitrary. Several factors have influenced the choice of size. For example, an expansion board cannot be larger than the space provided by the expansion slot; otherwise, it would not fit and could hardly fulfill its expansion function. The board has to be large enough to hold the circuitry it needs to do what it has to do. Manufacturers prefer smaller boards because they cost less to make. However, the board can't be too small or it cannot hold the required expansion connector.

Several components define the physical reality of the standard computer expansion board. The board proper is an ordinary printed circuit board fabricated with pin-in-hole or surface-mount technology or a combination of both. An expansion connector connects the board to the electronic circuitry of your computer. A retaining bracket secures the board inside your computer and provides a place to put peripheral connectors.

Substrate

The substrate is the board itself, a slice of glass-epoxy upon which the various circuit components and connection traces are bonded. Computer-makers fabricate expansion boards using exactly the same technologies as motherboards.

The original design for expansion boards envisioned one end of each board sliding into a card guide, a thin slot at one end of the expansion slot, to stabilize the board inside the computer and keep it from bending or flapping in the breeze. These expansion boards that stretch from one end of the slot to the other are often called full-length expansion boards.

Most modern expansion board designs don't require all the area allowed for the substrate in the computer and are classed as short cards. Because of their diminutive dimensions and low mass, they are adequately secured in your computer by the expansion connector and their retaining brackets.

Retaining Bracket

Nearly all expansion boards have an L-shaped bracket attached at one end. Manufacturers use a number of terms for this bracket, perhaps the most colorful being "ORB," an acronym for option retaining bracket. The current trend is to refer to this bracket as the bracket.

In a computer, the bracket serves two functions. It secures and stabilizes the expansion board in its slot. It provides a mounting space for port connectors that may be required for connecting peripherals to the expansion board, and it helps shield your computer, keeping electrical interference inside your computer's case by plugging up the hole at the end of the expansion slot.

In most computers, a screw secures the bracket to the computer's chassis. When installing an expansion board, you should always ensure that this screw tightly holds each expansion board in place. Properly installing each board with a screw will prevent you from accidentally pushing the board out of the expansion connector when you plug into the connector on the board. (Tilting the expansion board can cause the contacts on its edge connector to bridge across several pins of the expansion connector, thus shorting them out and possibly crashing or even damaging your computer.) In addition, firmly screw-mounting the bracket ensures electrical continuity between the bracket and computer chassis.

Connector

The card-edge connector on each expansion board is little more than an extension of the etched copper traces of the printed circuits on the board substrate. The chief difference is that the connector pads are gold plated during the fabrication of the expansion board. The gold does not tarnish or oxidize, so it ensures that the edge connector will make a clean contact with the expansion connector on the motherboard.

The chief current expansion board standard uses the placement of pad areas and slots to key the board so that expansion boards fit only in slots designed for them.

Nearly all expansion standards for desktop computers use edge connectors for one very good reason: They are cheap. The connector's contacts get etched onto the board at the same time as the rest of its circuit traces. The only extra expense is the thin gold plating on the contact area to stave off the oxidation of the copper or lead-and-tin-coated traces.

After the pragmatic choice of an edge connector, the creator of an expansion standard still has a variety of choices. Most important is the spacing between contacts in the connector. The spacing, along with the number of contacts, determines the size. In a dream world—the one in which many designers operate—the size of the connector is no concern. In the real world, however, it has two dramatic effects. It determines how much space must be given up to connectors on the motherboard, and it governs the insertion force of an expansion board into the connector.

In true bus-style computers, the board space given up to the expansion bus is immaterial. The computer chassis is nothing but the bus, so there is no problem in devoting the whole back or bottom of the machine to the expansion bus. In the traditional computer design, however, the bus takes up space on the motherboard, which also has to provide the basic circuitry of the computer. The more motherboard space taken up by the bus, the less is available for building the basic computer. Consequently, the bus area must be as compact as possible to yield the largest possible circuit space on the motherboard.

The larger the connector, the more area of the expansion board that rubs against the contacts inside the socket when you plug the board in. To ensure a reliable electrical connection, the socket contacts must press forcefully against the contact tabs on the circuit board. Sliding an expansion board into a socket requires enough force to squeeze the board contacts between the socket contacts. The more area devoted to contacts, the greater the required insertion force. When a connector is too long, the insertion force may be greater than some people can comfortably apply, even greater than the automatic insertion machinery used by computer manufacturers can apply. Worse, if the insertion force is high enough, sliding in an expansion board may overly stress the motherboard, potentially cracking it or one of the conductive traces and putting it out of action.

Making the contact smaller shortens the connector, cutting down on the motherboard space required for the bus and reducing insertion force. It also requires greater precision in the manufacture of expansion boards. Nevertheless, newer expansion board standards are marked by closer spacing of their edge connector contacts—just compare an ISA board to a PCI board inside your computer.

More specialized bus standards, such as those for notebook computers and industrial computers, rely on pin connectors. These necessarily cost more because they add another part that must be soldered to each expansion board. But because the connector is a separate part, it can be manufactured with greater precision, and the greater precision allows for smaller, more compact connectors. Moreover, pin connectors can use more than two contact rows to further reduce the space that must be devoted to the bus. Pin connectors are also more reliable because they allow their contacts to mate on multiple sides. Pin connectors are also easier to shield, making them more desirable as concerns about emissions increase along with bus speeds.

    [ Team LiB ] Previous Section Next Section