An Introduction to Electrical Systems

by John R. Carlsen

All electrical circuits consist of essentially two parts: a power supply and a load on that supply. A product as a whole may be thought of as a complete system (though power from an external source is often an input), with power supply and load subsystems. Among the most simple examples is a typical flashlight, which ultimately consists of only a storage battery connected to a light bulb, usually through a switch.

For most products, as the complexity of the load side increases, the complexity of the supply side increases proportionately, so the power supply side typically makes up about half of an electrical product. Readers familiar with the insides of personal computers will no doubt immediately disagree, citing that power supply modules are usually far smaller than half the enclosures; this of course is true, but misleading, as much of a computer's power supply circuitry is distributed throughout its various assemblies, mostly to provide voltage regulation at each point of load.

Power Supplies

Though the load components are obviously different for each application, the purpose of a power supply subsystem remains simple: it distributes power as needed to the system's various loads.

The functions performed by power supplies are less simple, but are still straightforward. To understand the function of a power supply, one must first understand some basic rules of power.

Within electrical circuits, power has two components: power (represented as P, the number of watts) is defined as potential energy (represented as E, the number of volts) multiplied by current (represented as I, the number of amperes or amps), as expressed in the following equation:

P = E * I

Power supply subsystems are designed to accept specified ranges of potential and current as inputs and to provide specified ranges of potential and current to their loads. Each range of potential is usually specified as a nominal voltage and allowable tolerance for variance, often expressed as a percentage of the nominal voltage (such as 5% or 10%). Current ranges are usually expressed as marginal (minimum and maximum) values, though the minimum outputs current margins are usually zero and thus not specified. (In order to run, some less-robust switching power supplies may require the presence of higher minimum loads.) For example, a specification for a personal computer's "bulk" power supply module might include a table like the one below.

A power supply subsystem provides power at each desired potential via voltage regulation circuits. Most modern power supply subsystems include several voltage regulation circuits, often supplying one from another.

Note that each connector pin may carry only up to the amount of current specified by the connector's manufacturer. To carry greater current, multiple pins are used for many outputs and for the common ground return path.

Power Budget

When crafting a system specification, one should create a power budget. A power budget is a detailed list of loads and the potential and current needed for each.

As a general rule, no load in a finished product should exceed 85% of its regulator circuit's rated capacity.

Efficiency

The overall quality of power supplies is determined by their reliability and efficiency.

Power-On Sequence

Typically, when power is applied to a state machine such as a computer, its power supply momentarily inhibits the product's operation until the potential of all of the power supply's outputs are within the allowable ranges. At this point, the power supply signals "power good" to the rest of the system, usually through de-asserting a master reset signal. This sequence of events is often referred to as initial power load.

Power States

Personal and handheld computers are complex state machines. At the top level of an electrical state machine's design, its states may be described in terms of its activity levels, which relates to its power consumption.

First released in December 1996, the Advanced Configuration and Power Interface (A.C.P.I.) specification is an open industry standard and the key element in Operating System-directed configuration and Power Management (O.S.P.M.). (wiki: ACPI)

By definition, state G3 may be entered from any state; by design, it should ordinarily only be entered from state G2. State G0 may be entered from state G2. All other states may be entered only from state G0.

Designing Battery-Operated Devices

To prevent potential data loss, all state machines including computers should store energy in a battery to provide enough power to store data to less-volatile media when a failure of external input power occurs.

What Is a Battery?

A "battery" is technically a set of many objects of a given type.

In the context of storing electrical energy, a battery is formed from cells, usually wired in series to increase their output potential, which is measured as voltage. For example, a nine-volt battery consists of six 1.5-volt cells, similar to AAAA, AAA, AA, C, and D cells.

Energy storage cells may be designed for one-time use or rechargeable operation, but both types of cells store energy through electrochemical processes. Despite improvements in construction technologies, the processes remain ongoing and volatile. So, both types of cells degrade over time. Both types have shelf lives of roughly three to five years. The performance of rechargeable cells typically degrades faster under heavier usage.

Planned Obsolescence

Because energy storage cells inherently degrade over time, designing products to use cells that are not field-replaceable artificially limits the useful lives of the products. This practice of "planned obsolescence" is environmentally unsound and, if these limitations are not disclosed prior to sale, also deceptive to consumers. So, sale of devices designed with non-replaceable batteries should be prohibited.

Perhaps with some irony, environmental activist Al Gore sits on the board of directors of Apple, which stubbornly refuses to improve this aspect of its iPod portable music players, even after settling a class-action lawsuit and subsequent design generations.

Energy Density

Note: To reduce negative environmental impact, devices should generally no longer be designed to use one-time-use (non-rechargeable) cells and batteries. This section is intended only to demonstrate a design process.

If designing a product to be powered by one-time-use (non-rechargeable) cells or a battery, that with the greatest energy density should be used.

Energy density may be calculated as amount of energy stored per unit of volume.

The table below lists approximate energy densities for conventional zinc-manganese dioxide alkaline (ZnMnO₂) cells.

As shown in the table above, AA, C and D cells offer the greatest energy densities. In addition, due to their added size, D cells may often provide an economy of scale advantage over C cells. So, AA and D cells should be used in designs in which the use of disposable energy cells is appropriate.

Summary

In summary, I offer the following design tips:

• Each electrical system consists of power supply and load subsystems, of approximately equal size.
• Each of the power supply subsystem's regulators should include a power good output connected to the system's reset signal.
• All state machines that store data should include a battery to provide enough power to store data to less-volatile media when a failure of external input power occurs.
• Cells and batteries have limited lives, so should be designed in as field-replaceable units.
• All state machines should include a simple, secure mechanism (such as a reset button) that allows the user to return the product to the state it was in when first sold.

Related Reading

For more general information on electrical engineering, see An Introduction to Printed Circuit Boards.