This is an edited and condensed excerpt from The Modem Reference, written by Michael A. Banks and recommended by the Associated Press, The Smithsonian Magazine, Jerry Pournelle in Byte, et al. The right to reproduce this article is granted on the condition that all text, including this notice and the notice at the end of the article, remain unchanged, and that no text is added to the body of the article. Thanks! --MB Copyright (c), 1988, 1989, 1990 Michael A. Banks All Rights Reserved (From Chapter 3) HOW TELECOMPUTING WORKS You don't have to understand how something works to use it, but understanding sure makes things easier--especially when trouble pops up. This is true of using any complex device or system--be it an automobile, a VCR, or a library. And it's especially true of using a computer in any application. When you understand even a little of what's happening "behind the scenes," as it were, you'll find that you have a lot more control over what's going on with your equipment and software. Which brings us to my purpose in writing this excerpt: I want you to have a solid picture of what's happening when your computer communicates with another system. We'll first examine the basic elements of telecomputing, followed by data formats. Then, we'll segue into how data are transmitted (focusing on hardware and software elements), and wrap up with communications parameters and error control protocols. (And, yes--you can understand all this "technical" stuff; what one human creates, another human can understand, if he or she so desires.) You won't have to pull out engineering texts to keep up with me. I'll explain terms and concepts as necessary, and you can supplement those explanations with the Glossary of Terms in Section Four of THE MODEM REFERENCE. If you're relatively new to telecomputing, read this chapter from start to finish. Otherwise, you may skim the headings, if you wish, to find only the information you need to fill the gaps in your telecomputing knowledge. TELECOMPUTING BASICS As indicated in Chapter 2 of THE MODEM REFERENCE, telecomputing (also called data communications or microcomputer telecommunications) is the transfer of data of any type between two or more computers via a transmission link. Most telecomputing (especially telecomputing involving dial- up systems) goes like this: Computer A transmits binary data (also known as digital data) to a modem, in the form of a sequence of bits. The modem converts the bits to an analog signal which mimics the distinction between the binary 1s and 0s. The analog signal is then transmitted over voice-grade telephone lines. At the receiving end, a modem connected to computer B converts the analog signal back into a binary signal that is basically a copy of what computer A sent to its modem. The modem then sends the binary signal to computer B. At this point, Computer B has in its memory a duplicate of the data Computer A originally sent to its modem. (This process works in both directions, of course.) This is a greatly simplified description of data transfer from one computer to another via modem, but it should give you a picture of how the basic elements of telecomputing (as described below) interact. (If any of the terms used in the description caught you off guard, don't despair--they are fully explained in the following pages.) Basic Elements of Telecomputing Successful telecomputing involves four major elements: data, data terminal equipment, data communications equipment, and a communications link. Various sub-elements are also involved, as described in the text following. Data. Computer data is defined as machine-readable information of any kind. The information may consist of business or personal messages, other kinds of text files (articles, contracts, documentation, reports), spreadsheets, "lines" of a realtime conference, graphic images, data base files, executable programs in binary data format, data files for programs, etc. Data are handled within a computer as binary information, and may be entered in "real time" from a keyboard, retrieved from a mass storage device like a disk, generated by a program's operation, or received via an interface from an external source. By the way, this chapter focuses more on the format and mode of data transfer--before, during, and after transmission--than on its content. The successful delivery of data with its content intact is of course the primary goal of data communications but, as you'll learn, data format in large part determines the mode-- and success--of transmission. Data Terminal Equipment. As imposing as the phrase may sound, Data Terminal Equipment (DTE) is nothing more than the computers or terminals used in telecomputing--the source and destination of data. The use of the word "terminal" in the phrase alludes to the fact that the computers or terminals involved are the beginning and ending points of data transmission--which is just what a terminal is: a place where a journey begins and ends, ala a train or bus terminal. Communications Links. Computer-to-computer data transfer may take place in a variety of ways. In its most basic form, telecomputing involves linking two computers directly, using what is called a "null-modem cable" to connect their serial ports. In this kind of setup, the computers are only a few feet apart, and data are transferred in true binary format. (Appendix E contains instructions on making a null-modem cable.) The majority of telecomputing activities, however, take place via telephone lines--either the public voice telephone system or what are known as "dedicated" telephone lines as the communications link. Although it is the least expensive and most readily available channel of data transfer, the voice telephone system cannot handle computer data in its native binary format. Thus, a modem must be used to translate data from binary (or "digital") format into analog format--a format that can be successfully transmitted via ordinary phone lines. (There is another type of phone system that does accommodate direct binary data transmission. Called a digital network, it is used in a applications where extremely speed and accuracy are necessary--but setting up and using such a system is an extremely expensive proposition.) A telephone link does not, by the way, consist solely of telephone wires. Electro-mechanical, electronic, and computer switching equipment is involved, as well as microwave transmitters and satellite up- and down-links. For the purposes of this chapter, however, such elements will remain transparent. Data Communications Equipment. Simply put, Data Communications Equipment (DCE) consists of modems and their associated interfaces, connectors, and cables. Modems have a number of functions, including but not limited to: * establishing and maintaining a communications link * translating data from digital to analog format, and vice- versa * transmitting and receiving data The interface between a modem and its computer is typically a serial interface (specifically, an RS-232C interface), although some few modems use a parallel interface. For the dial-up applications discussed in this book, you'll probably use a serial interface. DATA FORMATS AND DATA TRANSFER IN COMPUTERS As noted above, computer data must be converted from digital to analog format before it can be transmitted via voice telephone lines. Before you can understand how data are physically converted and transmitted over communications links, however, you need to understand how computers handle data internally. This section presents important information about digital data organization and handling in both your computer and portions of a communications link. Basically, I'm going to show data to you as your computer sees it. Computer Data Format Continuing improvements in computer hardware and software over the years (as well as competition among various manufacturers) have made for a veritable Tower of Babel when it comes to mainipulating, storing, and transfering information. The differences in how data are stored on, say, an IBM AT quad- density disk versus a Commodore 128 disk are so great that there is no possible comparison of the two (and there's certainly no way to transfer data between machines using one machine's disk and the other machine's drive). Too, how various brands and models of computers handle data internally differs substantially. Fortunately, all modern computers have one thing in common: they handle data in digital format (which is why they're called digital computers). This means that they "see" data characters as strings of binary digits. Note: The terms "digital" and "binary" are sometimes used interchangeably. "Digital" refers to discrete, uniform signals of any type--binary or otherwise--that do not vary in a continuous manner. Rather, such signals are identified by specific levels or values such as "on" or "off." Digital signals change immediately from one state to another, and are the antithesis of analog signals. Because we are concentrating on binary data and binary signals in this book, "digital" refers to binary data and signals. Further, almost all computers (some mainframe computers excepted) use the same numeric code to represent each character-- numbers from the American Standard Code for Information Interchange (ASCII). As you'll see in the following pages, computers that share these attributes can easily exchange data. Binary data. Even if you're not technically oriented, I'm sure you've at least heard rumors about data being stored and manipulated by digital computers in something called "digital" or "binary" data format. Binary data format means that each character (letter, number, symbol, or control character) a computer handles is operated on and stored as a specific binary number. (There are several excellent reasons for this, one of which you'll discover presently.) Binary numbers. A binary number is a string of binary digits, such as "1010" or "10011." Only 0s and 1s are used in binary notation (as opposed to the numerals 0 through 9, which are used by the decimal system). Unlike the decimal system, the values of the numerals themselves (0 and 1) are not used to determine the total value of a binary number; instead, the values of the places marked by a 1 are summed. Each place, of course, has a set value. The first place on the right in a binary number has a value of 1, the second place a value of 2, the third place a value of 4, and so on, with the value doubling with each place. Again, the value of a binary number is determined by adding up the values of the places that contain a 1. If there is a 0 in a place, that place's value is not counted. Consulting Table 3.1 in THE MODEM REFERENCE, it is easy to see that the binary number "11" is the same as the decimal number "3" (add the values of the places: 2 + 1 = 3). Similarly, the binary number 1010 is the same as the decimal number 10 (add the value of the places that contain a 1: 8 + 2 = 10). Nothing to it, right? Right! Keep this up and you'll be a math wizard in no time. Bits and bytes. A binary digit is called a "bit," by the way, and it happens to be the smallest unit of computer data. A binary number representing a computer character (sometimes called a data "word") is called a byte. The ASCII character set. I hope the explanation of binary data left you wondering what the key to the numeric "code" of binary numbers might be--which is to say, I hope it started you thinking. There is indeed a code to those numbers, a "Rosetta stone" you can use to translate the decimal equivalents of binary numbers into characters. It's called ASCII. Each and every character that a computer handles is assigned a number in the ASCII character set, which is shown in Table 3.2. (Note that only the first 128 ASCII characters are shown; certain computers--such as the IBM PC and Apple's MacIntosh--use an additional 128 ASCII characters, many of them dedicated to graphics. Not all computers recognize such extended ASCII character sets. Also note, for later reference, that the first 128 ASCII characters--0 through 127--can be represented by a string of seven binary digits, with leading 0's added as necessary.) As you may have inferred from the foregoing, a digital computer manipulates characters as the binary counterparts of these ASCII numbers. For instance, the letter "A" (ASCII 65) is represented within a computer as "1000001"--the binary counterpart of 65. Thus, any computer that uses binary data format and the ASCII character set recognizes the binary digit "100001" as the letter "A" when it is received as input--via its keyboard or from an external source such as a modem. Binary signals. Within a computer, each string of binary digits representing a character is manipulated as a discreet unit. This unit, as you may have noticed a few paragraphs back, is called a byte. A byte is normally composed of 8 bits. When a computer sends data to a peripheral--printer, modem, etc.--each byte is transmitted as a binary signal. The signal is literally a series of negative and positive voltages, as illustrated in Figure 3.2. The negative state represents a binary 1, while the positive state represents a binary 0. This particular signal contains the binary number "11000001." And yes, the order of the binary digits is reversed; that is, the figure does show "10000011," even though the number it represents is "11000001." The figure is reversed because the binary digits that make up a character are transmitted that way (i.e., the "low-order" bit [the bit with the lowest-value] is transmitted first). This is done so parity can be calculated as a character is sent (more on that later in the chapter). (NOTE: I wanted you to see this now, so you won't go slowly nuts trying to figure out why the digits are reversed in the figures in this book that illustrate binary signals. That's pretty much what happened to me the first few times I studied tutorials on data communication. None of the authors bothered to explain why the figures showed the data bit order as the reverse of what it was in the text. (Whether this was out of ignorance or due to a lack of consideration for the reader, I don't know. And I never did find a book that explains this situation; I eventually found out what was going on by asking several engineers who work in the field. (All of which is not to plug this book as the ultimate resource on data communications for the layman, but to make this point: Books and manuals are not infallible. Writers sometimes assume too much knowledge on the part of the reader, and are occasionally themselves confused. Too, the computer field is notorious for having several terms with the same meaning, or several meanings for one term. So, if something doesn't make sense when you read about it in your modem or software manual, check it out! See if this book provides the information you need. If not, ask someone who knows (or should know); contact the modem manufacturer or software publisher whose manual is the source of your confusion.) Why binary? The binary data system is used with modern computers primarily because it is faster and more reliable than analog systems. The two-state binary system is extremely simple; a signal is either there or not there, negative or positive, etc. Therefore, there is very little chance of information getting lost or "scrambled." Such information can be received and handled by a computer at extremely high rates, as cumbersome as this system is for humans to use. Using the analog technique for information transfer and manipulation, on the other hand, requires reading the relative strengths of any of a large number signals, as well as the calibration and checking of those signals. This slows down data manipulation tremendously. The electronic components used in digital computers (initially transistors and later integrated circuits of various types) exactly mimic the binary system, switching off and on (between two states) at extremely high speed. Data Transfer The astute reader will have realized by now that all that's necessary for the letter "A" to be transmitted from computer A to computer B is for computer A to send the binary digit "1000001" to computer B. If you've figured that out, pat yourself on the back--it represents the very essence of telecomputing. This all looks good on paper, but we still have to move that "1000001" from computer A to computer B. This requires a chain of devices and connections, the first link in which is a computer's port. SERIAL PORTS AND THE RS-232C STANDARD Computers aren't telepathic, and they seem immune to magic spells, so they must be physically connected before they can communicate. Moreover, they must be connected with one another or with their associated modems in a special way. The hardware medium for such a connection, whether the computers are linked via a null-modem cable or modems, is known as a port. Ports A computer's ports provide the physical (electrical) connections by which a computer's internal workings communicates with peripherals such as printers and modems. (A seaport might be used as a valid analogy for a communications port, as this is where data are sorted, organized, and shipped.) A port consists of a group of connectors in the form of pins (male) or sockets (female). Not all connectors are used by every computer, but each connector that is used is connected to a specific part of the computer's circuitry where signals are received and/or sent according to instructions from a program or the computer's operating system. (It gets more complicated, but I promised you wouldn't have to drag out the engineering texts, so we'll leave it at that.) There are two kinds of ports--parallel and serial. The name of each describes how it transfers data. Parallel ports. A parallel port sends and receives data over at least eight wires at once. This means it simultaneously transmits all the data bits that make up a character. Parallel ports are rarely used for data communication; they are most frequently used to connect computers with devices such as printers, where extremely high data transmission rates are necessary and the short distance involved won't cause problems. Serial ports. A serial port is a type of port which sends and receives data over one wire. Figure 3.3 shows typical serial port configuration. Note that Figure 3.3 shows two kinds of serial ports--one with 25 connectors and one with nine connectors. The difference in the number of connectors is a bit confusing, until you learn that not all 25 pins are used. The average microcomputer uses only eight of the pins in data communications applications (the IBM PC, among others, uses nine pins.) Serial ports and their associated cables function as an interface between computers and modems. Or, if you like technical jargon, as a "DTE/DCE interface." The RS-232C Standard This is where we un-buzz a buzz word (which should be done whenever possible). You've undoubtedly heard that this or that port or modem or cable conforms to something called the "RS-232C" standard. And you may well have wondered just what RS-232C means--and whether there's an RS-232A or B, for that matter. Well, I hope I don't shatter any cherished illusions, but RS-232C isn't as mystic as it may look in print. Let's look at it. What it is. RS-232C (often called simply RS-232) is a reference on how a serial port should physically communicate. Functionally, it is a standard for the design of the interface between a computer and its modem. This recommendation was established by the Electronics Industry Association (EIA) in 1969 to provide a standard for manufacturers of data communications equipment to follow in the design of such equipment. Semantically speaking, the designation breaks down thus: RS is an acronym for Recommended Standard; 232 is the ID number for this particular standard; and C is the latest revision of this standard. (The full name of RS-232C, by the way, is "Interface Between Data Terminal Equipment and Data Communication Equipment Employing Serial Binary Data Interchange.") What it does. The RS-232C recommendation covers the electrical and mechanical characteristics of the interface, the function of each signal (pin), and, for certain applications, secondary functions of signals. The reason for designing a serial interface that follows the RS-232C standard is to enable it to exchange data with other serial interfaces in proper form. To this end, the standard dictates how data are to be handled by the port, which indirectly dictates how the port's pins are connected to the computer's or modem's circuitry. Elements covered by the standard include voltage levels, which pins on the port are used to send and receive data, which pins detect various status signals, and more. In addition to the above, the RS-232C standard dictates that DCE (modems) use a female RS-232C connector, and DTE (computers) use a female connector. The actual configuration of a connector is not covered by the RS-232C standard, however. What a computer or modem does with a signal received at an RS-232C port is its own business; the important thing is to have the receiving device perceive data it receives in exactly the form it was sent, and at the appropriate location in its circuitry. How it works. Serial data ports that are designed in conformance with the RS-232C standard have standardized pin assignments. That is, each pin in the port's connector has a designated purpose, as shown in Figure 3.4 As you can see, the pin assignments cover just about every job that might come up in transmitting and receiving data. Again, each pin is connected to its device's circuitry as necessary to route signals to and from the appropriate elements. Thus, a computer with a properly-designed RS-232C port should, for example, send data at pin 2, and receive data at pin 3. In transfers via telephone line, each computer's serial port is connected to a modem via a cable that is wired in conformance with the RS-232C standard. (Presumably, the modem's serial port likewise conforms to the RS-232C standard.) In null-modem transfers, the computers' serial ports are connected via a cable that is wired to match the RS-232C standard for pin connections-- but with two connections reversed. Figure 3.4 covers pin assignments for DTE only, by the way. A DCE serial port (as wired for a modem) has slightly different pin assignments (such as using pin 2 to receive data from a computer, and pin 3 to send data). Other RS-232C Functions. A serial port is much more than a simple data conduit. In most telecomputing applications, the serial port performs additional tasks, some of which may not be covered by the RS-232C standard. These include organizing the parallel data bits transmitted by a computer into serial form, parity checking, adding and stripping start and stop bits, and-- on relatively rare occasions--flow control (more on those later in this chapter). Limitations. Because of the low voltages used to transmit a signal via an RS-232C port, the length, or "run," of a cable is limited to about 50 feet. If a longer cable is used, there is a good chance that data will be lost. However, the cable between a computer and modem is generally only two or three feet in length. Thus the RS-232C is ideal for this application. The maximum transmission speed of an RS-232C transmitter is 20,000 bps, but this is far faster than most telecomputing applications. Connectors and Cables The cables (and the connectors used with those cables) that connect a computer's serial port with a modem are obviously very important elements in the data communications chain. Like serial ports, connectors and cables used with serial ports must conform to the RS-232C standard. Connectors. There are two types of RS-232 connectors in common use--DB-9 and DB-25. Remember the diagram in Figure 3.3? It showed a port connector with nine pins and another with 25 pins. Nine- and 25-pin connectors are known as DB-9 and DB-25 connectors, respectively. Each type has numbered connectors (very important if you intend to make your own cables--saves a lot of messing around with a continuity tester). You'll sometimes see the letter "P" or "S" appended to DB-9 or DB-25. This letter indicates whether the connector is male ("P" for "plug") or female ("S" for socket). DB connectors are essentially "mirror images" of the serial ports on your computer and modem, and are used at each end of the computer-to-modem connecting cable. As noted earlier, the physical configuration, or shape, of an RS-232C connector is not defined by the RS-232C standard. DB- style connectors have become the de facto standard, however, simply because almost all manufacturers use them. Cables. A connecting cable is typically a "ribbon cable"--a flat cable with multiple conductors, as shown in Figure 3.5. Mixing connectors. DB-25 and DB-9 connectors can be used at opposite ends of a cable if necessary (as when a computer's serial port has a DB-9 connector and its modem has a DB-25 connector). Only eight of the pins are normally used, and all that's required for the connection to be successful is that the pins on each connector be properly wired (i.e., one connector should be wired as DCE [usually the DB-9], and the other as DTE [usually the DB-25]). Some modem manufacturers, such as U.S. Robotics and Migent, provide a DB-25 to DB-9 adaptor cable with certain models of their modems. (See Appendix E for more information on cable/connector hookups and making your own cables.) Connector and cable "gender". You may recall my saying a few paragraphs back that the RS-232 standard specifies a that female connector should be used on modems, and a male connector on computers. Thus, a "standard" RS-232C cable has a male connector on one end (to connect with the modem) and a female connector on the other end (to connect with the computer). Unfortunately, not all manufacturers follow this standard regarding the gender of their serial ports. So, with certain equipment you may find what is called a "gender problem." (No sex-change jokes please--this is serious stuff!) When this is the case, you'll have to buy or make an appropriate cable with both female or both male connectors. (See Appendix C for more information on this.) RS-232 Caveats Warning: Human nature being what it is, a few unscrupulous manufacturers and suppliers are rather more extravagant in their product claims than they should be. The result is that not every cable or device labeled "RS-232C compatible" is set up to operate exactly as a RS-232C device should. Generally, if a reputable manufacturer (or someone you trust) tells you that a cable, modem, or serial port conforms to the RS-232C standard, it's a fact . . . but watch out for Brand XYZ--especially if the company's home office is a post office box. You should of course refer to your computer's manual and your modem's manual before making connections of any type. Connecting your computer to a modem may require a cable of modified gender. You may find that, due to the peculiarities of your computer's design, you will need a special cable (as is the case with the IBM PCjr, which doesn't use a DB connector with its serial port). Or, it may happen that your computer's RS-232 port is wired slightly different from the standard. # Now that you've seen how a computer handles data internally, and how it transmits that data to the outside world, let's take a look at the next link in the data communications chain: modems. # MODEMS By now you probably have a mental picture of 0's and 1's hanging around inside a computer in groups that represent various characters, and, on demand, zipping through circuitry and pouring out a serial port and along a cable in tight formation. (If you don't have that picture, take a break, do whatever it is you do to clear your mind, and reread the preceding portion this chapter.) With that picture in mind, let's look at where data goes after it leaves a serial port, what happens to it, and why. What Does a Modem Do With Data? As you know, data intended for modem transmission normally leaves a computer via its serial port. A cable, connected to the port via a properly-wired connector, conducts the data to the modem's serial port. The modem, in the meantime, is busy maintaining the communications link that it originally established, watching for incoming data from the communications link, and watching for data and commands from the computer to which it is connected. When data arrives at a modem's serial port, the modem has two jobs to do: 1. Translate the data from digital to analog format. 2. Transmit the data via the communications link to a remote system. Similarly, when data arrives via the communications link, the modem must perform these tasks: 1. Translate the data from analog to digital format. 2. Send the data to its computer. Depending on its capabilities and the form of transmission, the modem handles flow control and performs certain kinds of error checking. These jobs are more often handled by software, however (and you'll learn more about them in a few pages.) The primary functions of a modem, then, are data translation and transmission. Data Translation: Digital vs. Analog Signals I've indicated more than once that data must be translated from digital (or binary) signals to analog signals before it can be transmitted over a voice-grade telephone line. You're already familiar with digital signals, so let's take a closer look at analog signals and why they must be used in data communications. What is an analog signal? Strictly defined, an analog signal is a signal that varies in a continuous manner, as opposed to a digital signal, which varies in a discontinuous manner. Figure 3.7 provides a visual comparison of analog and digital signals. As the diagram illustrates, analog signals vary continuously between their minimum and maximum values, while digital signals do not vary between values. A digital signal is always at either a minimum or a maximum level or value (or in one state or the other), with no in-between. Analog signals, on the other hand, cover the entire range between maximum and minimum values. Note that the analog signal in Figure 3.7 is in the form of a sine wave, while a digital, or binary, signal (as illustrated in Figure 3.2) is in the form of a square wave. Why analog? Modems were developed because the nature of the telephone system places several limitations on how data may be transferred: * Telephone lines and their associated switching equipment are designed for voice communications, which means they accommodate a limited range of frequencies. This frequency range, called the bandwidth, does not include all frequencies used by digital data. * Various amplifiers and filtering circuits involved in telephone line transmission sometimes "cut off" the upper end of a square-wave signal. * The sometimes-poor quality of telephone lines (which you've experienced aurally during long distance and even local calls) makes tolerance for variations in signal quality necessary. A computer's digital data signals must be changed to a format compatible with these limitations. To be compatible, the signal format must use the telephone bandwidth, and must be flexible enough so that minor variations in the signal do not result in lost data. Analog signals meet all the requirements dictated by the telephone system's limitations (the telephone system was, after all, designed to carry analog signals). And, as the diagram in Figure 3.8 shows, an analog signal can be made to approximate the variations of a digital signal. Data Transmission Once a signal has been converted from digital to analog format, it is a relatively simple matter to transmit the analog version by telephone lines. The signal is placed on the telephone line carrier wave via a process called modulation. This takes place in much the same way that radio transmission takes place. Carrier wave. A carrier wave is a tone transmitted on a telephone line. The carrier is a tone of constant frequency and strength--a sine wave, as shown in Figure 3.9. Carrier waves are so named because they are periodic electromagnetic impulses (waves) that literally carry information. Modulation. Modulation is a process whereby a signal is placed on a carrier wave by varying one or more of the wave's characteristics. These characteristics include amplitude, frequency, and phase. 1. Amplitude modulation. If the strength of a carrier wave is varied to carry a signal, it might appear much as in Figure 3.10 In this example, each period of maximum high and low variation might represent a binary 1, while each period of low variation might represent a binary 0. Because of the prevalence of varying amplitude on a telephone carrier (in the form of voltage spikes and other line "noise"), straight amplitude modulation is not commonly used. 2. Frequency modulation. If the frequency of a wave is varied, it might look like the wave in Figure 3.11. Each area of high frequency (i.e., when the waves are closest together) could represent a binary 1, while low- frequency waves could represent a binary 0. A simple variation of frequency modulation, called Frequency Shift Keying, is used by low-speed (300 bps) modems. Four frequencies are used in this technique, two of which represent the binary 1 and 0 of one modem, and two which represent the binary 1 and 0 of the other. The frequencies are turned on or off as necessary to transmit the 1s and 0s of each modem. (If you listen in on a telephone line while a modem using this kind of modulation is connected, you can hear the frequencies change.) 3. Phase modulation. Phase modulation is represented in Figure 3.12. Two waves of the same frequency are transmitted, and modulation is achieved by varying the time lag between the two waves. A variation on phase modulation is commonly used in higher speed modems, because it requires the smallest bandwidth of any modulation techniques. In the example in Figure 3.12, each change in phase represents a digital 1, while periods of no change represent a digital 0. (A change in phase occurs when the variable carrier wave is started at a new time in relation to the other carrier wave.) Several combinations of and variations on these modulation techniques are used in higher speed data transmission, but we won't concern ourselves with such here. The foregoing explanations and examples of modulation techniques are intended to help you visualize what happens when data are transmitted, rather than to serve as a tutorial. Data Reception You've seen how data are translated and transmitted by a modem, but what happens when a modem receives data? As you might expect, the reverse of the modulation process takes place. When a modem receives data, it senses the modulation of the carrier wave and demodulates the signal to recover the data from the carrier wave. This involves reconstructing a true digital signal from the analog signal's approximation of a digital signal. Once the translation is achieved, the modem transfers the data to the computer, sending it through its serial port to the computer's serial port. # Not incidentally, the word "modem" is a contraction of the full name of the device: "modulating/demodulating device." This name, in turn, is derived from the modem's functions of modulating and demodulating signals. Modem Standards As with serial data ports, there are recommended standards for modem operation. Rules and conventions for the design of modems have been established by several trade organizations and corporations. The standards cover the modulation and transmission techniques used by modems, as well as other elements of their operation. Virtually all BBSs, online services, and packet switching networks in the U.S. and Canada use modems that are compatible with Bell 103 (300 bps) and Bell 212 (1200 bps) standards. So, it's a good idea to use a modem that conforms to these standards, which were developed by Bell Laboratories and adopted by the telecomputing industry in the U.S. 2400 bps modems generally conform to the international CCITT V.22bis standard, described below. If your modem does not meet the appropriate standard(s), you won't be able to communicate with the vast majority of dial- up systems in the U.S.; in fact, you'll probably be unable to communicate with any system unless it uses the exact brand and model of nonstandard modem! (NOTE: The Bell standards have nothing to do with the RS- 232C standard. You must, of course, make sure your modem uses an RS-232C connector at its serial port.) If you're going to use a modem in another country-- especially in a European nation--it will probably have to conform to a different standard, known as the CCITT standard. This standard was developed by the Consultive Committee on International Telegraphy and Telephony, an international telecommunications standards committee. It is the standard for 1200 baud communications in western Europe and most of the world outside North America. Thus it is necessary to have a modem that operates on the CCITT standard if you wish to access systems outside North America. (The current versions of the CCITT standard in use are CCITT V.22 and V.32 for dial-up modems up to 1200 bps, V.22bis for 2400 bps dial-up modems, and V.29 for leased line modems.) 1200 bps modems that meet the CCITT standard are manufactured by several U.S. modem manufacturers, by the way, and it is becoming more and more common to find modems manufactured in the U.S. that recognize both Bell and CCITT standards. You'll even find some modems, such as the Hayes Smartmodem 2400, that are compatible with CCITT V.22 bis at 2400 bps, CCITT V.22 and Bell 212A at 1200 bps, and Bell 103 at 0 to 300 bps. (With a modem like this, you'll have no problem communicating with any system in the U.S. or Europe!) FCC Registration. It is worth noting here that any device sold in the U.S. for connection to a telephone must be registered with the Federal Communications Commission. You'll find evidence of such registration on a plate attached to the modem, and probably in the modem's manual, as well. Don't not buy a modem that isn't FCC registered. # If you found this excerpt useful, you may want to pick up a copy of the book from which it was excerpted: THE MODEM REFERENCE by Michael A. Banks Published by Brady Books/Simon & Schuster ISBN # 0-13-586646-4 $21.95 In addition to explaining the technical aspects of modem operation, communications software, data links, and other elements of computer communications, the book provides detailed, illustrated "tours" of major online services such as UNISON, CompuServe, DELPHI, BIX, Dow Jones News/Retrieval, MCI Mail, and others. It contains information on using packet switching networks and BBSs, as well as dial-up numbers for various networks and BBSs, and the illustrations alluded to in this excerpt. You'll also find hands-on guides to buying, setting up, using, and troubleshooting computer communications hardware and software. (And the book "supports" all major microcomputer brands.) For more information, contact: Michael A. Banks P.O. Box 312 Milford, OH 45150