Pocket calculator

The invention: The first portable and reliable hand-held calculator

capable of performing a wide range of mathematical computations.

The people behind the invention:

Jack St. Clair Kilby (1923- ), the inventor of the

semiconductor microchip

Jerry D. Merryman (1932- ), the first project manager of the

team that invented the first portable calculator

James Van Tassel (1929- ), an inventor and expert on

semiconductor components

An Ancient Dream

In the earliest accounts of civilizations that developed number

systems to perform mathematical calculations, evidence has been

found of efforts to fashion a device that would permit people to perform

these calculations with reduced effort and increased accuracy.

The ancient Babylonians are regarded as the inventors of the first

abacus (or counting board, from the Greek abakos, meaning “board”

or “tablet”). It was originally little more than a row of shallow

grooves with pebbles or bone fragments as counters.

The next step in mechanical calculation did not occur until the

early seventeenth century. John Napier, a Scottish baron and mathematician,

originated the concept of “logarithms” as a mathematical

device to make calculating easier. This concept led to the first slide

rule, created by the English mathematician William Oughtred of

Cambridge. Oughtred’s invention consisted of two identical, circular

logarithmic scales held together and adjusted by hand. The slide

rule made it possible to perform rough but rapid multiplication and

division. Oughtred’s invention in 1623 was paralleled by the work

of a German professor,Wilhelm Schickard, who built a “calculating

clock” the same year. Because the record of Schickard’s work was

lost until 1935, however, the French mathematician Blaise Pascal

was generally thought to have built the first mechanical calculator,

the “Pascaline,” in 1645.Other versions of mechanical calculators were built in later centuries,

but none was rapid or compact enough to be useful beyond specific

laboratory or mercantile situations. Meanwhile, the dream of

such a machine continued to fascinate scientists and mathematicians.

The development that made a fast, small calculator possible did

not occur until the middle of the twentieth century, when Jack St.

Clair Kilby of Texas Instruments invented the silicon microchip (or

integrated circuit) in 1958. An integrated circuit is a tiny complex of

electronic components and their connections that is produced in or

on a small slice of semiconductor material such as silicon. Patrick

Haggerty, then president of Texas Instruments, wrote in 1964 that

“integrated electronics” would “remove limitations” that determined

the size of instruments, and he recognized that Kilby’s invention

of the microchip made possible the creation of a portable,

hand-held calculator. He challenged Kilby to put together a team to

design a calculator that would be as powerful as the large, electromechanical

models in use at the time but small enough to fit into a

coat pocket. Working with Jerry D. Merryman and James Van Tassel,

Kilby began to work on the project in October, 1965.

An Amazing Reality

At the outset, there were basically five elements that had to be designed.

These were the logic designs that enabled the machine to

perform the actual calculations, the keyboard or keypad, the power

supply, the readout display, and the outer case. Kilby recalls that

once a particular size for the unit had been determined (something

that could be easily held in the hand), project manager Merryman

was able to develop the initial logic designs in three days.Van Tassel

contributed his experience with semiconductor components to solve

the problems of packaging the integrated circuit. The display required

a thermal printer that would work on a low power source.

The machine also had to include a microencapsulated ink source so

that the paper readouts could be imprinted clearly. Then the paper

had to be advanced for the next calculation. Kilby, Merryman, and

Van Tassel filed for a patent on their work in 1967.

Although this relatively small, working prototype of the minicalculator

made obsolete the transistor-operated design of the much larger desk calculators, the cost of setting up new production lines

and the need to develop a market made it impractical to begin production

immediately. Instead, Texas Instruments and Canon of Tokyo

formed a joint venture, which led to the introduction of the

Canon Pocketronic Printing Calculator in Japan in April, 1970, and

in the United States that fall. Built entirely of Texas Instruments

parts, this four-function machine with three metal oxide semiconductor (MOS) circuits was similar to the prototype designed in 1967.

The calculator was priced at $400, weighed 740 grams, and measured

101 millimeters wide by 208 millimeters long by 49 millimeters

high. It could perform twelve-digit calculations and worked up

to four decimal places.

In September, 1972, Texas Instruments put the Datamath, its first

commercial hand-held calculator using a single MOS chip, on the

retail market. It weighed 340 grams and measured 75 millimeters

wide by 137 millimeters long by 42 millimeters high. The Datamath

was priced at $120 and included a full-floating decimal point that

could appear anywhere among the numbers on its eight-digit, lightemitting

diode (LED) display. It came with a rechargeable battery

that could also be connected to a standard alternating current (AC)

outlet. The Datamath also had the ability to conserve power while

awaiting the next keyboard entry. Finally, the machine had a built-in

limited amount of memory storage.Consequences

Prior to 1970, most calculating machines were of such dimensions

that professional mathematicians and engineers were either tied to

their desks or else carried slide rules whenever they had to be away

from their offices. By 1975, Keuffel&Esser, the largest slide rule manufacturer

in the world, was producing its last model, and mechanical

engineers found that problems that had previously taken a week

could now be solved in an hour using the new machines.

That year, the Smithsonian Institution accepted the world’s first

miniature electronic calculator for its permanent collection, noting

that it was the forerunner of more than one hundred million pocket

calculators then in use. By the 1990’s, more than fifty million portable

units were being sold each year in the United States. In general,

the electronic pocket calculator revolutionized the way in which

people related to the world of numbers.

Moreover, the portability of the hand-held calculator made it

ideal for use in remote locations, such as those a petroleum engineer

might have to explore. Its rapidity and reliability made it an indispensable

instrument for construction engineers, architects, and real

estate agents, who could figure the volume of a room and other

building dimensions almost instantly and then produce cost estimates

almost on the spot.

Mark I calculator

The invention: Early digital calculator designed to solve differential equations that was a forerunner of modern computers. The people behind the invention: Howard H. Aiken (1900-1973), Harvard University professor and architect of the Mark I Clair D. Lake (1888-1958), a senior engineer at IBM Francis E. Hamilton (1898-1972), an IBM engineer Benjamin M. Durfee (1897-1980), an IBM engineer The Human Computer The physical world can be described by means of mathematics. In principle, one can accurately describe nature down to the smallest detail.

In practice, however, this is impossible except for the simplest of atoms. Over the years, physicists have had great success in creating simplified models of real physical processes whose behavior can be described by the branch of mathematics called “calculus.” Calculus relates quantities that change over a period of time. The equations that relate such quantities are called “differential equations,” and they can be solved precisely in order to yield information about those quantities. Most natural phenomena, however, can be described only by differential equations that can be solved only approximately. These equations are solved by numerical means that involve performing a tremendous number of simple arithmetic operations (repeated additions and multiplications). It has been the dream of many scientists since the late 1700’s to find a way to automate the process of solving these equations. In the early 1900’s, people who spent day after day performing the tedious operations that were required to solve differential equations were known as “computers.” During the two world wars, these human computers created ballistics tables by solving the differential equations that described the hurling of projectiles and the dropping of bombs from aircraft. The war effort was largely responsible for accelerating the push to automate the solution to these problems.The ten-year period from 1935 to 1945 can be considered the prehistory of the development of the digital computer. (In a digital computer, digits represent magnitudes of physical quantities. These digits can have only certain values.) Before this time, all machines for automatic calculation were either analog in nature (in which case, physical quantities such as current or voltage represent the numerical values of the equation and can vary in a continuous fashion) or were simplistic mechanical or electromechanical adding machines. This was the situation that faced Howard Aiken. At the time, he was a graduate student working on his doctorate in physics. His dislike for the tremendous effort required to solve the differential equations used in his thesis drove him to propose, in the fall of 1937, constructing a machine that would automate the process. He proposed taking existing business machines that were commonly used in accounting firms and combining them into one machine that would be controlled by a series of instructions. One goal was to eliminate all manual intervention in the process in order to maximize the speed of the calculation. Aiken’s proposal came to the attention of Thomas Watson, who was then the president of International Business Machines Corporation (IBM). At that time, IBM was a major supplier of business machines and did not see much of a future in such “specialized” machines. It was the pressure provided by the computational needs of the military inWorldWar II that led IBM to invest in building automated calculators. In 1939, a contract was signed in which IBM agreed to use its resources (personnel, equipment, and finances) to build a machine for Howard Aiken and Harvard University. IBM brought together a team of seasoned engineers to fashion a working device from Aiken’s sketchy ideas. Clair D. Lake, who was selected to manage the project, called on two talented engineers— Francis E. Hamilton and Benjamin M. Durfee—to assist him. After four years of effort, which was interrupted at times by the demands of the war, a machine was constructed that worked remarkably well. Completed in January, 1943, at Endicott, New York, it was then disassembled and moved to Harvard University in Cambridge, Massachusetts, where it was reassembled. Known as the IBM automatic sequence controlled calculator (ASCC), it began operation in the spring of 1944 and was formally dedicated and revealed to the public on August 7, 1944. Its name indicates the machine’s distinguishing feature: the ability to load automatically the instructions that control the sequence of the calculation. This capability was provided by punching holes, representing the instructions, in a long, ribbonlike paper tape that could be read by the machine. Computers of that era were big, and the ASCC I was particularly impressive. It was 51 feet long by 8 feet tall, and it weighed 5 tons. It contained more than 750,000 parts, and when it was running, it sounded like a room filled with sewing machines. The ASCC later became known as the Harvard Mark I. Impact Although this machine represented a significant technological achievement at the time and contributed ideas that would be used in subsequent machines, it was almost obsolete fromthe start. It was electromechanical, since it relied on relays, but it was built at the dawn of the electronic age. Fully electronic computers offered better reliability and faster speeds. Howard Aiken continued, without the help of IBM, to develop successors to the Mark I. Because he resisted using electronics, however, his machines did not significantly affect the direction of computer development. For all its complexity, the Mark I operated reasonably well, first solving problems related to the war effort and then turning its attention to the more mundane tasks of producing specialized mathematical tables. It remained in operation at the Harvard Computational Laboratory until 1959, when it was retired and disassembled. Parts of this landmark computational tool are now kept at the Smithsonian Institute.