Pioneering in Electronics
Chapter Twelve - A Science Transformed
A writer interviewing one of RCA’s laboratory directors in 1961 ventured the conclusion that solid-state materials research had done for electronic science much the same thing that the Rosetta Stone had done for the science of Egyptology.
“That seems reasonable enough, “ said his host, “except that it doesn’t go far enough. The Rosetta Stone was a key to the past. Our materials research has given us a translation in the same sense for much that we’ve seen without fully comprehending. What is even more important, though, is that it’s giving us a prescription for the future.”
What materials research has been prescribing is a new approach to nearly every facet of electronics, based upon new knowledge of the various ways that electrons behave in different types of materials. As this greater understanding has been brought to bear through the past ten to fifteen years, even those effects familiar through years of use have been so radically enhanced or transformed that they have frequently amounted to new phenomena from the viewpoint of practical application.
The semiconductor effect employed in the transistor, for example, is a significant advance in this sense beyond the same effect employed years ago in the crystal set. The magnetic effect of deflection yokes that control the scanning beam of a television picture tube is a vast improvement over (p. 303A) the phenomenon familiar through centuries past. At the same time, new light has been shed upon other phenomena that have been observed and noted by scientists in previous generations, and put aside as curiosities having no evident use.
The growing importance of materials research had led in 1953 to establishment of RCA’s Physical and Chemical Research Laboratory, first under [Douglas] Ewing and later under [Humboldt W.] Leverenz. The new laboratory incorporated research groups responsible for basic studies of electron behavior, and for preparing, developing, and analyzing new materials. In the revolutionary new environment being shaped by solid-state electronics, it provided RCA Laboratories with a potent source of fundamental knowledge and advanced technology.
This was a stimulating risk, as Leverenz wrote:
Since the story of this challenge and the response of the RCA staff carries up to the present and into the future, it is useful here to consider the principal electronic effects that have been enhanced or developed in new forms in the materials research program.
Semiconductivity, the effect used in the transistor, is a main theme with useful variations. Basically, the term describes the characteristic of many materials that are neither good conductors nor good insulators, but conduct current in controllable amounts. In crystal germanium, silicon, (p. 304) gallium arsenide and other materials used in transistors and diodes, for example, the amount of current is controlled by an applied voltage. In related types of materials known as photoconductors, it is controlled by radiant energy such as light or x-ray radiation, as demonstrated in a light amplifier or a Vidicon pickup tube.
Luminescence is another effect with variations. One of these is cathodoluminescence, characteristic of materials that emit visible light under bombardment by electrons from an external source, as seen in the phosphors on the viewing face of a television picture tube. Another is electroluminescence, a phenomenon unknown in nature, seen in man-made materials that emit visible light in response to the direct application of an electric field or current, as in the output layer of the light amplifier.
Magnetism, a time-honored property of certain materials in nature, has acquired new potency as a result of materials research. Continued progress in the improvement of such materials as ferrites has resulted in far more effective magnetic devices.
Superconductivity is a characteristic of materials that lose all electrical resistance in very low temperature surroundings close to absolute zero. Discovered in 1911 but relegated thereafter to the status of a laboratory curiosity, the effect has been enhanced by materials research to form the basis of a new electronic technology relating to computer memories and circuits, amplifiers, and high-field magnets.
Related to the principal effects are various others that defy simple classification. One of these is the photovoltaic effect, the conversion of radiant energy directly to electrical energy in certain materials, usually (p. 305) semiconductors, that are exposed to light or nuclear radiation. Another is the thermoelectric effect, a two-edged phenomenon involving the interchange of heat and electrical energy in junctions of certain materials, as employed for example in the electronic refrigerator.
The emphasis upon basic studies that has brought the enhancement of these and other specific effects also has led to a parallel advance in wider areas of electronic science and technology.
Older techniques of amplification have been improved and new ones developed, employing pumping and resonant effects provided by the motion of electrons in materials. In this general category is the laser, an acronym applied to a new family of solid-state and gas devices in which light is amplified through concentration into narrow and powerful beams of a single frequency, as opposed to the mixture of frequencies in conventional light.
With the advance in materials research has come notable progress in processing and assembling solid-state devices. This has added to the vocabulary of electronics such terms as thin-film technology and integrated electronics, describing the arrangement of devices and circuits into unbelievably small packages or single units formed of thin films deposited by evaporation onto a solid surface, or of single crystals grown in such a way as to incorporate two or more electronic functions.
Electronics research is contributing basically to the opening of new scientific vistas in the broad field known as plasma physics. Plasma is a fourth state of matter comprised of free electrons and ions—atomic nuclei (p. 307) from which the electrons have been stripped. Plasmas have unique electrical properties, and they are of special interest to electronics today in connection with energy conversion, thermonuclear power, microwave communications, and electric propulsion for spacecraft.
To these new and expanding fruits of today’s electronic research should be added an interesting side-effect that reaches into the past. Solid-state studies have led also to a further advance in tube technology by improving both the structural and electronically active materials used in tubes. For example, research leading to new techniques for bonding together electrically conducting and insulating materials aided in the development by the RCA Electron Tube Division in early 1959 of a new family of rugged miniature electron tubes known as “Nuvistors,” rivaling transistors in compactness.
These examples of growing diversity and deeper penetration in the research process itself indicate the extent to which electronics has become today a science transformed. At RCA Laboratories, the effect has been far-reaching both in the organization of research and in the nature of its accomplishments. The new emphasis upon electron behavior in materials, beginning in the mid-1950s, produced a flow of interesting and significant results, and it carried the research staff into fruitful new directions that can best be described by the process of random selection.
New Materials, New Effects
From its organization in 1953, the Physical and Chemical Research Laboratory set out to become the prime source of new basic knowledge and materials for the remainder of RCA’s research and development (p. 307) organization. By 1961, when its name was altered to Materials Research Laboratory, still under the management of Leverenz, it had achieved a number of major successes and was on its way to as many more.
The materials program was aimed along three main channels:
1) to provide materials with superior properties;
2) to develop the best methods of synthesizing these materials and fabricating them in appropriate form for use in devices; and
3) to develop understanding of the relationships between the nature of the material, the method of preparing it, and the resulting properties.
These efforts ranged across a widening spectrum of materials and effects. By the early 1960s, the program had broadened to embrace magnetics, luminescent phosphors, laser crystals, insulators, semiconductors, photovoltaic materials, superconductors, electron emitters, and thermoelectrics.
Browsing through these varied useful and promising effects, the materials research effort produced, among other results, the most effective thermoelectric power material yet devised; a new technique for growing high-temperature insulator crystals; a significant advance in magnetic materials for tape recording; improved infrared-sensitive photoconductors; and interesting new phosphors and laser crystals employing rare earths. (p. 308)
Illustrating the possible ramifications of effective materials research is the case of the photoconductors, a family of materials that had been a subject of laboratory interest for many years. By 1950, improvement in photoconductive materials provided a key to the development of the Vidicon television pickup tube. Thereafter, basic studies were undertaken by a number of research staff members, including Albert Rose, Richard Bube, S. M. Thomsen, and Henry DeVore. In the early stage of this new work, one of the most important advances was Thomsen’s development of a simple and economical method for making a cadmium sulfide photoconductive layer suitable for large surfaces.
The large-area photoconductor opened the way to the light amplifier panel demonstrated to General Sarnoff as one of his “birthday presents.” It also awakened a lively interest at Lancaster, where product engineers of the Tube Division saw its possibilities as the basis of a large-area photoconductive device that could replace the phototube in certain familiar applications—such as streetlight controls and automatic headlight dimmers for automobiles.
In the laboratory, improved photoconductors were applied in 1954 to an entirely different function as a sensitive coating on paper for a simple, high-speed electrostatic printing process known as “Electrofax.” This was a development by Charles J. Young, Harold Grieg, and Edward Giaimo, resulting from an effort to find a simpler and faster recording method to fit into a special facsimile system that had been developed a short time earlier for the Navy by H. C. Allen. (p. 309)
“Electrofax” used a special photoconductive coating that could be applied to any solid surface. The key to its usefulness was the ability of the coating to hold an electrical charge in darkness, and to give up the charge upon exposure to light. This was the basis for a rapid automatic process in which the surface was charged, an image was exposed by projection or through a transparency onto the charged area, and the resulting latent charge pattern was rapidly developed by brushing the surface with a magnetized dark powder or a liquid solution containing magnetized particles. The entire process took less than a minute even with rudimentary hand-operated equipment, and it produced a high-resolution print of any desired photographic or alphanumeric information.
Basic studies of the photoconductive material in “Electrofax” led to improvements in speed, resolution, and economy in the system. In 1959, the process was coupled with a new thin-window cathode-ray tube in equipment developed by Roger Olden for printing at speeds of up to 100,000 words per minute. “Electrofax” received a warm welcome, too, from the graphic arts industry, and a number of companies were licensed by RCA to use the process for various applications, including high-speed office copiers that began to appear on the market in substantial numbers in the early 1960s.
Having contributed directly to an important advance in television pickup tubes, a light amplifier, an automatic headlight dimmer, and an electronic printing process, photoconductor research has remained an important part of the RCA materials program in the 1960s. Among its continuing objectives have been application of thin-film techniques to (p. 310) produce photoconductive cells, and the enhancement of infrared sensitivity in certain photoconductors for use in detectors desired for military applications.
Photoconductors are perhaps unusual in their application to such a variety of end products. Other outstanding achievements in the materials program have been associated with more limited but possibly even more significant applications in such fields as computer technology, energy conversion, and integrated circuitry, and they will be cited in the account of those developments.
Among other significant aspects of RCA’s materials research have been the invention and application of new processes for synthesizing materials with certain desired characteristics. One notable example is the development by Joseph Hanak of a chemical diffusion method of depositing superconducting niobium-tin in crystal form on wire, ribbons, and larger surfaces, opening the way to practical high-field superconducting magnets.
Another is the development by George Goldsmith, Michael Kestigian, and Maxwell M. Hopkins of a new technique for producing extremely high temperatures needed for growing certain potentially useful dielectric and insulator materials. The objective has been to obtain new insulating materials that remain stable at high temperatures in order to perform active and passive circuit functions in environments hostile to devices employing conventional semiconductor materials.
Through such advances as these, RCA’s materials research program (p. 311) has become within a few short years an organized and prolific source of new knowledge and electronic effects for application by device and circuit research groups in the laboratories.
From Transistors to Integrated Circuits
The transfer of transistor technology from the laboratory to the product division in the mid-1950s had enabled the research staff at Princeton to turn its full attention to translating the new materials and effects into new devices and circuitry. And it is here, perhaps more than in any other aspect of electronics, that the revolutionary implications of the new solid-state art became most clearly evident.
The technology in 1955 was based largely upon transistors and related circuit devices that were limited for the most part to performing functions similar to those performed by tubes, albeit with considerable savings in size and power consumption. Within little more than five years, new solid-state devices had been developed to operate at frequencies far beyond the operating range of conventional transistors; a new technology had been developed that permitted devices to be literally grown in crystal layer form; a thin-film transistor had been achieved; and astonishingly small integrated circuits had started to appear on the market.
In all of these advances, substantial contributions came from the experimental and applied groups of RCA Laboratories, often in close cooperation with development teams of the product divisions.
Among the highlights of the reoriented device program was the evolution in 1958 and 1959 of the tunnel diode—a wholly new type of solidstate device capable of oscillating, switching, and amplifying at frequencies (p. 312) in the kilomegacycle range—billions of cycles per second. The project was a notable demonstration of a prime laboratory function—the recognition of potentially useful scientific developments that occur elsewhere.
The story began in early 1958 when Dwight O. North read an article published by a physicist in Japan, describing a peculiar effect occurring in a very thin semiconductor junction. North recognized the effect as being caused by an electronic process called tunneling, and it occurred to him that the extreme speed of the process might be harnessed in a device that could operate far above the frequency range of transistors.
He discussed the possibility with Henry S. Sommers, Jr., an experimental physicist of the Electronic Research Laboratory, and together the two scientists concluded that such a device could be extremely important—a verdict that was enthusiastically endorsed by another laboratory group concerned with developing ultra-high-speed computer circuitry for a government project known as Lightning.
Sommers then set about the complex task of making the device, for which no precedent existed. He was aided at various stages by chemists and metallurgists, and by microwave and electronics engineers who provided the solution to problems raised by the impossibility of coupling to the device and measuring its performance by any standard techniques. (p. 313)
Before 1958 had ended, Sommers had produced two of the devices, and tests confirmed their ability to perform basic electronic functions at kilomegacycle frequencies. There still remained the problem of converting a hand-made experimental component into a reliable and uniform device that could be reproduced in quantity. The solution came from Herbert Nelson, who devised a metallurgical technique that could be used to make many uniform tunnel diodes at one time. The new method was used first for small-scale production of the devices at the laboratories for distribution to engineers working on circuit applications. At the same time, the Semiconductor and Materials Division at Somerville began work on the devices, with such success that they were being produced in quantity for the Lightning computer project and for commercial sale by RCA before the end of 1959—less than two years from North’s original reading of the article describing the phenomenon.
The fabrication problem that entered so critically into the tunnel diode project tended to appear under one guise or another in much of the continuing work on new solid-state devices and components. This was inevitable in view of the extremely small dimensions of the devices and the infinitesimal size of the regions in which the electrons performed their useful functions.
Relatively early in the solid-state research program, it was evident that a number of new and useful devices might be obtained if it were possible to grow one semiconductor on another. It was evident, too, that the performance of transistors might be improved if it were possible to (p. 314) reduce substantially the thickness of the semiconductor layer through which the current carriers pass in moving from one junction to another.
An answer to both problems came with the development of the process known as epitaxial growth, in which a thin crystal wafer is used as a “seed” to grow another thin layer of crystal of the same material or a different material with a similar crystal structure. In this process, the wafer or “seed” crystal is held in a vapor containing ingredients that react chemically at its surface to form another thin layer of crystal. By this means, it is possible to make crystalline layers of semiconductors less than a thousandth of an inch thick, to form junctions of two different semiconductor materials (e.g., germanium and gallium arsenide), and even to grow in thin-layer crystals certain semiconductors that are difficult to grow otherwise in crystal form.
The epitaxial process evolved gradually from work in many university and industry laboratories. Studies and experiments relating to epitaxial layers were carried on at RCA Laboratories through the latter 1950s by various staff members including James Amick, Herbert Nelson, Frederick H. Nicoll, Roscoe M. Williams, and others. In 1959, a process developed at the Laboratories was introduced at Somerville for the production of tunnel diodes.
A major contribution came from the Laboratories in the early 1960s with Nicoll’s development of extremely simple and compact equipment that grew epitaxial layers much more rapidly than with earlier techniques, and could be set up in almost any laboratory. The new equipment permitted the expansion of research on many new devices that (p. 315) could not have been developed practically in the Laboratories with previous epitaxial growth methods. It also offered a more rapid and economical means for producing epitaxial devices at Somerville.
The general trend of solid-state device work at RCA, as elsewhere, thus continued through the late 1950’s in the direction of smaller elements, fabricated by new and extremely precise methods, and forming new types of circuit arrays. The logical next steps were toward fully integrated circuits employing groups of devices and their interconnections fabricated simultaneously on a single wafer base or even in a single crystal.
Initial progress toward integrated devices at RCA Laboratories resulted from a program conducted largely by J. Torkel Wallmark of the technical staff and Sanford Marcus of RCA Defense Electronic Products. They achieved in 1959 an experimental computer logic circuit incorporated entirely in a sliver of silicon so small that it could pass through the eye of a sewing needle. It was evident at the same time, however, that other approaches were needed in order to develop fully practical types of integrated circuits for general application.
Two major developments in 1960 and 1961 suggested where the solutions might lie. One was the thin-film transistor; the other, the metallic oxide semiconductor (MOS) transistor.
The thin-film transistor was developed by Paul K. Weimer and a group of associates including Joseph Dresner, Frank Shallcross, and Harold Borkan. Their achievement was the first successful production (p. 316) of operating transistors formed of thin films of material evaporated onto an insulating base. It resulted in a device so small and a technique so effective that thousands of transistors and their connections might be produced simultaneously on a wafer no larger than a postage stamp. In the original development, the new transistors were made with cadmium sulfide, a compound with considerably greater insulating properties than germanium, silicon, and other semiconductors used in standard transistors. Subsequent work brought continued improvement in the performance characteristics of the device, and led in 1963 to the development of a second type employing tellurium as the semiconductor material and involving an even simpler fabrication process.
The metallic oxide semiconductor transistor came during 1961 from an integrated devices group including T[homas]. O. Stanley, Frederic P. Heiman, and Steven R. Hofstein. It created substantial excitement because of performance characteristics which combined some of the best features of both tubes and transistors, having the simple circuit requirements of the former and the size, durability, and power economy features of the latter.
The MOS was fashioned from a wafer of silicon on which a conducting path, a high-resistance gap, and a metal control electrode were placed by a simple process of oxidation and diffusion. In operation, the control electrode, or gate, controlled the current across the gap in much the same way that the grid of a vacuum tube controls the current through the vacuum. (p. 317)
The device appeared to be a practical building block for complete integrated circuits, both in size and in ease of production. At one point in the experimental program, 2,000 of the devices were prepared in a single processing step on a square inch of silicon crystal, and more than 95 percent of the units operated effectively. During 1962, pilot production of the MOS transistors was undertaken by the Semiconductor and Materials Division at Somerville, and arrays of the devices were fabricated and used in communications and computing equipment by the Applied Research group of RCA Defense Electronic Products at Camden.
New Sources of Energy
In one of its many aspects, solid-state electronics illustrates how one technical advance tends to encourage another. Application of the new materials frequently resulted in a sharp reduction of the power needed to operate electronic equipment. Better understanding of electron behavior in materials pointed at the same time toward new techniques of converting radiant energy directly into sufficient electrical energy to power the new apparatus.
This prospect grew in importance with the rise of space technology because of the urgent need for long-lived, portable sources of power to run electronic equipment in spacecraft. The military was interested, too, in the possibility of generators that might operate in complete silence, and in power packages that could be used in remote locations over long periods.
Because some of the most promising energy conversion techniques (p. 318) employ either solid-state or vacuum tube phenomena, RCA’s scientists and engineers turned logically to their development. The result was a continuing program that received growing emphasis after the mid-1950s and produced a number of outstanding contributions.
As early as 1950, Ernest Linder and Schuyler Christian had performed an interesting experiment which led to the development by Linder, with Paul Rappaport and Joseph Loferski, of a rudimentary “atomic battery.” The device, more properly called a nuclear generator, converted radiation from a radioactive source to electric current in a wafer of silicon, producing an output sufficient to run a transistor oscillator. The experiment attracted wide attention at its demonstration in 1954, although there was little prospect of practical use for the device because the silicon deteriorated rapidly under radioactive bombardment.
This early work furnished valuable background, however, for further work by Rappaport, Loferski, and others in compiling data on radiation effects on silicon and other types of semiconductors used in solar cells. This led in turn to their development of improved solar cells having up to 100 times greater resistance than did earlier types to particle bombardment, promising substantially longer operating life in the Van Allen Belts and other possible regions of high radiation in space. An interesting sidelight to this development was the experimental employment of the new radiation-resistant cells in one of the successful Relay communications satellites designed and built by RCA for the National Aeronautics and Space Administration. (p. 319)
The solar cell development program, based on application of the photovoltaic effect, flowed logically from RCA’s work in the broad field of semiconductors. This is also the case with another portion of the energy conversion effort—the development of high-performance thermoelectric power generators.
The thermoelectric power program grew directly out of earlier work on the improvement of thermoelectric materials for cooling, in connection with Lindenblad’s development of the electronic refrigerator and air conditioner. By 1959, a 50 percent improvement in the cooling capability of these materials had been achieved by a research group under Fred D. Rosi, and emphasis was shifted to the other promising aspect of thermoelectrics—the conversion of heat directly to electrical energy.
A research team including Benjamin Abeles, George D. Cody, John P. Dismukes, and Eric F. Hockings, working under Rosi’s direction, produced a family of improved thermoelectric power materials for efficient operation at different temperature levels. The culmination of their work was the development of special techniques to obtain the most precise measurements ever made of heat flow in semiconductor materials, and their application of these techniques to achieve in 1961 one of the most significant gains in thermoelectric power—a new germanium-silicon alloy that generated power efficiently from heat sources up to 1,000°C. (1,800°F.), far beyond the upper limit for any existing practical thermoelectric power materials. The new alloy was both strong and durable, (p. 320) and of less density and weight than previous materials, adding to its attractiveness for space application. It also possessed high power density; laboratory measurements indicated that a square-foot panel of germanium-silicon units operated from a 1,000°C. heat source could generate up to 10 kilowatts of power—about three times the consumption at any given time in an average home.
The new alloy provided RCA with a unique capability in thermoelectric junctions to produce power from sources across the temperature spectrum from ordinary fossil fuel heat to the high-temperature output of nuclear reactors. Shortly after announcement of the new alloy, RCA’s Electron Tube Division received a government contract for a thermoelectric generator to be used in space with the SNAP 10A nuclear reactor to generate power in a satellite. A major engineering and production effort was launched at Harrison to develop manufacturing techniques and product designs for thermoelectric power modules.
A third approach to energy conversion at RCA Laboratories was the technique of thermionic generation, drawing upon RCA’s long experience in electron tube development. Thermionic conversion employed gas-filled tubes in which the application of heat to an electrode released electrons to flow to another electrode, producing an electrical output. The method lent itself to substantial power output from heat sources in the range of 1,100° to 2,500° C., well beyond the upper limit for thermoelectric devices. (p. 321)
From the mid-1950s, research by Karl G. Hernqvist, aided at various times by Fred Johnson and F. H. Corregan, produced a number of advances in thermionic tubes for generating power by direct conversion from solar and nuclear heat sources. A principal objective in this work was to lower the temperature levels at which thermionic tubes would operate most effectively, in order to take advantage of a greater variety of sources. Continued progress resulted from Hernqvist’s design of different internal arrangements as well as the improvement of materials used in the active elements.
By 1959, this work provided the basis for an applied research and development program organized at the Lancaster plant of the RCA Electron Tube Division under Frederick G. Block. The Lancaster team established a new thermionic tube technology based upon the flow of research results from Princeton and augmented by a substantial amount of invention and tube know-how in the product division engineering staff.
Among the various initial results of this close cooperation were an experimental tube designed to generate electricity from the exhaust heat of a rocket, and several successful designs to be employed inside nuclear reactors and with the heat-radiating circulation systems associated with reactors.
A notable aspect of these energy conversion programs at RCA Laboratories was the relative speed with which advances were achieved in all three techniques as new knowledge was obtained from research in materials and electronic phenomena. As a result, the emphasis of the (p. 322) mid-1950s upon laboratory studies and original development had shifted by the early 1960’s to a laboratory program in support of energy conversion device development in the product divisions.
Advancing the Computer Art
More than any other principal field of electronics, computer technology is a child of the solid-state revolution. Prior to the appearance of semiconductor devices, printed circuits, and magnetic memories, there was no practical way of building and operating economically a computer or data processing system having either the speed or the capacity found in any present-day commercial system.
The pioneering advances by RCA scientists in analogue computer techniques for fire control in World War II formed the root of a continuing and expanding RCA Laboratories program which ultimately probed into all aspects of computer technology, from theory to circuit design. This program turned soon after the war to digital computer techniques, and the various aspects of the total effort in subsequent years embraced such research activities as the conception and use of ultra-high-speed tunnel diode circuitry; the study and development of image-recognition techniques for computer input; the invention of high-speed read-out systems; and the various research phases of the extensive Project Lightning, a U. S. Navy program seeking a thousand-fold increase in computer speed.
Perhaps the most substantial and enduring RCA Laboratories contributions were made, however, in the esoteric art of information storage, or, more specifically, computer memories. (p. 323)
The memory is the heart of every digital electronic computer, and the capacity and speed of the memory determine the maximum capacity and speed of the computer itself. In solving complex problems having many steps, or in handling information from many sources simultaneously, the computer must have means for storing interim data that can be recalled as it is needed. This can be done electronically since the language of the digital computer is a binary code in which all information is expressed by combination of two symbols, corresponding to “0” and “1”, or, in electronic terms, “off” and “on.” Thus, detecting and handling the symbols is largely a switching problem to select one or the other. In modern computers, however, information may be handled at a rate of tens of thousands of bits per second. It is therefore critically important that the memory be able to store and recall bits of information at these speeds.
The problem of information storage attracted Jan Rajchman immediately after the war, and his first solution was the development during 1946–47 of an electrostatic storage tube containing a matrix of 256 small memory elements in which bits of “on-off” information could be stored in the form of electrical charges and recalled as needed. Known as the “Selectron,” the tube was created initially for use in a pioneer digital computer at the Institute for Advanced Study in Princeton. Coming in the early period of computer development, the device attracted such widespread interest that it was put into production at Lancaster for commercial and military users. (p. 324)
In view of progress being made at this time in materials research, the “Selectron” was in reality a stopgap solution soon to be superseded by a vastly more effective solid-state technique. The genesis at RCA of this new development was the discovery by Imre Hegyi of a new magnetic material combining magnesium, manganese, and iron in a ferrite whose magnetic field could be switched almost instantly from one direction to the opposite simply by applying an electrical pulse.
Rajchman saw immediately in the new ferrite a means for obtaining a computer memory of vastly greater speed and capacity, employing tiny doughnut-shaped ferrite cores to store ‘‘on-off’’ information in the form of a magnetic flux in one direction or the other. He placed each core at the intersection of two wires in a grid of horizontal and vertical wires. Information was stored by sending a pulse simultaneously along a horizontal and a vertical wire to reach the single core at their intersection, causing magnetic flux in the desired direction in the core. To recall the information, a read-out pulse was sent in similar fashion along two intersecting wires to determine the direction of the magnetic flux in a particular core, in a manner analogous to using a grid system to locate a given point on a map.
The ferrite matrix memory was a true breakthrough in computer storage because of its ability to store thousands of bits of information in a small area and to recall the information selectively at electronic speeds. Unknown to Rajchman and his associates at RCA Laboratories, (p. 325)
his pioneering work was paralleled by an independent program along similar lines at Massachusetts Institute of Technology. Thus, while his invention cannot be claimed as the first, it was in fact an original conception and it resulted in the first commercial application of the principle.
The new memory was the first in a long and continuing series of developments by Rajchman and his associates of memory devices providing greater speed and larger capacity in various combinations, and employing a variety of storage techniques. An early advance was Rajchman’s development, with Chandler Wentworth, of a far more compact ferrite memory formed by perforating a small plate of the magnetic material with rows of minute holes. Ingenious fabrication techniques were devised by Wentworth to produce many thousands of these aperture plates at the laboratories even before the development was announced in 1956.
The subsequent course of computer memory research was determined by both technical and economic objectives, and the work was spurred by mounting competition in the industry. One primary objective was to miniaturize the information storage elements in order to increase operating speed or, alternatively, to economize on the power needed to operate at slower speeds. Another aim was to devise a type of structure that could be mass-produced at reasonable cost.
A number of paths toward these objectives were suggested by solidstate research with its discovery of new materials and thin-film techniques. (p. 326) However, Rajchman and his associates at RCA felt that the potential of ferrites was far from exhausted, and that major advances could still be achieved in capacity, speed, and economy while enjoying the advantage of working with familiar and proven materials. The primary emphasis therefore remained on ferrite memory development, but the program was expanded at the same time to include some study and experiment with other possible approaches. Both avenues led to fruitful results.
The ferrite memory work was carried on by a group including George Briggs, Rabah Shahbender, Rolf Lochinger, Thomas J. Nelson, James Tuska, and Joseph Walentine. Among their achievements through the early 1960s were a specialized low-cost, high-capacity miniature memory for microelectronic computers, a new “microferrite” technology for building memories capable of operating speeds 15 times greater than the fastest then available, and an extremely high-capacity laminated ferrite memory formed of thin sheets in a sandwich only a few thousandths of an inch thick, yet capable of storing millions of bits of information.
From the exploration of other materials and techniques came a promising new high-speed thin-film superconductive memory developed by Leslie L. Burns with the aid of George Leck and others in Rajchman’s group. Experiments indicated that such superconductive devices, operated at liquid-helium temperatures, could provide an economical solution for systems requiring even higher speeds and greater storage capacity than could be achieved practically with magnetic memory techniques. (p. 327)
With the general advance of computer technology, the memory development program became the nucleus of an even broader technical effort under Rajchman’s general direction. Supporting the RCA Electronic Data Processing Division in the Navy’s Project Lightning, a group of laboratory specialists made significant contributions to high-speed circuit development employing tunnel diodes. Other experiments were focused upon possible use of the new thin-film transistor in computer circuitry. In another important area, various approaches were made to methods of character recognition, including an experimental system developed by Ivan Sublette and Juri Tults in the quest for a reading machine capable of processing documents typed imperfectly or printed in various fonts. A theoretical group under Saul Amarel explored problems of switching theory, systems analysis, and artificial intelligence in order to develop a foundation for improved computer design and to suggest new types of systems and applications.
Feeding directly into, and frequently interacting with, the development activities of RCA’s Electronic Data Processing, Defense Electronic Products, and Electronic Components and Devices* divisions, the Laboratories program provided a firm technical base for RCA’s expansion in one of the most rapidly growing branches of electronics.
One of the more eventful accomplishments of the new electronic science has been to demolish any lingering idea that light is useful only as a means for overcoming darkness. With the broader understanding of electrons and their behavior in materials has come a sweeping advance in methods of producing and controlling light electronically in order to display information, communicate across millions of miles of space, locate objects beneath the sea, and even weld and machine metals that defy treatment by conventional means.
In reality, the term “electronic
light” is ambiguous. It can be interpreted to mean either of
two entirely different phenomena having in common only the production
of visible or near-visible radiation resulting from electron
One of these is the luminescence caused by electronic bombardment of, or movement within, materials known as phosphors. The other is the generation of electromagnetic radiation in controllable form at the immensely high frequencies of visible light in crystal, gaseous, and semiconductor devices called lasers.
Luminescence had long been a subject of research in RCA because of the importance of improved luminescent phosphors in hastening the development of more satisfactory television pictures in the 1930s. In the immediate post-war era after 1945, the talents of the phosphor research team, headed by Leverenz, concentrated upon improved materials for the color television picture screen, with the successful results that were incorporated in the tricolor kinescope. (p. 329)
The effect employed in television was cathodoluminescence. With the greater emphasis upon materials after 1950, the new and more sophisticated effect of electroluminescence took precedence in research.
The electroluminescent effect was observed originally by a French physicist in the late 1930s, but, as with so many basic discoveries, it amounted to little until the upsurge of interest in new materials and phenomena in the postwar era. Because of the desire for improved electronic methods of visual display, attention was inevitably focused upon this demonstrated means of producing light electronically without a vacuum device. As Leverenz pointed out, electroluminescent materials compressed within a single crystal the functions that are performed by the electron gun at one end of a television picture tube and a phosphor crystal on the tube face. It was a small step from this thought to the prospect of new and versatile display devices, even including a wall-mounted television receiver formed of a thin screen of electroluminescent material with appropriate circuitry built into its frame.
Through the middle 1950s, electroluminescence was thoroughly explored by a research team including Simon Larach, Ross Shrader, and Imre Hegyi, and they developed a complete range of electroluminescent phosphors producing light in all colors of the visible spectrum as well as near-visible ultraviolet and infrared. By 1957, they were able to fabricate (p. 330) such materials to meet specific requirements of color, rate of decay, and other performance characteristics.
Further research led to improved materials emitting light under the influence of alternating as well as direct current, and their potential applications were shown in demonstration units for home lighting, automobile instrument panel illumination, and numerical displays. Elsewhere in the laboratories, the availability of better electroluminescent phosphors led to a series of interesting applied projects such as an initial exploration of mural television techniques by Rajchman and some of his associates, and the development of direct-view storage panel devices by Egon Loebner and Harvey Hook.
The continuing quest for practical thin-screen display equipment, including television receivers, received further impetus in the early 1960’s with the development of a technique for making phosphors, both cathodoluminescent and electroluminescent, in thin self-supporting sheets, instead of depositing them in the form of a layer of powder on a solid surface. Larach, with Niel Yocum and Herbert Moss, explored ways of improving the uniformity of performance in the new phosphor sheets, and the results indicated that it might be possible to devise display panels consisting of such luminescent plates with circuit and driving elements deposited at appropriate places on them.
The phosphor research group at the same time continued research in cathodoluminescent materials in the unending pursuit of improvements in efficiency and economy. In 1963, they drew upon new knowledge developed (p. 331) in laser research to experiment with a new type of cathodoluminescent material employing the so-called rare earth elements to activate light emission. One of the initial developments was a high-performance phosphor that emitted unusually bright light in a narrow band of the spectrum, suggesting a means for achieving a substantial increase in brightness in television tubes and other displays.
The result of research in luminescent materials was a solid achievement in improving upon a generally familiar type of electron action. The laser phenomenon that provided another form of electronic light was, by contrast, something entirely new and different.
The term is an acronym for Light Amplification by Stimulated Emission of Radiation. What is radiated is light, a form of energy emitted by electrons as they return to a normal state after having been raised to a higher energy level by some external stimulus, such as heat, or electrical voltage, or radiation. All light originates in this way, including that emitted by luminescent materials. But laser light differs in an extremely important way: it alone is coherent.
All other forms of light are composed of many different frequencies in the visible or near-visible spectrum, emitted by the random action of energized electrons. In the laser, the process is rigidly controlled so that the light is emitted in a single frequency, having all of its energy concentrated in a narrow beam of such intensity as to rival the sun. The result is a new form of light whose coherence lends it to use as a communications carrier of immense capacity, and whose concentrated power gives it the ability to melt any known material. (p. 332)
The theoretical basis of the laser was laid in 1958 by Charles Townes of Columbia University and Arthur Schawlow of Bell Laboratories, who asserted that electrons could be made to emit light in a controllable manner in the proper environment. Many laboratories, including RCA’s, undertook studies to achieve the proper environment, and success came first in June 1960, with demonstration of the first working laser at the laboratories of the Hughes Aircraft Company. By the end of 1960, many such lasers had been built and operated by scientists of RCA and other companies.
At RCA Laboratories, the laser program was assigned to a research group under Henry R. Lewis in William M. Webster’s Electronics Research Laboratory. Beginning with a laser formed of ruby crystal, similar to that demonstrated at Hughes, the RCA group launched an expanded program to explore other, better, materials and techniques. Their materials quest was supported by Larach and his materials research group, who first provided an effective method for growing calcium fluoride laser crystals of outstanding optical quality. At the same time, a program to explore methods of demodulating laser beams for possible communications use was undertaken by members of the Microwave Applied Research group of the RCA Electron Tube Division, working at the Princeton center.
The scope of the laser program, and its concern with a completely new effect and device, resulted inevitably in a number of significant achievements. Many new laser materials were developed, such as a new and more efficient emitter based upon (p. 333) a suggestion by Donald McClure for the use of certain types of rare earths to absorb a larger percentage of incoming light energy and to emit far more light than could earlier types of laser crystals. This advance formed the basis for further developments by Zoltan Kiss, Robert Duncan, and others in Lewis’s group. One significant example was a technique for magnetic tuning in order to modulate the amplitude or frequency of the light before its emission, pointing the way toward a practical broadband laser communications system.
A complementary advance came from Donald J. Blattner and others in the Microwave Applied Research Laboratory and from George Morton and John Ruedy in the Conversion Devices Laboratory. For the receiving end of such a communications system, they developed an extremely sensitive multiplier phototube known as the Lasecon. The device received the laser beam on a semitransparent photocathode, and the resulting emission of electrons from the cathode into the tube was enhanced by multiplier elements before entering a helix structure like that of a traveling-wave tube to induce the output signal.
Among the other facets of the expanding laser program after 1960 were the development of gas devices as well as crystals, and the application of lasers to various promising military devices by the RCA Aerospace Communications and Controls Division at Burlington, Mass., in cooperation with the research groups at Princeton. Among these applications were optical radar devices, laser range-finders, and experimental approaches to an underwater laser for detection and communications. (p. 334)
Up to 1962, the laser work concentrated upon devices that were activated by an input of light to start the laser action in the crystal or gas. During 1962, an advance of great potential importance came with the independent development at RCA Laboratories and elsewhere of laser action in an entirely new environment—the junction of a semiconductor diode. The action was caused in the junction by pumping with an electric current rather than with light, as in previous lasers, and the importance of the development lay in the greater ease with which the laser could therefore be modulated.
The diode, of gallium arsenide, was turned into a laser by altering its shape, by reducing the resistance in the contacts, and increasing the current density through the junction. The laser action occurred in the initial types when the diode was cooled to liquid helium temperature and subjected to a strong pulse current. Further experiments opened the way to a system of semiconductor materials in which the frequency of emitted light could be varied and controlled, and to a new diode laser that operated at room temperature. Contributing to this advance at RCA were many physicists and device specialists, among them Jacques Pankove, Stephen Ochs, Herbert Nelson, and George Dousmanis.
The emergence of the laser brought to electronics in the early 1960s a new concept and a new capability that would require some years of further research and development before it could be fully appreciated and harnessed to all of the practical ends which it suggested. (p. 335)
It was evident, however, that the science had come within sight of a basic goal toward which it had been striving since the 1920s. A primary objective of tube research from the earliest years had been to expand the range of useful frequencies for communications and for the growing number of other information-handling functions. This became, too, a basic objective in transistor development in the 1950s. The laser suddenly opened a door into an entirely new and vastly higher range of frequencies, at and near the visible portion of the electromagnetic spectrum, offering a potential capacity far beyond the maximum likely to be needed for generations to come. At the same time, the discovery posed a great number of new and immensely challenging problems for solution before the potential could be realized.
In a broad sense, perhaps this and the other unprecedented developments of the solid-state revolution did not alter the primary task of electronics to provide new and better means of handling information in all of its forms, from mass communications to computation. The transformation of the science itself, however, had far-reaching effects upon the type of research to be done, and the nature of the organizations responsible for doing it. The new era demanded flexible organization, new scientific skills, and highly complex facilities. For RCA Laboratories, the situation called for a new balance in research. (p. 336)
|Home | Back to Top|