Sharp QT-8D Electronic Calculator
The Sharp QT-8D, also known as the "micro Compet", has the distinction of being recognized as the first electronic calculator to have its calculating logic implemented entirely with MOS (Metal-Oxide Semiconductor) large-scale integrated circuitry. While essentially true, Hayakawa Electric Co. (renamed Sharp Corp. on January 1, 1970) was actually beaten to the punch by nearly three years by Victor Comptometer and their Victor 3900 electronic calculator that had its logic implemented entirely with bleeding-edge state-of-the-art MOS Large Scale Integrated (LSI) circuitry. The QT-8D was introduced in March, 1969 at the IEEE exposition in New York, NY.
Victor Comptometer introduced the Victor 3900 electronic calculator in October, 1965. The Victor 3900 used large-scale (for the time) MOS integrated circuits for all of its calculating logic. (For the story of the Victor 3900, see the Old Calculator Museum's essay).
The calculating logic of the Victor 3900 used 29 IC's, whereas the QT-8D used only four, clearly indicating the advancement of MOS LSI technology in the just over three years between the two calculators. In terms of historical accuracy, the Victor 3900 really was the first to use large-scale MOS integrated circuits as the sole means of implementing the calculating logic of an electronic calculator. The Victor 3900 was not made in large quantities (somewhere between 1000 and 2000 were made), and had a relatively short market lifetime, causing its historical significance to be diminished. While the 3900 may have been first, the Sharp's development of the QT-8D was most certainly a breakthrough, paving the way for MOS/LSI to dramatically increase the capability, decrease the size, and more importantly, reduce the cost of electronic calculators.
The QT-8D's Dust Cover
Desktop electronic calculators were the first consumer application for digital integrated circuits, and thus were prime candidates for further miniaturization through the use of MOS Large-Scale Integrated circuit technology. Prior to the debut of Metal Oxide Semiconductor techniques, integrated circuits utilized bipolar technology that required fairly complex structures to create transistors. This limited the amount of logic that could be practically placed on a single chip. In the early days of MOS ICs, only a few logic circuits could be placed on a chip simply due to limitations of the fabrication technology. With bipolar or early MOS ICs calculators would require 100 or more small and medium-scale integrated circuits, along with fairly substantial quantities of individual discrete components such as transistors, diodes, and resistors. An example of Sharp's pre-LSI calculator technology utilizing small and medium-scale MOS integrated circuits is the Sharp Compet 16.
Size Comparison, QT-8D(Left), Compet 16(Right)
At that time, Small-Scale Integration (SSI) IC's could contain the equivalent of 20 or so logic gates, combined to form a flip-flop or two, or a selection of a few logic elements. Medium Scale (MSI) integrated circuits were able to contain more logic, topping out at around 60 to 100 or so logic elements per chip. MSI devices provide higher level functions, examples of which would be an 8-bit shift register, 4-bit data selector, or binary full-adder. Later medium-scale MOS (Metal Oxide Semiconductor) devices could place an entire 64-bit shift register in a single chip. These devices paved the way for the development of Large Scale integrated circuit technology. LSI devices made it possible to cram many hundreds, and later, thousands of logic elements onto a single chip. Today, millions of logic elements make up the chips that are in our PCs, tablets, cars, TVs, and smart phones. The advancement of integrated circuit technology made it possible for an entire subsystem of an electronic calculator (for example, the arithmetic unit) to be integrated onto a single chip by late 1967.
One of the ICs used in the QT-8D
LSI devices, while initially expensive to develop, became easier and easier to manufacture as the processes became better understood and more refined. Use of LSI integrated circuits in calculators allowed calculator manufacturers with access to LSI chip fabrication technology (or alliances with those that had the technology) to substantially reduce their costs to make a calculator, which in turn resulted in lower cost for the consumer.
The story of the development of the QT-8D, as well as the emergence of Hayakawa Electric as a major force in the Japanese electronics industry, is a rather long journey. It follows the pattern of many other successful high-technology businesses. The force behind Sharp's dominance in the electronic calculator marketplace was driven by a gifted individual, someone with the genius, vision, discipline, and persistence to lead the company into a new realm. For Sharp, that man was Tadashi Sasaki(5/12/1915-1/31/2018).
Tadashi Sasaki was born in May of 1915. The exact place of his birth isn't known for sure. He may have been born in Taiwan, or in Japan. In any case, Sasaki considered his homeland to be Japan. As Sasaki progressed through his schooling with a focus of study being Japanese literature, one of his teachers, seeing his curiosity and aptitudes for problem solving and creative thought, recommended to him that he should seriously consider re-directing his studies into the sciences. Sasaki took this to heart, and began studying the sciences with great energy and diligence.When he finished his primary schooling, graduating at the top of his class, he had an opportunity to go to Germany to study at prestigious Dresden University, which he quickly seized. He graduated from Dresden University in 1938 with a doctoral degree in electrical engineering, as well as being named as an honorary professor of the university. After earning his doctorate, the now Dr. Sasaki returned to Japan and joined the Electrotechnical Laboratory, an arm of the Japanese government's Ministry of Telecommunications that was devoted to development of advanced electronics technology. During this time he worked on various electronic technologies, including improved telephone communications systems and improvements to the design of vacuum tubes. As the flames of World War II ignited, Sasaki, under direction of the wartime Japanese government, was assigned to work for a Japanese aircraft manufacturer, Kawnashi Kogyo (renamed Kobe Kogyo after the war). At Kawnashi Kogyo, Sasaki worked on the design of radar and anti-radar systems for Japanese war planes. After World War II ended, the United States began a process of helping to reform and rebuild Japan. Sasaki became involved with teams of engineers from the U.S. focusing on rebuilding the telecommunications infrastructure in Japan, which was largely destroyed during the way. As a result of this work, Sasaki was able to travel to the U.S. to study telecommunications technology at technology centers such as Western Electric, RCA, and Bell Laboratories. In 1947, at Bell Labs, Dr. Sasaki had occasion to come into contact with Dr. John Bardeen, one of the co-inventors of the transistor. In late 1947, immediately after Bardeen told Sasaki about the transistor, Sasaki went back to Japan to try to spearhead development of transistor technology there. Sasaki had the foresight to realize that transistors were going to be a huge revolution in electronics. He wanted his homeland to be able to reap the monetary and technological benefits that development of semiconductor electronics would bring. Upon his return to Japan, Sasaki went back to work for Kobe Kogyo. He worked at continuing to improve and miniaturize vacuum tube devices, and later began work on semiconductor technology. His valuable contributions to Kobe Kogyo continued through the 1950's and into the early 1960's. During this time, Sasaki was very active in promoting semiconductor technology to the Japanese government, to encourage Japanese academia to begin training budding engineers in semiconductor technology, as well as working with business to foster a commercial interest in transistors. Through his efforts, a strong momentum built behind research, development, and commercial realization of transistor technology in Japan.
Sumlock Comptometer's Anita Mk 7In 1962, something happened that captured Dr. Sasaki's attention and interest. Dr. Sasaki learned of the development of the first desktop all-electronic calculator in the UK, called ANITA. With strong sales of the Anita Mark 7 (Model C-VIII) in the UK, Sumlock Comptometer introduced a somewhat improved version for world-wide sale, the Anita Mk 8, which was an instant hit around the world, including in Japan. Even though the Mk 8 was more expensive than existing motor-driven mechanical calculators, the benefits of its much faster and silent operation worth the additional cost. The ANITA calculators utilized cold-cathode tube technology that was invented in the mid-1930's. By the time the ANITA calculators were introduced, this technology was approaching obsolescence due to the commercial introduction of transistors. Sasaki knew that the ANITA calculators were a technological dead-end. He was positive that the transistor could do the same to a calculator that it did to the pocket AM radio. The transistor had reduced the AM radio from a tabletop box with vacuum tubes that was tethered to the AC power outlet to a pocket-sized, battery-powered device that could go anywhere, not to mention making the radio less-expensive and much more reliable. While an electronic calculator is vastly more complex than a transistor radio, Dr. Sasaki felt that transistors would eventually become so small that an electronic calculator could be made small enough for a person to carry with them all the time, perhaps in a shirt-pocket, just as the transistor had enabled with the AM radio. Little did Sasaki know at the time how quickly his vision would come to pass. By late 1963, Sumlock's ANITA calculators were selling like hotcakes all over Europe. At the same time, Fujitsu, a giant Japanese manufacturing conglomerate, acquired the company that Sasaki was working for. By the time of the acquisition Dr. Sasaki had advanced through the management ranks of Kobe Kogyo and was a now a member of the board of directors. Although Fujitsu realized his tenure, Dr. Sasaki didn't relish the idea of becoming just another cog in the wheel of giant Fujitsu. This prompted him to leave, taking a job at consumer electronics manufacturer Hayakawa Electric Co., Ltd. (which became known as Sharp Corporation in 1970). Sasaki already had contacts at Hayakawa Electric, because they had been a consumer of electronic components manufactured by Kobe Kogyo for use in radios and televisions. Dr. Sasaki knew that Hayakawa Electric (hereafter referred to as Sharp) was very interested in the potential of electronic calculators, and in fact had tasked a small group of its most talented engineers with the development of an electronic calculator. Sharp executives knew of Sasaki's expertise with semiconductor technology. Sharp's hiring of Dr. Sasaki was a classic win/win scenario. None of Sharp's executives could have guessed just how much of an effect on the company the hiring of Dr. Sasaki would have.
The Sharp Compet 10, Sharp's First Electronic CalculatorImmediately after starting at Sharp, Sasaki became heavily involved with the team working on the electronic calculator developmnt, bringing a great deal of his skills to bear to complete the design and prototyping of what became the Hayakawa Electric the Sharp CS-10A (also known as the Compet 10), announced in March of 1964. The Compet 10 was among the earliest transistorized electronic calculators in the world, introduced at nearly the same time as the Friden 130 and Mathatron calculators made in the US; and the IME 84 made in Italy. The Compet 10 was a fully-transistorized calculator using Japanese-made Germanium transistors manufactured by Nippon Electric (NEC) and Hitachi. Germanium transistors were prone to reliability problems, somewhat difficult to manufacture, and were also quite expensive. Sasaki knew that transistors made from Silicon rather than Germanium would solve a lot of the problems inherent with Germanium transistors of the Compet 10, and immediately started up a project to develop a calculator based on Silicon transistors. At the same time, he incorporated improvements in the architecture of the calculator to make the machine easier to manufacture, as well as making the calculator easier to use. The result of this project was the Sharp Compet 20 (CS-20A), announced in September, 1965. The Compet 20 was markedly smaller, lighter, much more reliable, and much easier to use than the Compet 10. These traits made the Compet 20, and a few follow-on machines, the Compet 21, 0 and 15, great commercial successes. This success provided a firm foundation for Sharp to build the very strong electronic calculator business that still prospers to this day.
The Sharp Compet 20Like any visionary, Dr. Sasaki was always looking forward. While the Compet 20 was a wonderful success, Sasaki was not content to rest on his laurels. He had been keeping a close eye on the development of a new type of semiconductor technology in the United States. The technology involved the integration of many semiconductor devices onto a single chip of silicon. Like he did with the transistor, Sasaki knew that the Integrated Circuit, or "IC", as it had come to be known, would make it possible for calculators to grow more capable, as well as shrink dramatically in size, potentially making Sasaki's dream of a pocket-sized calculator a reality. At around the time that the Compet 20 was introduced, Victor Comptometer stunned the world with the introduction of the Victor 3900. This machine was proof to the world that Large Scale MOS ICs had the potential to revolutionize the electronics industry. Dr. Sasaki realized that if Sharp didn't have this technology, the company would not be able to retain its leadership position in the electronic calculator industry, nor would his vision for Hayakawa Electric to develop a pocket-sized calculator ever become a reality. The technology used to make the ICs in the Victor 3900 had its genesis at Fairchild Camera and Instrument in the US. The IC's in the 3900 were designed and manufactured by a spinoff from Fairchild called General Micro-Electronics (GM-e). The folks that founded GM-e had performed a lot of the research into the development of large scale MOS devices at Fairchild, and had published some research papers describing how to make MOS chips with large numbers of transistors on them. The founders of GM-e had become frustrated because it seemed like Fairchild was content to rest on the laurels of its bipolar integrated circuit technology and not invest in MOS ICs as anything but research projcts. As the news of Large Scale MOS integrated circuit technology becoming a reality in the US spread, Dr. Sasaki made a tireless effort to learn as much as he could about MOS/LSI technology. Sasaki made a number of trips to the United States, making arrangements to meet with the best engineers in the field. In early 1966, Sasaki had amassed the information he needed, and put his best engineers to the task of developing large-scale MOS devices within Hayakawa Electric. In late 1966, world was leaking out that Texas Instruments had developed a prototype of a four-function, battery-powered handheld printing calculator called Cal-Tech that used only four Large-Scale ICs. These developments of advanced IC technology in the US further drove Sasaki to push his engineers to the edge to develop Sharp's own advanced MOS IC technology. Over the next year and a half, Sharp's engineers toiled at developing their own PMOS (P-Channel Metal Oxide Semiconductor) IC technology, and had built some prototype chips that had respectable levels of integration. It was during this time of frenzied development that Dr. Sasaki earned the pet nickname of Mr. Rocket, referring to his jet-set lifestyle of hopping around the globe to learn the latest developments in technology that could help his company lead the world in the development of electronic consumer goods, with the chief money-maker at the time being the electronic calculator.
The Texas Instruments Cal-Tech Prototype Calculator, a source for inspiration for Dr. Sasaki
Image Courtesy Texas Instruments, Used with Permission
The Built-in Carrying Handle, and Model/Serial TagThe introduction of the QT-8D created major turmoil in the calculator business. The competitor's calculators were large, heavy, power-hungry, and cost between $500 and $1300. The QT-8D was small, easily portable (it would fit comfortably inside a briefcase), used only seven Watts of power, and came with a built-in carrying handle that made it easy to carry. Sharp and Rockwell were raking in the profits, while other calculator manufacturers had to scramble to cut prices (and thus margins) on their existing calculators, as well as having to make large investments to forge their own alliances with LSI chip manufacturers. The QT-8D marked the beginning of a shakeout in the calculator industry, with many players, including the likes of Busicom, ending up getting out of the calculator business, or out of business altogether. The QT-8D wasn't particularly fancy in terms of its capability. It is a basic eight digit, four function, AC-powered portable desktop calculator with automatic floating decimal. It has no memory capabilities or additional math functions. However, that didn't really matter in the market that the machine was designed to sell into. The QT-8D was targeted at business, where the need for small, easy-to-use machines that could add, subtract, multiply, and divide was very strong. The QT-8D was even marketed (see an early Advertisement for the QT-8D and it's later battery-operated counterpart, the QT-8B) to affluent executives as the perfect briefcase wonder to help them grind through their latest business figures, as well as helping to manage their own personal finances.
Inside the Sharp QT-8DMoving inside the machine, the QT-8D is truly a marvel of electronics in its day. The entire guts of the calculator fit on two circuit boards, stacked one atop the other, both plugged into an edge-connector backplane that interconnects the boards. The backplane itself is a small printed circuit board with traces connecting the various edge connector pins. The circuit boards are made of phenolic, with traces on both sides of the boards, and components only on the top surface. Plated-through (along with solder-filled) feed-throughs provide interconnection between the two sides of each board.
The Unique Digit Rendition of the Itron DisplayThe top circuit board contains the display subsystem, consisting of the unique 8-segment Itron display tubes, discrete transistor display driver circuitry, and a single hybrid device that also helps with display driving duties.
The Iseden-made "Itron" Display Tubes in the QT-8DThe display consists of eight of the eight-segment Itron DG10B tubes made by Japanese manufacturer Iseden, which form numerals using an unusual arrangement of segments that make the numbers look very stylized. The most unique digit rendition is that of the digit zero, which looks kind of like a squashed, half-height zero. A special ninth tube, also made by Iseden (part number SP10), is positioned at the right end of the display panel, contains a "dot" and a minus-sign. The dot lights to indicate an overflow condition, and the minus sign lights to indicate a negative number in the display.
The Main Calculating Circuit Board of the QT-8DThe bottom board in the stack contains the main brains of the machine, with the Rockwell-made LSI chip set, along with the clock generator chip (in a can-type package), and a single Hitachi HD3103(Quintuple P-Channel MOSFET Transistors) IC providing some support circuitry.
A Closeup View of one of the LSI's in the QT-8DThe four LSI's are packaged in ceramic staggered pin flat-pack carriers, with each package having 42 pins. The part numbers on the IC's are AC2261, AU2271, NRD2256, and DC2266. The AU2271 chip (which has 900 transistors) contains the arithmetic unit, consisting of adder, complementer, binary to BCD (Binary-Coded Decimal) correction, and carry logic. The AU chip also hosts the working registers of the calculator. The AC2261 (AC stands for Address Control) chip is the master control, overseeing the interactions of the other chips. The NRD2256 chip (for Numeric Read-in/Display) uses 900 transistors to generate basic timing signals, manages scanning the display, and encodes key presses from the keyboard. Lastly, the DC2266 chip (with 740 transistors) takes care of the logic to keep track of decimal point positioning. These same chips were used in another groundbreaking machine made by Sharp, the EL-8, the first "handheld", rechargeable battery-powered electronic calculator. The chips in the exhibited calculator all have date codes from the mid part of 1970, and the serial number of the machine indicates that it was built in July of 1970.
Keyboard Arrangement on the QT-8D
The QT-8D uses an unusual keyboard layout. Note the combined [X÷] key. Sharp used a novel means to save a key on the keyboard (thus saving the cost, as well as decreasing the real-estate needed for the keyboard) by making the [X÷] key serve both functions. The determination as to which function the user wishes is made by the user selecting either the [+=] key to generate the result for multiplication, and the [-=] key to cause division to occur. For example, to multiply, the user would enter the first number, press the [X÷] key, then enter the multiplier, then press the [+=] key for the answer. To divide, the user would enter the dividend, press the [X÷] key, enter the divisor, and press the [-=] key to calculate the answer.
The keyboard uses magnetic reed switches. This is one of the most expensive means for implementing a keyboard, but is extremely reliable. The magnet/reed switch design results in a keyboard that still works perfectly over 40 years later, with no double entry or inconsistent operation problems. The keyboard assembly connects to the backplane via a neatly bundled wire harness. The key caps have molded-in nomenclature, and are made of high-quality plastic. The keyboard has a great feel to it. Sharp's calculators tended to always have very nice keyboards, and the QT-8D is no exception.
The rear part of the calculator contains the power supply electronics. The machine uses a traditional transformer/rectifier setup, with transistor regulation. The power supply, like all of the other parts of the machine, is of high quality, using large filter capacitors and generous heat-sinking. A special three-prong cord plugs into a socket on the rear panel of the machine.
The QT-8D operates at a decent speed, though it isn't a speed demon compared to some of its less highly-integrated predecessors. The "all-nines" divided by one benchmark takes about 200 milliseconds (0.2 second) to perform. Addition and subtraction complete with no discernible delay. Negative numbers are indicated by a "-" symbol lighting up in a special tube at the right end of the display. The QT-8d accurately detects overflow, lighting the a circular indicator in the same tube as the negative indicator. Despite detecting overflow reliably, the QT-8D does not flag an error when division by zero is attempted, and it will go into a strange state if that occurs. Pressing the [C] key will return to the calculator to normal operation if division by zero occurs.
The Facit 1115 and Addo-X 9354
OEM Versions of the Sharp QT-8D
Images Courtesy Serge Devidts, Calcuseum
Like many other Sharp calculators, the QT-8D's electronics were offered to Sharp's OEM customers for repackaging and sale by other manufacturers. Included among these were long-time Sharp OEM customer Facit, with the Facit 1115, as well as Facit's subsidiary Addo-X, with the model 9354. The OEM versions of the QT-8D were identical inside, with the difference being cabinetry styling, color scheme, and badging.
The Sharp QT-8D also appears to have a place in electronic calculator history within the former Soviet Union. A Soviet copy of the QT-8D, called the Electronika 24-71, was developed by the Soviets reverse-engineering the QT-8D. Leadership of the Soviet Union had become very concerned about the state of electronics within the Union in the 1960's. Many Soviet military systems still utilized vacuum tubes, and had only recently began utilizing solid-state (transistor) devices. The Soviet Union only had small-scale micro-electronic devices in laboratory facilities, which were generally copies of devices developed in Europe and the United States. Large Scale Integration, which had begun to make tentative inroads into consumer electronics as early as 1965 in the US, and by late 1967 in Japan, was only a gleam in the eye of Soviet technical leadership. An aggressive program was put into place in 1969, with a promise made to then Soviet President Leonid Brezhnev that he would have a Soviet-made Large-Scale Integration-based calculator on his desktop at the opening of the 24th Communist Party Congress in 1971. (Note the Soviet machine's model number of 24-71 was based on the 24th Congress, and 1971). The Soviet scientists and engineers were successful in duplicating the LSI chips themselves, but had great difficulty with getting the chips packaged and having them still work afterward. In order to make the presentation to President Brezhnev as promised, a Sharp QT-8D was acquired through indirect means, disassembled, and its electronic guts repackaged into a Soviet-made cabinet, with a Soviet-made display based on the Japanese-design Itron vacuum-fluorescent display tubes used in the QT-8D (with a comma instead of a decimal point), and a Soviet-made keyboard assembly with Cyrillic key caps. Soviet President Brezhnev had no idea that his prized possession was just a repackaged Japanese calculator. It took until 1973 before the chip packaging issues had been solved, and the calculator went into production and sales, with a selling price of 100 Rubles (about US$85 back them).
Original Vendor's Sticker
The exhibited example of the QT-8D appears to have been purchased new in Boston, Massachusetts, at a business machine merchant called "I.C.B.M.". The vendor affixed a tag with their name, address, and telephone number on the side of the calculator case near the power switch. It seems ironic that a business machine company would name themselves ICBM, given that at the time, the threat of a nuclear attack by Soviet Inter-Continental Ballistic Missiles (ICBMs) was a very scary concern throughout the US.