Dr An Wang was a physics graduate of Harvard University. He remained at Harvard to perform research in the Harvard electronic computing laboratories and this placed him at the forefront of developments in electronic calculation in the late 1940s. He was closely involved with the academic work and innovations that established magnetic cores as viable electronic memory units and he was awarded early patents on this technology. It is no surprise that Wang calculators of the 1960s all used core memory architecture.
Dr Wang had a strong entrepreneurial drive and in the early 1950s he established his own electronics research and development company which produced control systems based on electronic logic modules that he had developed. Wang Laboratories products from this era included numerical control systems for machine tools and a data processing system to automatically justify text before it was submitted to a typesetting machine. These products were successful and Dr Wang was able to use some of the profits to pursue his interest in developing an electronic calculator.
Dr Wang's desire to produce electronic calculators was perhaps driven by another of his significant innovations. This innovation responded to the fact that while addition and subtraction were relatively easily done with electronic logic, multiplication and division were much more complicated and extraction of roots was even more complicated. Other engineers had designed circuits to directly perform some of these more complex operations but Dr Wang took a different approach. He invented circuits to convert numbers stored in his calculators to their logarithms, and the logarithms back to their conventional base ten representation. Once the logarithms were available, multiplication and division were reduced to simple addition and subtraction of logarithms. Squares and square roots could be produced by simple doubling or halfing the logarith and further exponents or roots could be produced by multiplying or dividing the logarithms. The logarithm circuits provided Dr Wang with a unique means to build calculators with powerful multiplication, division, logarithmic and exponentiation capabilities that were until that time only available with full scale computers.
The LOCI machines were the first calculators based on Dr Wang's logarithm circuits. They came to market in 1964, with the name being derived from "Logarithmic Computing Instrument". They were complex to look at and had a peculiar operating method that was based on the underlying engineering and paid little heed to user interface design. Nevertheless, scientists and engineers readily understood the logarithmic approach and appreciated the calculating power available from a machine that operated on these principles.
The LOCI-1 was a short-lived first version, rapidly replaced by the LOCI-2.
The LOCI-2 has two further significantly differences from all other first generation machines. Firstly, it is programmable, with loop and branch abilities. Secondly, it has an interface connection that allowed data to be input, processed by the calculator and then output. These features made the LOCI more than just a desktop calculators, it could also be used in process control and data acquisition/analysis. No doubt Wang Laboratories' past experience in these fields suggested the use of LOCIs in this way, an application that was overlooked by almost all other calculator companies. Wang Laboratories had a department that assisted in the design of LOCI based data acquisition and processing systems and one interesting system that resulted was an automatic spacesuit testing system that was used by NASA throughout the 1960s.
For further information about the LOCI machines, see the Old Calculator Museum's excellent site and LOCI exhibit.
Wang Laboratories built on the success of the LOCI machines, replacing them in 1966 with the 300 series. The 300 series machines appear to be an attempt to provide the power of the LOCI architecture in a smaller and more usable package. The packaging was certainly more appealing with a compact desktop keyboard and display unit but the user interface remained rather unfriendly with terse keyboard symbols and puzzling features like two sets of "+" and "-" keys.
The 300 series machines were more powerful than the LOCIs and were still readily understood and appreciated by scientific and technical markets. They rapidly became very successful and expanded into a family of successful machines, in the process making Wang Laboratories the dominant manufacturer of high-end calculators in the late 1960s.
The 300-series were less popular in business and non-technical markets because of a side effect of the logarithmic approach. Most numbers do not have exact finite logarithmic representations, so logarithmic operations generally involve approximations, to 14 digits of precision in the case of the 300 series. Scientific, engineering and technical users understood these approximations and recognised that the answers were quite accurate enough for building bridges, surveying land or launching rockets.
Business and non-technical users were disturbed however, when "100 divided by 2" produced the answer "49.999999999999" and Wang responded with the 200 series, which was essentially the 300 series with additional rounding off logic.
While the 300 series machines were very successful, they remained quite expensive. Where multiple users required calculation facilities the 300 series allowed tee-connectors to be used to connect multiple keyboards to one electronics package. When tee-connectors were used only one keyboard could be active at a time, all others were blocked. The active keyboard had to be switched off before another could be switched on.
The "SE" machines allowed customers to achieve additional value for money by plugging up to four keyboards into one electronics package. The SE-series electronics packages provided true timesharing and separate numeric storage for all four keyboard ports, allowing all to be used simultaneously and independently.
It was still possible to use one or more tee-connectors on any or all ports on an SE package, allowing six, eight or even more keyboards to be connected to one electronics package. In this situation some keyboards would block others, while others again would be independent - a recipe for departmental confusion but an indication of the value placed on desktop calculators in the 1960s, and their high price.
The 700 series was the successor to the 300 series and was a major step forward in capability, underlying architecture and electronic construction. It was the pinnacle of Wang's calculator achievements but it also marked the point from which Wang's dominance of the high-end calculator business began to decline.
The capability of the 700 series was greatly increased by its having 120 storage locations, inbuilt higher mathematical functions and most importantly a programming facility with branch and loop constructions. The 700 series machines were actually less like calculators and more like small computers, this was because the hardware of the 700 series had started as a project to build a Wang minicomputer. However, before the computer project was complete Hewlett Packard released the HP 9100 and Dr Wang saw a major threat to his very successful and profitable 300 series calculators. Seeking an immediate response to Hewlett-Packard, Dr Wang realised that the minicomputer project could be redesigned to become an advanced programmable calculator.
The 700 series was introduced in mid 1970. The 300 series' discrete transistor logic was superseded by integrated circuits and the hard-wired logarithmic calculation electronics were was superseded by a basic electronic logic engine plus a stored microprogram that drove this engine to express the higher level machine that was seen by the user. Magnetic core memory was retained, perhaps reflecting Wang's long association with this technology. The 700 series also restored the data input/output connector that was available on the original LOCI machines and this allowed Wang to cater to the data acquisition and processing market once again. For example, a 700 series machine was the central element of a lightning surveillance and warning system for the Apollo launch complex at Cape Canaveral.
The 600 series was a simplified and cheaper range based on the 700 series. The machines were physically similar to the 700 but had operating features similar to the 300 series, perhaps to assist 300 users to update the the 600 series.
The 400 series was the final family of Wang calculators, introduced in mid 1972. By this time LSI calculator-in-a-chip machines were becoming common and while the 400 series machines were very powerful fully-programmable calculators, their construction using many discrete chips was outdated and was becoming uneconomic to produce.
The 400 series machines were the first Wang calculators that did not use core memory, ending a design tradition that stretches back to the LOCI machines, and then back further to Dr Wang's postgraduate research in the 1940s.
The 400 machines did nevertheless maintain the unique operating style of previous generation machines, with Wang's typical user interface and the associated disregard for steep learning curves. There were still multiple buttons with terse and/or cryptic labels, and a variety of locking buttons that modified or completely changed the function of the cryptically labelled buttons. Once mastered, the system was actually very powerful and very economical of keystrokes but the learning curve could be long and steep.
The combination of unconventional operation and high production cost accelerated the decline of Wang's high-end calculator business and Dr Wang elected to leave the calculator field rather than try to compete his way back in an increasingly crowded and competitive market.
The 400 series machines are interesting not just as the final stage of Wang's calculator development but also because they are constructed from many relatively simple integrated circuits. Their design can therefore be fully analysed and the microcode can be extracted, allowing the engineering of these final Wang machines to be fully understood and appreciated.
In the early 1970s Wang still had a large and profitable calculator business but its position as the dominant manufacturer of high end calculators had been lost. Hewlett-Packard's HP 9100 series had started the process and the subsequent HP 35 handheld calculator established HP as the new leader in high end calculators. At the same time, low-end integrated circuit calculators were becoming ever cheaper and more powerful. Wang was pressured from both sides and Dr Wang responded by moving the company's skills and focus, leaving the calculator market. The 400 series were the last Wang calculators.
Dr Wang shifted the company's focus in a manner that brought together its distant past history and its current expertise to produce a new product for a new market. Revisiting the text-processing problems that he was involved with in the 1950s, Dr Wang re-used the small-computer-like hardware of his programmable calculators by developing new microcode so that the hardware no longer acted as a calculator, but rather as a word processor. Wang was one of the ground-breaking developers of business word processors and then was an important early developer of networked small business computers. Wang word processors and minicomputers became very successful in the late 1970s and through the 1980s, taking the company to a greater level of success than had been achieved in the calculator business.
The Small Wang Museum by Jan van de Veen is an excellent site that presents this next stage of Wang's development.