Commodore US*14

This calculator bears a strong resemblance to a another calculator in the museum, the Panasonic 1000. The resemblance seems more than coincidental, and it is, because Commodore had an OEM relationship with Panasonic (Matshushita) that began sometime in the late 1960's to early 1970's. That explains why there are so many similarities in designs of calculators from the two vendors. The Panasonic 1000 is a slightly earlier calculator, being made in 1971, and using less-complex Large Scale Integrated circuits. This US*14 was made in the late part of 1972, and benefits from advances in integrated circuit technology, trimming the IC count, and expanding the functionality of the machine. It appears that the US*14 was introduced sometime in late 1971.

The US*14 is yet another in a fairly long line of "US*" calculators sold by Commodore. The Commodore US*12M is virtually identical to the US*14, but has twelve digits of capacity the fourteen digit capacity of the US*14. Other calculators in the museum in this line include the US*1, US*1M, US*8, and the >US*10. All of the US* line of calculators were designed to be inexpensive, making the price of entry into the world of electronic calculators less of a barrier for consumers to consider for use in the home or small business. The US*14 had an introduction price of $159.95, putting it within the means of mid to upper-income consumers of the time, as well as fitting comfortably within the budget of small businesses. The line was not necessarily released to market in the numerical order of the models. For example, the US*1 came to market in 1973, while the US*8 came out in the mid-part of 1972.

View of the Sperry Gas-Discharge Display in Operation

The US*14 uses a somewhat unusual display technology consisting of a Sperry-made multi-digit (one three-digit, and six two-digit) planar gas-discharge display modules, for a total of fifteen digit positions. The Sperry display modules are similar in design to the multi-digit Burroughs Panaplex display panels, however, rather than having a whole display panel integrated into one module, the Sperry modules come in groups of two or three digits, allowing creation of displays with an arbitrary number of digits by combining combinations of the modules.

The Display Circuit Board

The Sperry display modules are designed so that they can be butted up next to each other, maintaining consistent digit spacing. In the case of the US-14, the display modules are all soldered to a separate circuit board that plugs into an edge connector mounted on the main circuit board. Sperry offered special sockets for these display modules which allowed for easy replacement without de-soldering should a module fail(and, they did over time). These sockets were used in the much more expensive Tektronix 21 and 31 calculator displays, but in the case of the US*14, cost was a major consideration, so the sockets were skipped and the display module leads were soldered directly to the circuit board. The display modules use the standard seven-segment digit rendition, with the digits slanted slightly to the right for a more aesthetic appearance. A triangle shaped decimal point is positioned at the lower right of each digit. Heathkit used Sperry modules with larger digits in its first electronic calculator kit, the HK 2008 and Heathkit 2008-A calculators.

The right-most digit of the display in the US*14 is used as a status indicator. A negative number is indicated by lighting the bottom most horizontal segment of the status digit. The upper three segments of the status digit light up when the memory register has been accessed using the "M+" or "M-" keys, and the status digit shows "E" when the calculator has detected an overflow or error condition.

US*14 Keyboard Detail

The US*14 provides the standard four math functions, along with a percent calculation. It has a single memory accumulator, with [M+] and [M-] keys to add/subtract the content of the display from the memory accumulator. The [RM] key recalls the content of the memory register to the display, and the [CM] key clears the memory register. The memory status indication mentioned above operates differently than most. Most memory status indicators are lit anytime the memory register contains a non-zero value. Not so with the US*14. A depression of the [M+] or [M-] keys lights the indicator, even if the memory register contains zero as a result of the operation. Pressing the [CM] key clears the memory register, and turns off the memory status indicator. It appears that rather than check the memory register for zero content, the memory status indicator of the US*14 is simply a latch that is set by the operation of the [M+] or [M-] keys, and cleared by the [CM] key. A push-on/push-off [Σ] key allows results of multiplications or divisions to automatically be accumulated in the memory register. It is important to remember to clear the memory register before engaging the summation function, as any as any existing number in the memory register will be included in the total.

The US*14 operates in fixed or floating decimal as selected by a thumb-wheel selector at the left of the keyboard panel. Fixed decimal point operation can be selected by dialing in zero through seven digits behind the decimal point on a thumb-wheel switch located to the left of the keyboard. Two positions of the thumb-wheel switch contain an "F" setting, which selects floating decimal operation. The US*14 provides leading zero suppression, and when in floating decimal mode, it always right-justifies the display so that any trailing zeroes are eliminated. A round-off mode switch is located next to the decimal mode thumb-wheel. This switch provides three different settings for force up, 5-up/4-down rounding, and force down (truncate) modes. The percentage function operates as an [=] key, terminating multiplication and division operations, for example to find 25% of 16, one would enter "16 [X] 25 [%]" and the display would show the answer of 4. The [K] key, a push-on/push off switch, provides for constant calculations, and operates only in multiplication and division operations.

Error detection on the US*14 is unusual, with error and overflow being reliably detected and indicated, however, the calculator does not logically lock the keyboard (e.g., ignoring key presses) when such an error/overflow condition exists. The right-most digit position in the display shows an "E" when an error or overflow occurs, but the machine will merrily continue calculations. Pressing the [C] (Clear All) key extinguishes the error indication. The "E" indication will also occur if the memory register overflows. The [CI] key clears the display for use in correcting input errors. The [EX] key exchanges the order of operands for multiplication and division operations, and must be pressed before the [=] key in order to take effect.

Inside the US*14

The US*14 is one of the few low-cost calculators that was manufactured in North America at the time. By the early 1970's, the Japanese had pretty much won the market for low-cost electronic calculators, due to a number of factors, including heavy government backing for development of advanced electronic technology; a highly-educated and driven workforce, and significantly lower wages than in America and Europe. Most of the calculator manufacturers that still manufactured their calculators in North America were focusing on high-end scientific and engineering calculators. Manufacturers such as Hewlett Packard, Computer Design Corporation (Compucorp), and Wang Laboratories made their calculators in the US, but these were machines that cost thousands of dollars and offered much more complex functionality than the Japanese calculators. In the high-end calculator market, price sensitivity wasn't as critical as it was in the low-end market, which allowed the high-end North American manufacturers to market their machines successfully against the flood of low-cost calculators from Japan. The US*14, was conceived as a machine to fill a gap in the higher-end of the basic home office/small business electronic calculator market. The calculator bucked the trend of being made in Japan, and still managed to sell for an introduction price of $

In spite of its low-cost, the US*14 is a nicely built machine. The calculator's guts are all contained on one high quality fiberglass circuit board. The circuit board is single-sided, with wire-jumpers making connections on the component side. Very legible silk-screened component identifications are printed on the circuit board, making it much easier for repair technicians to identify the correct part when doing service. The weakest part of the design of the US* line of calculators is the keyboards. The keyboards use spring-type contacts in sealed modules. For whatever reason, gunk tends to get inside the switch module through the key stem, gumming up the works and causing inconsistent keyboard action, including double and triple entries, as well as missed entries. Every Commodore US*-series machine in the museum has in its possession has had keyboard problems which required remedial efforts (generally a small squirt of De-Ox-It down the key stem of each key) to fix.

Two Versions of the Main Circuit Board in the US*14
Note Added Transformer and Aluminum Heat-sink for Power Supply Regulation Transistors on Second Main Board

The main logic of the US*14 uses an unusual three-chip Large Scale Integration (LSI) chip set that was designed by relatively obscure MOS IC design firm known as Integrated Systems Technology(IST). IST would design (not manufacture) full-custom MOS integrated circuits for anyone who had a logic design that required implementation in MOS/LSI integrated circuit form. IST did not manufacture IC's -- they only designed them, turning the logic into a set of huge rubylith masks that would be used by the customer's IC fabrication contractor to create the ICs. It was up to the customer to find an IC fabrication house to actually produce the chips from the masks produced by IST.

The story of IST is rather interesting, as it is a keen illustration of how the threads of technological history have unique ways of entangling. IST was founded in 1968, by a number of folks who had left Philco-Ford Microelectronics to start a custom MOS integrated circuit design firm. These highly skilled engineers left Philco-Ford, because in 1966, Philco-Ford had purchased an early pioneer in large-scale MOS Integrated Circuit design and manufacturing, a company called General Micro-electronics(GM-e). The folks that left Philco-Ford had been a part of GM-e, and did not want to become gobbled up in the huge bureaucracy of Philco Ford. GM-e had been formed by a group of engineers that left Fairchild Semiconductor to pursue their (later proved to be correct) belief that Metal-Oxide Semiconductor (MOS) integrated circuitry was the the future of digital integrated circuits. These folks had tried to get Fairchild interested in a large scale (forgive the pun) effort to take MOS IC technology beyond the modest investment the company had made in the technology, but were not able to gain traction with Fairchild's management, so they opted to go their own way. GM-e holds a significant place in electronics history, as the company produced some early high-density MOS devices that were used by the United Status military and national security sectors to significantly advance the state-of-the-art in cryptographic systems, secure communications systems, weapons guidance, and navigation systems early in the company's history. Not much later, GM-e would become famous for developing the first implementation of the digital logic of an entire electronic calculator in a Large-Scale Integration (LSI) MOS IC chip set, which at the time was something that was considered to be in the realm of science fiction. The development of the calculator and the chip set that made it possible was performed by GM-e under contract to Victor Comptometer. Victor Comptometer marketed, sold, and (with the help of GM-e) serviced the world's first production integrated circuit-based electronic calculator, the Victor 3900 (the referenced link points to an essay that outlines the history of the development of the Victor 3900), introduced in October of 1965 - nearly two full years before small-scale bipolar IC technology appeared in any other electronic calculator. The GM-e chip set that contained all of the logic of the Victor 3900 consisted of 23 unique integrated circuit devices, each of which contained a staggering (at the time) 250 logic elements per chip. A total of 29 of the chips were used (with one chip used six times) in the Victor 3900. The twist of the story is that the IST-designed chips used in the Commodore US*14, far removed from the Victor 3900, have their roots in folks that developed Victor's history-making calculator.

The chip set used in the US*14 consists of two 16-pin, and one 18-pin package devices. The part numbers are 7061, 7062, and 7063. The packages are ceramic dual-in-line, with gold plated pins. The chips are plugged into sockets similar to Molex strips.

Closeup of the "IST"-made LSI IC Chip set

Along with the three LSI's, there is a sprinkling of six TTL 7400-series devices that provide various glue functions. Among these, a 7442 (a fairly rare IC today) is used to convert the Binary Coded Decimal (BCD) output of the calculator chip set to seven-segment form for the display. Two more IC's, made by Sperry (DD-700), are specialized chips for driving the Sperry gas-discharge display modules. Date codes on all of the chips range from mid-1972 to late-1972. Along with the IC's, the machine has a surprising number of discrete transistors. Some are used in the power supply section, others appear to be related to clock generation, and lastly, some are used in the display driving circuitry.

The US*14 is about average in speed for a 14-digit calculator, with the all-nines divided by 1 problem taking about 1/3 second. During calculation, the display is blanked.