| +Home | Museum | Wanted | Specs | Previous | Next |
Casio 121-A (AS-A) Desktop Calculator
Updated 7/29/2024
The Casio 121-A, also known as the Casio AS-A (more on the AS-A model name later), is distinctive because of its unique horizontal form-factor, placing the keyboard beside the display rather than below it. The 121-A/AS-A was the first calculator on the market to use this form-factor. The rationale behind the design was that it was easier for the operator to see both the display, and their hand on the keyboard than the traditional form-factor. Whether or not this translated to fewer errors in use of the machine is debatable, though Casio did introduce three later versions using this form-factor (121-B/AS-B, AS-C, and AS-L) that improved upon function and features as technology advanced.
A Casio AS-A badged 121-A calculator
Image Courtesy Museo Nazionale degli Strumenti per il Calcolo (National Museum of Computing Instruments), Pisa, Italy.
By Permission of Dr. Christina Paoli
The earlier mention of the AS-A model number is included here because Casio marketed the same machine under two different model names. It appears that the machine was introduced and sold its native market (Asia/Europe) as the model AS-A, but when exported outside of Casio's native market, the machine assumed the 121-A model name. Inside and out, the machines were the same. The only difference was the badge on the front panel of the calculator designating "121-A" or "AS-A", and the model/serial-number tag located on the bottom cover of the calculator. It is not entirely clear why Casio did this, but this differential in model names continued with an improved version of the 121-A/AS-A which was marketed by Casio as the Casio 121-B/AS-B. Later, there were machines using both the 121-x and AS-x nomenclature, but after the 121-B/AS-B, the 121-x and AS-x model lines diverged, with 121-x models differing from AS-x models.
Casio 121-A Model/Serial Number Tag
The Casio 121-A/AS-A is implemented with mostly second-generation small and medium-scale Japanese MOS (Metal-Oxide Semiconductor) integrated circuit technology made by NEC and Hitachi, along with a relatively small amount of discrete components compared to earlier Casio electronic calculators. The reduction in the number of discrete components, which were mostly inexpensive diodes and resistors used as low-cost logic gates where using an IC wasn't as cost-effective, as well as in the Nixie tube display drive circuitry was due to higher levels of integration within the IC's. While the 121-A/AS-A utilized improved IC technology compared to earlier machines by Casio, some sacrifices were made in the design in order to reduce the physical size of the calculator. When announced to the Japanese market in June of 1969, the 121-A/AS-A were touted as the smallest, lightest electronic calculators on the market.
The Casio 121
Photo Courtesy of
Mr. Serge Devidts, Calcuseum
The "A" in both 121-A and AS-A are hints that there might have been a "Model 121" or "Model AS" that preceded the 121-A/AS-A. In fact, that is true -- there was a Casio model 121, but it only marketed as the 121, and never as the model AS, further confusing the situation with regard to the dual model numbers of the 121-A/AS-A and 121-B/AS-B. The Model 121 was the beginning of a long line of Casio 121-series calculators. The original machine in the series, the Model 121 (See the Old Calculator Museum's Commodore 1121 exhibit for an OEM version of the Casio 121 sold by Commodore), was a larger, had a more conventional (e.g., display located above keyboard) form-factor and was a more fully-featured calculator that utilized first-generation Japanese MOS Integrated Circuits combined with a large number of discrete components. The 121, and it's somewhat downgraded stable-mate, the Casio 120, were introduced in 1967, and were popular sellers. The 121, by virtue of its less advanced technology, was quite a bit larger, heavier, and significantly more expensive than the new 121-A/AS-A, and utilized the more conventional keyboard-below-display form-factor.
At the time that the 121-A/AS-A was being designed, desktop electronic calculators made by Casio's competitors were significantly more expensive, and generally not as portable as this calculator. Even though the 121-A/AS-A is still a mains-powered desktop device, the calculator is small and light enough (only 6 1/2 pounds) to easily carry around from place to place where calculations are required -- a feature which made the machine very popular. The 121-A/AS-A was designed to be able to fit inside a traditional briefcase, though, one would have to exercise care carrying the calculator in a briefcase without some kind of padding to prevent it from slamming around inside. To keep the price and physical size of the machine down, Casio made some compromises with the functionality and features of the calculator. These compromises included the lack of negative number handling, as well as the inability to input fractional numbers due to there being no decimal point key! Also, there was limited ability to calculate with fractional numbers. These limitations, though somewhat profound compared to other calculators on the market at the time, were not an impediment to sales, as the Casio 121A/AS-A and those marketed by OEM customers of Casio were hot sellers, setting a new low-price benchmark for an electronic calculator. The compromises in the calculator were such that the machine was still extremely useful for basic mathematics, especially for things like balancing statements and invoicing, were an assumed decimal point, which could be set by the sliding plastic digit grouping indicators as a visual reference, was not a problem.
Casio 121-A Circuit Board Showing "Model AS-A" Silk-screen
To further the confusion concerning the different model names of this calculator, the circuit boards inside the exhibited Casio 121-A calculator actually have "MODEL AS-A" silk-screened on them, even though the calculator has "Casio 121-A" clearly stated on the badge on the front panel, and has the model name of "121-A" on the model/serial number tag.
An Advertisement for the Remington version of the Casio 121-A marketed as the Remington Rand AS-A
Remington Rand in Australia was the first company outside of Japan to import Casio-made calculators under an OEM (Original Equipment Manufacturer) agreement. Remington Rand initially imported Casio-made electronic electronic calculators beginning with the Casio 101 in September of 1966. Remington Rand marketed the Casio 121-A in Australia as the "Remington/Casio AS-A."
The Commodore version of the Casio 121-A/AS-A known as the Commodore 512
Commodore Business Machines was a Canadian firm that became known in North America as an early importer of low-cost Japanese typewriters, and then later, electronics, distributing and servicing the business machines in the North American market. Commodore later became essentially a household word as a result of the development of a wildly-successful low-cost home computer (long before the IBM-PC) known as the Commodore 64, introduced in 1982. The Commodore 64 is computer that readers of a certain age range reading this exhibit may likely have used as their first home computer. The Commodore 64 was the most popular (by number sold) personal computer of its era.
In the latter part of 1966, Commodore trial-marketed Casio's second export-ready electronic calculator, the Casio 101E, as the Commodore 500E, to test the waters to see if it was viable to import Japanese-made electronic calculators and sell them in North America. The Casio-made calculators were significantly less-expensive than similar American-made calculators, even when the cost to import them was considered, given that large-scale OEM customers of Casio would receive a substantial discount on orders of large quantities of calculators. The results of the trial marketing were sufficient that in the first half of 1967, Commodore engaged in an official large-scale OEM contract with Casio. Casio designed and manufactured the calculators, and Commodore simply placed their own model/serial number tag on the imported calculators, packaged them with a user manual that was (poorly) translated from Japanese, substituting "Commodore 512" for "Casio AS-A", put it all in a box marked Commodore, and sold and supported the calculators under the Commodore brand within North America. The Commodore 512 is identical in all aspects to the Casio 121-A/AS-A, but marketed, sold and supported entirely by Commodore in North America. The only hint on the Commodore 512 was an import was the "MADE IN JAPAN" notation on the serial number tag.
Near the end of Casio's production line for Casio AS-A calculators, Circa 1970
The exhibited Casio 121-A appears to have been built in early 1971, based on the '1A' date codes which are the latest codes on any of the IC's in the machine. The 1A code the ICs with this date code were made in the first month (A) of 1971. The 121-A/AS-A's design was finalized in mid-1969, with trial production beginning at that time, with full production beginning in the summer of 1969. Production of the 121-A/AS-A likely continued into early 1972, with sales extending into 1973, with the price for the calculator discounted over time to account for the tremendous competition in the electronic calculator marketplace in the early 1970s.
Internal view of Casio 121-A
The Casio 121-A/AS-A has two main circuit boards interconnected by 38 individual wire jumpers. The circuit boards are made of laminated Phenolic paper, with silk-screened white annotations. There is no solder-mask present on the boards. The circuit boards contain components and traces on one side, and traces on the reverse side. There are some jumper wires on the component side of the circuit boards to make connections between components that were unable to be made via copper etch due to space limitations. The circuit boards have feed-through holes that are plated through to provide connectivity between traces on the top and bottom sides of the boards.
The "INDICATOR & MAIN" (top) Ciruit Board
The top circuit board, designated "INDICATOR & MAIN" contains the Nixie display tubes and their decoder-driver circuitry; the display multiplexing shift register; the 50KHz master clock-generation and divider circuitry; and the keyboard conditioning and encoding circuitry. The bottom circuit board, designated "SUB & REGISTER", contains the calculating logic; with the adder, carry logic, the three working registers of the calculator, and state sequencing logic.
Discrete Transistor Nixie Tube Drivers in Casio 121-A/AS-A
The display is made up of 12 individual Hitachi CD-71 Nixie tubes put together in a metal frame that provides mechanical stability and shock isolation for the rather delicate tubes. At the bottom of the metal framework is a circuit board that provides the connections for the Nixie tubes. The Nixie circuit board connects to the INDICATOR & MAIN circuit board with jumper wires. The Nixie tube support frame mounts to the INDICATOR & MAIN circuit board with screws. The CD-71 Nixie tubes contain the digits zero through nine, and have a right-hand decimal point. The digits are approximately 1/2" in height (13mm), and the tubes operate at a nominal voltage of 200VDC.
The "SUB & REGISTER" (bottom) circuit board in the AS-A/121-A board set
All together, the logic of the calculator is made up of only 35 Small- and Medium-Scale Integration IC's, mostly from Hitachi's HD3100-Series of PMOS integrated circuit devices. The small-scale IC devices consist of logic gates and flip-flops used for control and sequencing functions. The medium-scale integrated circuits in the 121-A consist of a number of NEC µPD108 shift registers (used as the working registers of the machine), a Hitachi HD3112 serial binary/BCD adder/subtracter (performing arithmetic operations), and a NEC µPD116 BCD to 1-of-10 decoder for converting the 4-bit binary (2'-4-2-1) internal representation of digits to the 1-of-10 signals needed to drive the Nixie tubes.
The 2'-4-2-1 internal representation of decimal digits used in the AS-A/121-A is rather unusual in the realm of electronic calculators. Most calculators of the time represented individual decimal digits in so-called "Binary-Coded Decimal", or "BCD", using an "8-4-2-1" ranking for the four binary bits which represented the digits 0 through 9 with their normal four-bit base-2(binary) encoding (e.g., 0000(0), 0001(1), 0010(2), 0011(3), 0100(4), 0101(5), 0110(6), 0111(7), 1000(8), 1001(9), with the remaining bit combinations (1010, 1011, 1100, 1101, 1111 being invalid cases, though in some calculators these invalid codes were used as markers for the end of numbers, or even as a decimal point place holder). Casio chose the unusual 2'(2 prime)-4-2-1 encoding to help reduce the amount of logic required to implement the calculator. 2'-4-2-1, also known as the Aiken code (the last name of an American pioneer in computer design Howard Aiken(3/8/1900-3/14/1973), is a representation that encodes the decimal numbers as so: 0000(0), 0001(1), 0010(2), 0011(3), 0100(4), 1001(5), 1100(6), 1101(7), 1110(8), 1111(9). This leaves the codings 0101, 0110, 0111, 1010, and 1011 as invalid. The main reason this encoding makes calculating logic less complex is that no correction is required to perform the addition of any two decimal digits.
Using BCD digit encoding, as an example, adding 9(1001) and 3(0011) together results in 1100, which in BCD encoding, is an invalid coding. The condition where a sum of two digits exceeds 9 must be detected by the logic, and a correction factor of six added to the result in a second addition cycle in order to properly generate a carry, as well as represent the final digit of the sum correctly. The logic to detect the invalid sums and correct them by performing an additional addition cycle is fairly complex, and adds to the time it takes to perform a calculation. In this case, after the addition results in 1100, the correction factor of six must be added, (e.g., 1100 + 0110), resulting in a total of 0010 with a carry out of the most significant bit. The carry represents the quantity of 10, with the sum bits of 0010 coding as 2, indicating a final sum of 12. Performing the same binary addition using Aiken coding, (e.g., 1111(9) + 0011(3)) results in an overflow (carry), and 0010, which, in Aiken code, represents two, giving a total of 12, without having to perform any correction. This makes the addition circuitry, which represents the least-common-denominator function within a calculator, significantly less complex to implement, reducing complexity and thus cost.
Technically, it doesn't matter what internal representation for the decimal digits is used in a calculator that processes individual decimal digits as a unit, as long as there the addition circuitry properly generates proper decimal sums based on the coding used (including a carry if necessary), a way of encoding the digits on the keyboard into the proper internal code, and some form of translation circuit that takes the internal binary representation each digit and translates it to the 1-of-10 signal that lights up the proper digit in the Nixie time for each of the codes. The benefit of the Aiken code is that the addition circuitry is implemented using a simple four-bit binary adder. This is in contrast to more complicated correction processes required for other forms of encoding a decimal digit into a four-bit binary number, performing addition on it, and converting it back to a form displayable on the calculator's display system.
Casio 121-A/AS-A Keyboard Layout
The 121-A provides the four basic math functions. The [=+] and [=-] (a rather unusual designation, usually these functions are labeled [+=] and [-=]) keys provide addition and subtraction functions, and operate adding machine-style. Multiply and divide functions are designated by the familiar keys, [X] and [÷]. The [C] key (located where one might seek a decimal point key) clears the machine and readies it for a new calculation. Multiplication and division operate as expected with the [=+] key pressed to calculate the result.
Counting Display on Multiplication Overflow
The 121-A/AS-A does not deal well with multiplication calculations that exceed the capacity of the machine. Multiplying 999999999999 X 999999999999 (the worst-case situation) results in a madly counting display, with multiple digits in some Nixie tubes lit at once. Pressing the [C] key recovers from this strange state. This is likely due to compromises made in the logic design that does not properly detect overflow conditions.
Upper Half of the 121-A Cabinet - Power Wiring, Keyboard
As mentioned, the 121-A/AS-A does not directly support negative numbers. Negative results are indicated by displaying the tens complement of the negative number. For example, -1 is displayed as 999999999999. Users performing calculations must be wary of this, as there is no indication given when such "underflow" occurs. Chain multiplication and division must be performed with intermediate presses of the [=+] key for each division. Even when that is done, chain division doesn't work very well with fractional numbers, as each time the divide key is pressed, the decimal point is reset to the right-most digit position, losing track of where the decimal point was. For example, performing 25 ÷ 2, ÷ by 2, results in 12.5 as expected for the first division, but when the second divide is carried out, the display changes to 625000000000., a grossly incorrect answer (at least in terms of magnitude). The multiply function key also resets the decimal point location back to the right-most position, making chain calculations on the 121-A/AS-A with fractional numbers an exercise in futility.
Division by zero results in the machine getting also quite confused. All of the decimal points light up dimly, and the machine acts strangely when keys are pressed when it is in this state. Clearing the machine with the [C] key remedies this condition. Divisions which result in quotients which have fractional portions don't always result in answers that make sense. For example, the simple division of 1 by 3 results in 333333333333. -- not quite the answer expected, likely due to the fact that the left most Nixie tube has its decimal point on the right, meaning there is no way top display a leading decimal point unless special logic were added to detect such a condition and shift the result to the right, adding a leading "0." in front of the quotient, which would yield "0.33333333333". Such logic would add significant complexity to the machine, which was simply omitted as a compromise in the quest for minimizing the amount of circuitry in the calculator.
Another compromise decision for this calculator is that it does not have any detection of input overflow;
entering numbers in excess of the twelve digits of capacity of the machine result in the high-order
numbers just shifting off the left end of the display with no warning or
other indication of a problem. As with many early electronic calculators,
the 121-A doesn't properly deal with large dividends properly. The machine
gives an incorrect result (999999999990) if the "all-nines divided by 1"
problem is fed to it. This is because one digit of the working register is
actually used as a counter for division operations, meaning that only
eleven digits can be operated on in a division operation. This tactic saves
on additional logic at the expense of being able to exercise the full
capacity of the machine for division operations. The closest speed
benchmark I can get is eleven nines divided by one, which takes about
300 milliseconds to complete.