+Home     Museum     Wanted     Specs     Previous     Next  

Sumlock Comptometer/Bell Punch Anita C/VIII Electronic Calculator

Updated 5/6/2002

This calculator represents a real piece of history. The Anita C/VIII (also known as the Mark 8) is generally recognized as the first mass-produced commercially-available all-electronic calculator. That distinction alone makes the machine interesting, but the really fascinating part about this machine is that the designers managed to come up with a complete all-electronic four-function calculator made from tubes. Yes, you read right -- tubes. Sumlock Comptometer Ltd., a company that made mechanical calculating machines, in collaboration with another London-based company, Bell Punch Company, built and marketed a machine that would herald the beginning of a new age...an age where the use of slide rules and mechanical calculators would fade into the mists of memory. From the introduction of the first Anita, and the subsequent proliferation of other electronic calculators, electronic calculating machines quietly and forever changed the way people dealt with the drudgery of mathematics. By the way, it appears that Sumlock realized that their creation was definitely a revolution in the way mathematics were dealt with, as the name Anita is an acronym for A New Inspiration To Arithmetic.

A Production Anita C/VII (Serial #510) - Predecessor to the Mark 8, See Text
Photo Courtesy Frank Eggebrecht/Friedrich Diestelkamp

To be most accurate, the Anita Mark 8 was not really the first all-electronic desktop calculator. There was a predecessor to the machine, the C/VII (Mark 7), that was truly the first electronic calculator to be commercially available. However, less than a thousand of the Mark 7 were produced, and the machine was superceded within a short time by the improved Mark 8. The Mark 7 used three Dekatron tubes, one as a master clock, and two as counting elements. There were some problems with using the Dekatron tubes as counting elements in the machine, which led to machines which would give incorrect results over time. Dekatron tubes prefer to be continually operating. The start/stop type of operation that occurs when the tube is used as a counter could lead to a condition called "poisoning", where the electrical characteristics of the tube would degrade. This condition would end up causing the calculator to make errors. Because of this situation, design changes were made that led to the Mark 8, which utilized ring counters made from individual Thyratron tubes as the main counting elements, leaving only one Dekatron which was used as a master timing generator for the machine. This improved the reliability greatly, and made the Mark 8 a true mass-market machine.

The back panel of the Anita showing the Sumlock Anita logo

The Anita Mark 8 is a 12-digit, four function (add, subtract, multiply, and divide) electronic calculator. The Mark 8 was originally introduced in September, 1961, clearly beating the American company Friden, and their fully-transistorized Friden 130 (1963) to market by over a year, and Japanese manufacturers Casio and Sharp (then known as Hayakawa Electric) by an even larger margin. In the late 1950's, prior to Anita, there were desk-sized (or larger) electrical (as opposed to electronic) calculators that used relays as logic and switching elements, such as the Casio 14-A. Relay-based calculators were a dramatic improvement over electromechanical (motor-drive mechanical) calculators, but were physically much larger than their electromechanical counterparts. Relay-based calculators were not truly electronic machines. Relays rely on mechanical motion to close switch contacts, and thus aren't considered true electronic devices. The thyratron tube devices used in the Anita have absolutely no moving parts, and are therefore pure electronic devices.

Internal View of Anita C/VIII

Sumlock Anita Ltd. (the name of the joint partnership formed by Sumlock and Bell Punch Company to market the Anita) managed to beat Friden and other manufacturers to market with an all-electronic calculator by using tried and true cold-cathode tube technology. By the mid-1950's, tube technology had fully matured, with miniaturization making it possible to cram enough tubes into a desktop-sized (similar in size to electro-mechanical calculators of the day) package to make a practical desktop electronic calculating machine. The notion of building an all-electronic desk calculator was an idea that Bell Punch Company had decided to persue in 1956, and engineer/scientist Norbert Kitz was assigned to the project. Kitz had plenty of experience working with tube-type digital logic, as he was heavily involved in the development Pilot ACE, an early electronic computer built at the National Physics Laboratory in Britain in the late 1940's. Back in '56, when the development of the Anita desktop electronic calculator began, transistor technology was still rather new, and was viewed as not mature enough to be a viable choice as a component for use in such a device. The only choice at the time was tube technology. However, even though tube technology was quite mature, tubes had numerous difficulties (high voltage requirements, comparatively limited lifetime, relative large size, and delicate glass construction) that made the design of a tube-based electronic desktop calculator quite a challenge. Challenges aside, development of the calculator proceeded, and in 1960, a truly all-electronic prototype calculator was running. In 1961, the Mark 8 was on the market, and selling like hotcakes, with production quantities in stated as being in the "10,000 units per-annum" in a paper on the use of cold-cathode tubes for computing applications presented by Kitz at a technical symposium in March, 1964.

A close-up view of the sub-miniature Thyratron tubes and Selenium rectifiers (the black cubes) used in the Anita C/VIII

In many ways, the Anita Mark 8 is essentially an electronic implementation of earlier electro-mechanical calculators. The calculator operates with an electronic implementation of base-10, using a '1 of 10' arrangement rather than Binary Coded Decimal or other binary-based notations that later electronic calculators use. In fact, in his symposium paper, Kitz made a forthright reference to his preference for digital circuity to operate in base-10(decimal) rather than base-2(binary) with the following statement: "In many ways, Anita is a small electronic computer working in decimal notation (no binary stunts here!) under the command of the operator."

The majority of the function of the Anita Mark 8 is created by a collection of 177 sub-miniature tube devices called Thyratrons. A Thyratron is not a vacuum tube like those used in old TV's and radios. Thyratrons do not have a vacuum inside them, but rather are filled with a gas. Neon gas, along with small amounts of other gases fill the glass envelope of a Thyratron tube. Thyratrons are "cold cathode" devices -- there is no heater element in them like vacuum tubes have. A thyratron is a modified version of the small neon glow tubes typically used in as power-on indicators in older electronics gear. When the voltage applied across two electrodes inside the device reaches a certain potential, the gas ionizes, and the tube becomes conductive. When the voltage across the electrodes drops below a certain level, the ionization collapses, and the tube turns 'off'. An additional electrode can 'trigger' the tube to turn on when a much smaller trigger voltage is applied. A thyratron acts much like a current day semiconductor device called a SCR (Silicon Controlled Rectifier). The action of a thyratron makes a simple latch. Interconnecting these latches allows more complex devices such as logic gating devices, ring counters, and flip flops to be created.

One of the 'digit' daughter cards [Click on image for more information]

The majority (120) of the thyratron tubes reside on daughter cards that plug into edge-card connectors that are located on a main board that makes up the 'backplane' of the machine. Each of the twelve daughter cards make up one digit of the calculator's accumulator, as well as providing the display for that digit. Each of these digit cards contains ten of the thyratron tubes, connected as a ring counter. A ring counter is a series of latching devices, of which only one is active at a time. A pulse coming into the ring counter causes the currently on latch to shut off, and the next latch in the ring to turn on. The last latch in the chain is connected back to the first latch such that the chain loops back on itself, forming a ring. Each of the thyratron tubes in the ring is connected to a digit electrode in the Nixie tube, so that each stage of the ring counter activates a single digit (zero through nine) in the Nixie tube. Each clock pulse coming into the digit card will cause the ring counter to advance one position, essentially counting the clock pulses and displaying the count on the Nixie tube. Additional circuitry takes care of generating carry pulses as the ring moves from the '9' state, back to the '0' state.

Closer view of the display panel

The Nixie tubes contain only the digits zero through nine, with no decimal point, with the digits arranged in the order 1029384756 (front-most to rear). The Nixie tubes are an active part of the counting logic of the ring counter, and this specific order helps the counter properly cycle through its steps. The Nixie tubes in the Anita are made by Hivac, a European electronic component manufacturer. An unusual aspect of the Nixie tubes is that the glass envelope of the device is coated with a red-tinted see-through plastic-like coating. I'm not sure why the Nixies are coated this way, except perhaps to provide a protective surface for the tube itself, as there is no lens between the outside world and the face of the Nixie. The Nixies shine through open windows in the cabinet.

Control Ring Counter Boards (with lots of thyratron tubes)

Along with all of the tubes on the display daughter cards, there are three other boards full of tubes mounted perpendicular to the main board across the bottom of the machine. These boards also contain ring counters used mostly for control purposes. The ring counters on these boards are configured for specific tasks, such as scanning the keyboard (in columns, left to right), counting up/down the digits of the multiplier/dividend when multiplying and dividing, and selecting which digit is getting count pulses delivered to its ring counter.

The GS-10D "Dekatron" Counter Tube

The keyboard in the Mark 8 is scanned both horizontally and vertically. The horizontal scanning of the keyboard is accomplished by one of the ring counters on the control boards mentioned above. The vertical scanning (starting with all of the '9' keys, working down to the '1' keys) is done by a special and rather unique device called a "Dekatron" tube (or, in the US, "Counter Tube"). The dekatron is a specialized cold-cathode device that serves as a ten-stage ring counter, all housed in a single tube. Counter tubes were originally developed in the 1940's out of the requirement to be able to count events which occurred at a rate much faster than any then-available counting device could cycle. High speed counters were needed to record the result of experiments related to development of the atom bomb. Counter tubes were used to count the generation of subatomic particles in high-energy particle experiments. A counter tube utilizes some of the same properties used in Nixie tube displays, but rather than using numeral-shaped electrodes, the electrodes are nothing but little wire nibs. The nibs are arranged in a circle around a special center electrode. The tube is filled with a carefully selected mixture of gases, including neon, similar to that used in Nixie tubes. At any given time, just one of the electrodes in the tube will have an ionized area of gas around it (which just happens to cause an orange-red glow around the electrode, making it possible to visibly observe the tube in operation). By giving the tube an electrical 'kick' via other special electrodes, the ionized area will jump to the next neighboring nib. By electronically monitoring the charge on the electrodes, it is possible to tell which electrode is the one with the ionized gas surrounding it. In this way, the tube can serve to count the number of pulses that have been sent to it. The counter tube used in the Anita (a Hivac GS-10D) has 30 electrodes, with ten of them being the primary counting electrodes, with two alternate electrodes for each primary electrode used to provide the 'kick' that move the charged area from one primary electrode to the next.

For more interesting information on counter tubes, CLICK HERE to see an article on the technology by Tom Jennings.

The keyboard assembly also has a set of thyratron tubes in it which provide the logic for decimal point placement. Below the Nixie tube displays on the front panel of the machine, a group of small individual Neon indicators are used to indicate the decimal point position. These indicators are tied into the decimal point circuitry in the keyboard assembly.

The Mark 8 is also loaded with a couple hundred selenium rectifier devices (the black cubical devices seen on many of the circuit boards). These rectifiers are used in an interesting way in the Anita. In most situations, rectifiers are used to convert alternating current to direct current, using diodes as switching elements to siphon off the positive going side of the AC wave to one tap, and the negative going side to another tap. In the Anita, the rectifier packages turned out to be a convenient way to pack more diodes in a small package, for use as logic gating elements. Each of the packages contains two back-to-back diodes, essentially providing two diodes in one package. Also involved in logic gating functions in the Anita are eleven more conventional triode vacuum tubes (ECC-81[Europe] or 12AT7[US]), like those found in old radios and televisions. These tubes are also used in a few other places within the machine for clock generation and signal conditioning functions.

The Anita C/VIII Power Supply

The power supply of the machine is physically large, but not terribly sophisticated compared to the transistor-regulated power supplies in later calculators. A rather large transformer is mounted on the power supply circuit board, along with quite an assortment of power diodes, filter capacitors, resistors, and other components. The heater elements of the ECC-81 vacuum tubes are driven in series directly off of the power line, through a diode and a series of current limiting resistors. The line voltage also is connected to a constant-voltage transformer that steps the line voltage up to around 495V, which is used as the base supply for the rest of the circuitry in the machine.

The master clock oscillator in the Anita Mk 8 operates at 3 KHz (3000 cycles per second). While this is quite slow compared to the early transistorized electronic calculators (which had master clock frequencies measured in 10's or 100's of kilohertz), it is orders of magnitudes faster than the speed of the counting devices in earlier electro-mechanical calculators, that typically operated from between ten to thirty cycles per second. This fast clock speed allowed the Anita to perform counting operations much faster (not to mention, totally silently) than electro-mechanical calculators.

The single transistor (a later design replacement) in the Anita

Surprisingly, the Anita does have a single transistor in it. An early transistor was used in the power supply section of the machine. It appears that the transistor was used as a component in voltage regulation functions. In the machine in this exhibit, it seems that a newer transistor was substituted for one that failed over time (see photo). It is clear that this was not an add-on done at a later date, because a pictogram showing the transistor case outline is clearly shown in blue on the circuit board.

A closer view of the Anita C/VIII Keyboard

Since the machine was designed to run on the European power grid (220V, 50Hz), and the risk of causing damage to the machine, I do not wish to attempt power it up at this point. Without being able to power the machine up to exercise it, and not having any manuals for the machine, I have to make some guesses as to how the user operates the machine. I have been able to figure out from observation how some of its functions operate. The majority of the real estate on the front panel of the machine consists of 10 columns of keys labeled one through nine (a zero is signified by no key pressed in a given column). These keys are the main entry method for inputting numbers into the calculator. Situated between these columns of keys, below the '1' keys, are twelve smaller keys, apparently used for entering the decimal point location. At the lower-left of the keyboard area, there are four keys which select the math function mode of the machine. These are labeled "-", ":" (for divide), "X", and "+". Only one of these keys can be depressed at any time, and selecting a key different than the currently selected key causes the previous key to release. When either the "+" or "-" keys are depressed, the main numeric entry keyboard keys do not latch down when pressed. In addition or subtraction mode, when a numeric entry key is pressed, the digit corresponding to the column the key is pressed in immediately counts up or down (depending on whether the machine is in addition or subtraction mode) the number of counts specified by the key. This method of addition/subtraction is much like earlier comptometer type devices, where pushing the key down would cause, via mechanical or electro-mechanical methods, the corresponding digit counter to count the appropriate number of steps.

A View with keyboard swung out of the way

Multiplication and division are somewhat more complicated, and not entirely clear to me yet in how they work. When in multiply or divide mode, the main keyboard keys lock down when pressed rather than being momentary as they are in addition/subtraction mode. It appears that a single column of keys labeled zero through nine on the right hand side of the keyboard becomes active when multiplying or dividing, but it isn't clear exactly how these keys are used in performing these operations. A mysterious slide switch to the left of the main numeric entry keys is labeled "NON-SHIFT". I have no idea what this switch does. Two keys situated between the main numeric entry key array, and what I'll call the 'multiply/divide' set of number keys are labeled "DEC PT" (for signifying where the decimal point is in the multiplier or dividend in multiply/divide operations) and "CHECK X", which allows the results of multiplications to be verified for accuracy. Two "CLEAR" keys are located below the main number entry key array. One is labeled "CLEAR REG", apparently used to clear the accumulator, and the other "CLEAR KB", which unlocks the keys locked down when the machine is in multiplication or division mode. If anyone out there knows more about the operation of this machine, I'd appreciate hearing from you so I can fill the missing detail in this page.

A view of the back side of the keyboard assembly

The display area of the Anita consists of the twelve Nixie tube displays, with openings in the cabinetry to allow the Nixie tubes faces to protrude through the housing. Below the Nixies is a row of small square indicators (lit by discrete neon bulbs) for indicating the decimal point position. There is no indicator for negative numbers, so when negative numbers are displayed, they show up in 10's compliment form, for example, -1 displays as 999999999999). There also is no indicator for error/overflow conditions. My guess is that an overflow simply wraps the count back to zero (For example, with successive additions of 1, starting at 999999999997, the machine would display 999999999998, 999999999999, 000000000000), with no indication to the user that the overflow has occurred. It would be very interesting to see what an operational Mark 8 would do in the event of being commanded to divide by zero.

External Device Interface Connector

On the rear panel of this C/VIII there is a DB-37 style female connector which allows external devices to be connected to the calculator. According to Mr. Gary Mann, a former Anita factory service technician, this connector provided external access to the main main accumulator register of the calculator. At least two different devices were made which utilized this connector to provide additional functionality for the calculator. In one case, an external memory device was available that had a storage register (implemented with Thyratrons just like the calculator itself) that provided a means to store the number in the display, and recall it, at will. Another application used this interface connector to allow an output typewriter to be interfaced to the calculator, to provide hard-copy record of calculations, mainly for accounting applications. Not all Anita Mark 8's came with this connector, it appears to have been an optional item that could be ordered with the calculator.

The case of the C/VIII is mostly plastic, probably to help reduce the weight of the machine (which is still rather heavy at 33 pounds). The case is a bit of a puzzle to take apart, with the back panel coming off first, when two screws on the bottom of the case are removed. Once the back panel is removed, two long-shanked screws are exposed, that, when removed, allow the cover over the displays to be taken off. Once the display cover is out of the way, it makes it possible to remove the keyboard bezel. The side panels and bottom panel of the case are all one piece of moulded plastic, and most of the components of the machine are mounted to this base. The keyboard assembly is on a hinge arrangement which allows the keyboard (after removing the four screws that hold the keyboard assembly in place) to be tilted toward the front of the machine to gain access to the workings under the keyboard. There are no retainers or any type of latch mechanism to hold the keyboard in place while it is swung out of the way for service operations. At the the corners of the bottom rear of the case, there are two assemblies which allow steel 'legs' to extend out of the bottom of the case to raise the back of the machine to provide adjustment to the angle the machine sits at. The legs have a number of stops on them that provide different angles to make the machine more ergonomically acceptable in differing office desk settings.

A profile view of the Anita C/VIII

This particular Anita C/VIII is serial number 3108. It isn't known exactly how many of these machines were built, but in a paper presented by Norbert Kitz in early 1964, he mentioned that there were sales of 10,000 units per year. Given that production began in 1961, and that the sales of the machine likely continued at least into the 1967 timeframe, it could be that 50,000 to 60,000 of these machines were built. However, with transistor technology advancing so quickly during the mid-'60's, it was assured that the days of tube-based calculators were numbered, even before the Anita was created. Transistorized circuitry and advances in circuit board technology made a machine like the C/VIII economically uncompetitive in a relatively short time. Even so, the machine was still being marketed as late as 1967. A company in the US, Inter-Continental Trading, apparently imported and marketed the machines in the US. The machines marketed by Inter-Continental must have had power supply modifications in order to operate on the US power grid.

The Anita C/IX (Mark 9)
Image Courtesy Galerie Alte Technik

In its day, the C/VIII was unique, and most certainly fit the bill for any enterprise that needed fast, quiet answers to their mathematical problems. Sumlock Anita Ltd. went on to make other electronic calculators, including a follow-on version to the C/VIII which also used tube technology (the Mark IX). After the Mark IX, Sumlock Anita calculators were all based on transistor technology.

Thanks to Peter Lee and Klemens Krause for historical and technical information included here.

Thanks to Mr. Gary Mann for information relating to the external interface on the Mk 8, and to Nigel Tout for relating the information Mr. Mann provided.

Text and images Copyright ©1997-2011, Rick Bensene.