| +Home | Museum | Wanted | Specs | Previous | Next |
Olivetti Programma 101 Electronic Calculator
The Old Calculator Web Museum currently has two Olivetti Programma 101 electronic calculators in its collection. Sadly, neither is operational, and it is going to be rather a challenge to get either one of them fully operating.
The first machine came to the museum in the mid-1990's through an EMail contact indicating that they had one, but it needed work, and was missing some parts. The museum had been looking for one of these benchmark machines for some time, and arrangements were made to acquire it. It was well-packed and shipped to the museum, and when it arrived, it was found to be in fair physical condition, with no apparent damage from it's journey to the museum, but was missing a few parts as mentioned. Some pieces of the power switch assembly were missing. One of the buttons that controls the magnetic card reader is missing. One of the rubber rollers that helps to drive the magnetic cards through the card reader has decomposed into a very sticky thick black goo, which is a quite common occurrence for synthetic rubber over the course of many years, and is rather difficult (any messy) to repair, requiring a complete disassembly of the magnetic card reader in order to attempt any kind of replacement of the rubber roller. The main logic circuitry looked to be in good shape, all of the logic cards are present, and there are no signs of any catastrophic component failure. The wiring harnesses all looked to be in good condition, with no indications of chafing or overheating. The power supply assembly also looked good. The power supply was isolated from the rest of the circuitry and gently powered up using a Variac. All of the power supply voltages came up clean and within tolerance after all of these years. The keyboard assembly of the Programma 101, a complex mechanical assembly that encodes the key-presses for the electronics through mechanical means, was rather gummed up and will require a thorough cleaning and lubrication. Likewise, the serial drum printer assembly is also somewhat grungy and will require cleaning and lubrication. All of the rubber belts that link up the main motor that drives the printer, card reader, keyboard, and forced air cooling system are in serviceable condition. The magnetostrictive delay line that holds the contents of the working registers and program steps shows no sign of external damage (it is a sealed assembly), and the circuit board that provides write and read amplifiers for the delay line also looks good on visual inspection. The cabinetry is in good condition, with the original paint a bit oxidized from age. The show-stopper for this particular Programma 101 is the the gear-train that turns the printer drum with all of the digits and symbols on it. There is a composite or plastic gear that has disintegrated, and was found in many pieces on the baseplate of the calculator. It isn't known if this gear was made of such a material so as to be sacrificial in the case of some kind of jam of the drum drive mechanism, or if it simply failed over many years. Without this gear, the drum will not spin, thus rendering the calculator non-functional no matter how functional the electronics or the rest of the machine is. This gear is not easily replaceable, as parts for these machines are essentially unobtainium. Since the gear has disintegrated, trying to fabricate a replacement is not possible since the proper dimensions, tooth spacing and gear cut are unknown. Until either an original part can be sourced, or someone manages to find an original gear and can use it as a template to manufacture a replacement part (perhaps by 3-D printing?), this calculator will unfortunately remain in its current state.
An Internal View of the Programma 101
In the early 2000's, the museum was notified by the daughter of man who had purchased an Olivetti Programma 101 brand new in 1965 for his structural engineering business. He had used the machine to develop programs for all kinds of structural engineering calculations that he used to primarily develop bridge structures. His daughter wrote that he had long-since retired, and was now somewhat infirm, but he absolutely loved that calculator, and it still sat in a prominent space, under its cover, in his office. It was time to start clearing things out, though, as her father was soon to be going into a residential care facility, and the family had plans to downsize into a smaller home and needed to clear things out to sell the current home. She had found the museum online, and noted that the museum was looking for a Programma 101 through the museum's WANTED page, and wrote explaining the situation. The calculator was relatively local to the museum, about an hour and a half drive. Arrangements were made to drive to the home of the owner, and if the calculator was in good enough condition for the museum to consider adding to the collection, a price would be negotiated, and the machine would become a part of the collection.
Upon arrival at the home, there was a warm greeting and introductions, followed shortly by the introduction of the original owner of the calculator. While he had some difficulty getting around, he was still very sharp mentally, and was eager to tell all kinds of interesting stories about the calculator. He made it very clear that the calculator provided him a major advantage over his competitors who were still doing their structural calculations with slide rules or desk-type motor-driven mechanical calculators, which was literally orders of magnitude slower, and much more prone to error. He was able to put together proposals for designs much more quickly than any of his competitors once he got his custom programs designed and written, which he did completely on his own, using Olivetti's very well-written documentation to understand the programmability of the calculator.
He then said that he had purchased the machine brand new at Portland Typewriter Co. in Portland, Oregon, in 1965. He said that it was the very first Programma 101 sold in the Pacific Northwest, and that fact gained his business some notoriety as a small story was published about the purchase of this newfangled "electronic brain" in the local newspaper. Remember, in 1965, the common public's conception of any kind of machine that could compute complex equations was encased in a glass room, with huge boxes with spinning tapes, and people dressed in suits under lab coats who tended to the needs of the computer. The idea that a subset of this huge computers of the day could somehow fit onto a desktop was very foreign to most people, which is why the purchase of the Programma 101 for use in this man's structural engineering business was such a big deal.
Portland Typewriter and Office Machine Co. was the largest business equipment retailer and repair facility in Oregon for many years, and had a reputation for providing the latest in office technology, standing behind what they sold very well, had splendid customer service, and had a service and repair facility that was second to none in the Pacific Northwest. Portland Typewriter was founded in March, 1957, and quickly grew to become Portland's leading office equipment supplier, selling typewriters, adding machines, calculating machines, bookkeeping machines, office furniture, filing solutions, copying machines, and general office supplies, making it a one-stop shop for any office needs. The company went out of business sometime in the 1990's, likely due to the replacement of typewriters calculators by desktop PCs and Macs, with sales of these machines in warehouse stores and specialized computer stores making it hard for an independent to compete.
Looking over the Programma 101, it looked like it was brand new. It was in unbelievably clean condition. There wasn't a scratch in the paint, the keyboard looked as if it had hardly ever been used, and the controls were all in place and operated smoothly. Of course, the calculator was not plugged in and tested, as doing so could be problematic due to power supply issues after not having been run for a long time. The owner said that the last time it had been powered up was over ten years past, and it worked then. However, ten years is a long time to sit without any power being applied, and the filter capacitors in the power supply can degrade with disuse. The aesthetic condition of the calculator was wonderful, and I complimented its owner on how much care he had lavished on the machine during the many years he used it until it was finally replaced with a PC. He said that it received regular scheduled servicing for at leas ten years after he purchased it. A Portland Typewriter technician would come to his office and clean, lubricate and adjust the machine so that it performed exactly to specifications. He would check the power supplies, print quality, programming tests, as well as testing the magnetic card reader and adjusting as necessary. A service visit usually took close to 90 minutes, but there was no charge because when the calculator was originally purchased, a lifetime maintenance contract was purchased with it, at nearly the same price as the calculator, but it was indeed a lifetime contract.
Of course, there at this man's home, it wasn't possible for me to do an in-depth evaluation of the calculator, which would require some degree of taking it apart to do. The only way I would be able to determine the state of its insides would be to buy it. So, I queried as to the asking price for the machine, and his daughter spoke up with a price that was out of the question for me to pay, but just as I was about to make a counter-offer, the original owner spoke up and said "That's not enough, it's worth much more than that!", which as absolutely not something that I wanted to hear. His daughter than asked one of her daughters to take grandpa to his room, which he vehemently did not want to do, but he didn't have much of a choice. I said that I could not pay the asking price, as it was in the territory of prices being paid for fully operational machines on auction websites such as eBay, and I had no idea if this calculator worked or not. I said it looked really nice, but could be a complete mess electronically and possibly even mechanically. I said it was in nice "exhibit-able" shape for a museum that has displays behind glass for no one to touch, but the Old calculator Museum is a "living" museum, where the calculators are restored to operational issue if needed, and visitors are encouraged to power up their favorite calculator or one they are curious about and play with it for a while. A display piece was not what I was looking for, and I was not going to pay the asking price for a display machine. After some dickering back and forth, a price was finally settled upon that I still felt was more than I would have liked to pay, and I very carefully lifted it up (these things are not lightweight) and hauled it out to the car. I had a bunch of fluffy moving blankets lining the trunk area, as well as additional pads to place around the calculator, then a really big blanket over the top. By the time I'd hauled the machine out and made the payment, the original owner was sound asleep in his bed, and I did not want to disturb him. I thanked his daughter, and asked her to pass on my thanks and admiration to him for taking technology so new and turned it into an indispensable tool that gave his business a huge competitive advantage, as well as for his taking such good care of it over all the years. I said my goodbyes, and headed back to the museum.
The Programma 101 was set up the workbench in the museum, but I did not get any chance to look into the details of the calculator's condition for a few weeks. One I got around to it, the cabinetry was removed, and inside the machine also looked like new. The only sign that it had been used at all was some paper dust around the printing mechanism, and a few black and red coatings of ink on some of the characters around the print drum. The print drum turned freely, but was a bit noisy. I also noticed that it didn't look right to me. Something was different. I went and got the other Programma 101 out of storage, and compared the print mechanisms, and found that they were completely different. There did not appear to be a common part of any of the drum drive mechanism between the two machines. The new mechanism utilized a cogged belt to transmit rotation from the main motor to the printer instead of the gear train of the other Programma 101. The drum itself was the same, but the way it was driven was not. The worst part is that the drive belt was nowhere to be found. There were not pieces of it, sludge left from the belt deteriorating, or even dust from the belt crumbling. Either the belt had been removed at some point (which would have been a chore without cutting it) for some reason, or the belt failed at some point and the pieces of it were removed, and the machine left without the belt. Further inspection of the machine revealed that the same thing had happened to the rubber roller that pulls the magnetic card through the reader as in the other machine. The rollers had also turned into a gooey black mess. Everything else in the machine was in like new condition. It was spotless inside other than a bit of dust, and the electronics all looked to be in great condition, with no signs of any circuit stress, blown components or anything that would indicate problems had occurred with the electronics. Like the other machine, the power connections from the power supply to the logic of the calculator were disconnected, and the power supply was slowly brought up to line voltage using a Variac. The main drive motor started spinning, and the ventilation fan made quite a bit of noise which is obviously attenuated by the cabinet when it is in place. I could see the cogged wheel inside the printer that turns to drive the belt that in turn drives the print wheel, but with no belt in place, the print wheel doesn't turn, and without the timing pulses supplied by the print wheel, the logic would never operate properly even if the electronics were in perfect condition.
The bottom line is, two Olivetti Programma 101's, each needing a different part that hasn't been manufactured for 40+ years, that is unique enough that it's unlikely that a current-day replacement will work. The gear that the first machine needs has helical cut teeth, and is somewhat small in diameter, but is quite wide. Since it was broken into a number of pieces, getting accurate measurements of it to perhaps have a custom part made using 3D printing is next to impossible because some of the pieces are missing. In this case, the only likely way to replace the part would be if old assembly documentation were available that had dimensional information included. Likewise, the belt that is missing for the second machine is a complete unknown, because there's no sign of the part, and taking measurements of the cogs and the distance between them is very difficult without disassembling the print mechanism. The belt appears had to be a cogged belt, as it goes between two cogged wheels that are on shafts that turn via the belt. In both cases, without these crucial parts, neither calculator has any chance of working.
As time permits, both of the magnetic card reader/writer assemblies will be removed from the chassis, and will be carefully disassembled to remove the shafts with the deteriorated rubber drive wheels to see if they can be rebuilt using silicone or rubber tubing that has the proper inside and outside diameter, and if something can be found that works, the mag card assemblies will be thoroughly cleaned and reassembled. The keyboard, printer, logic chassis, ventilation system, and power supply will be removed from the chassis, and then The chassis and cabinet components will be thoroughly cleaned. The circuit boards will be carefully removed from the logic chassis, inventoried, photographed, and a more complete visual inspection given to assure that there are no obvious problems.
Two Views of the Logic Chassis of the Programma 101
Edge connector fingers will be cleaned, the keyboard mechanisms throughly de-gunked, cleaned, and lubricated where it appears that lubrication is needed. Once all is checked out, the machines will be carefully re-assembled, and put on the shelf until the missing parts can be found or fabricated. Hopefully at some point the proper gear or belt can be found and testing of whichever machine benefits from the right part can be done, perhaps even resurrecting the calculator.
The Programma 101 is a truly fantastic machine for its time, with a magnetostrictive delay line for register storage, and beautifully made circuit boards that use what is called cordwood construction, which allows the boards to stack one atop the other with the cordwood modules interleaving between the boards, making for higher density circuitry than traditional circuit board technology, as well as simply looking beautiful. The cordwood modules were standardized with each different module having a specific function, so that troubleshooting was fairly straightforward. Once a failing module was identified, there were between 4 and 8 wires connecting the module to the circuit board that could easily be de-soldered, and a replacement module soldered in its place. The printing mechanism is an ingenious combination of a serial printer and a drum and hammer printer. A single solenoid-operated hammer rides on a carriage that moves back and forth across the paper, with the hammer firing into the paper at the instant when the print drum behind the paper is in the correct position to cause the desired digit or symbol to print when the hammer presses the ribbon against the paper, leaving the printed character on the paper. It's magical to watch in operation, though a bit on the noisy side. The magnetic card reader marks the first ever use of such a device for offline storage for an electronic calculator. The next calculator to use such a device was Hewlett Packard's 9100A. Hewlett Packard ended up having to pay a royalty to Olivetti for each 9100A and 9100B calculator sold because of the similarities between their magnetic card devices and Olivetti's. Olivetti had wisely patented their design in the US, given Hewlett Packard little choice but to pay the royalty fee.
Service Tag found on the logic chassis of the Programma 101
The museum's first Programma 101 is serial number 1311468, and appears to have been manufactured in the late part of 1969. Each of the eight 13 1/2" by 9 1/2" main logic boards is marked with a date, which ranges from late October to late November, 1969. A service tag found inside the machine indicates that the machine was placed in service in January of 1971, and a repair service to the main logic chassis was performed in February of 1971 under warranty due to calculation errors.
The Programma 101 was introduced in the fall of 1965, and while it was not the first monolithic programmable calculator on the market (that distinction fell to the Mathatronics Mathatron), it was superior in many ways to the Mathatron, providing a regular adding-machine style annotated printout of its calculations versus the "ticker-tape" printout of the Mathatron. Its programming features were also superior to the Mathatrons', with conditional branching capabilities that were much easier to use than the same feature on the Mathatron. The Mathatron did have the Programma 101 beat in a few crucial areas, though. The Mathatron provided full algebraic problem entry, including parenthetical expressions, and autometic PEMDAS (Parenthesis, Exponentiation, Multiplication, Division, Addition, Subtraction) in that order of precedence, following traditional mathematical rules. The Programma 101 used a method similar to most adding machines for addition and subtraction, with algebraic multiplication and division entry, but without any parenthesis for grouping expressions. This amazing feature of the Mathatron would not be matched until years later.
While the earliest models of the Mathatron did not have this capability, subsequent versions introduced roughly around the same time as the Programma 101 provided built-in high-level mathematical operations such as exponentiation, logarithms, and trigonometric functions that the Programma 101 lacked. While programs could be written for the Programma 101 to perform such calculations, they were considerably slower than the built-in routines provided by the Mathatron, and used up precious program space.
One of the 8 Main Logic Boards of the Programma 101
Despite the Mathatron having been introduced well before the Programma 101, the Programma 101 is considered a watershed machine in the history of electronic calculating machines, considered by many as the "first desktop personal computer", although that nomenclature is dubious in its accuracy, as the Programma 101 was technically not a computer, as it lacked a few critical aspects of a computer that preclude it from being classified as such. The Programma 101 lacks the ability for its program to modify itself, a characteristic that is considered part and parcel to the stored program capabilities of a von-Neumann architecture computer. The Programma 101 also had only the ability to manipulate decimal numeric quantities, whereas in a computer, the binary notation used could represent any form of data ranging from character strings to numbers in other bases such as hexadecimal or octal, fixed point or floating point. Even though it lacked these characteristics, the capabilities it did offer made it a true wonder of its day - an easy to use mathematical tool that could be programmed to perform virtually any kind of calculation automatically. The Programma 101 was an instant hit in the marketplace, an even with its $3,200 initial price (which was a lot of money in 1966), the machine sold very briskly. Initially, Olivetti had difficulties producing the calculators in sufficient quantities to meet demand, and had to dramatically ramp up production capacity. Because of its ease of use and programming, and Olivetti's reputation in the business machines market with superlative electromechanical calculators and adding machines, the Programma 101 sold in much greater quantity than Mathatronics' calculators, thus giving Olivetti the spotlight in history.
The Cordwood Logic Modules
The design of the Programma 101 began in 1962, making it one of the earliest designs for an integrated desktop stored program calculator. At that time, the only electronic calculator on the market was the Anita C/VIII, a simple desktop four-function calculator that used thyratron tube technology that was considered by many to be outdated technology as soon as it was introduced, though the maker of the Anita calculators claimed that their tried and true thyratron and dekatron circuitry was more reliable than any transistor design, which was true for a time as Germanium-based transistor technology was beginning to mature. The Programma 101 was an aggressive design, utilizing computer-like transistorized logic modules, a large-capacity magnetostrictive delay line for working register and program step storage, the ability to load and save programs and data to magnetic cards, and an fast one line-per-second 30-column serial drum impact printer for recording the results of calculations. The circuit board technology used in the Programma 101 was ahead of its time, with traces on both sides of the board that utilized plated-through feed-throughs. Miniature cordwood (aptly named because the modules were stacked such that they interleaved with each other to optimize space utilization) modules that form circuit assemblies are populated on the circuit boards. These modules were placed so that two circuit boards could be placed component-side to component side, and the modules would neatly enmesh, allowing high component density in a comparatively small space.
The "interleaved" circuit board layout provides for high component density
The electronic design of the Programma 101 was brilliant, with genius of its inventor and project manager, Pier Giorgio Perotto(1930-2002) showing in all aspects of the design. Not only was the machine brilliantly-designed, it was also quite easy to learn to use, which made it an instant hit among educational institutions as the choice for early computer education. The relatively low cost (compared to any true computer) and high quality documentation for the Programma 101 led it being the "computer" that many people in the nascent career of computer programming learned the foundations of what became their career on a Programma 101.
As much as possible in my research relating to early electronic calculators, I would try to contact people that were involved in the design, manufacture, documentation, and marketing/sales of the machines at the time. It is becoming more and more difficult to find such folks anymore, as sadly, many of them have passed away, including Mr. Perotto. However, before he passed, I had made significant efforts to contact him to perhaps learn more about the process by which the machine came to be from the man who mad it possible, but sadly, had not made any headway. At some point, through sheer happenstance, I stumbled across an EMail address that might have been Mr. Perotto's, and using Google Translate, wrote an EMail in Italian asking him if it'd be possible to reach him by telephone for an interview. The Email address proved to be legitimate, and we ended up having quite a lengthy EMail discussion, and finally he provided me his telephone number, and we agreed upon a date/time to talk. I called up at the prescribed time, and there, on the other end of a string of electrons, was Mr. Perotto himself. I felt so privileged to be able to speak with him. His accent was pretty thick, but his English was good. We talked for some time about the development of the machine, about Olivetti's culture, and how the project to develop the electronic calculator was received within a company that was known for its exquisitely-made electromechanical calculators.
Pier Giorgio Perotto
When it came to talking about Mr. Perotto's assertion that the Programma 101 was the first "personal computer", I expressed my disagreement, saying that the machine was a programmable calculator, not a computer. His voice rose in intensity, and he very firmly stated that the machine was a computer, insisting that it was "not just a programmable calculator". He steadfastly refused to accept the definition of the machines he designed as a programmable electronic calculators. We agreed to disagree, and I will state here that despite our disagreement, I thoroughly enjoyed the approximately 1/2 hour of time I was able to speak with Mr. Perotto, and will never forget it. He was a genius and a gentleman who lead a team of some incredibly bright people to develop a calculating machine that completely re-defined "state of the art" in the electronic calculator industry for a great length of time. This accomplishment will be remembered, be it "personal computer" or "programmable calculator" by history as a true watermark in the development of personal computing devices.
Sadly, Mr. Perotto passed away on December 23, 2002, not all that long after I had my telephone conversation with him. On the day he passed away, the world lost a man who's brilliantly-designed machine set in motion the very beginnings of what became the personal computer revolution of the mid-1970's. While I maintain his machine was not truly a computer, the personal nature of the machine combined with its storage capacity and programmability did make the machine a true personal device capable of highly complex calculations at electronic speed. This was something that no other calculator, except perhaps the Mathatronics Mathatron could lay claim to. The Programma 101's dominance in the marketplace for almost three years in an intensely competitive market is a true testimony to the brialliance and foresight of its designers.
Assembly Line for Programma 101 Calculators, Ivrea, Italy, Circa 1967
The famous Italian designer Mario Bellini developed the stunning cabinet design of the Programma 101, which won countless industrial design awards and still looks artfully stylish even by today's standards. Altogether, the Programma 101 stood as the premier programmable desktop calculator until Hewlett Packard introduced its 9100A calculator in 1968. The fact that the Programma 101 was a design which was class-leader for almost three years, during such a period of dramatic growth in the electronics industry, gives clear insight to the engineering excellence of this calculator.
Some time was finally found to perform a checkout of the main drive motor and power supply of the second Programma 101. After checking the mains supply resistance, it seemed safe to try powering it up. The logic chassis was disconnected frm the power supply, and then power was applied to the system, and the drive motor spun up just fine, and the power supply seemed to be operating properly. The noise that the machine makes is substantial, with the main drive motor, two squirrel-cage fans for cooling (one for the motor itself, and another that cools the electronics through built-in ductwork), and the mechanical components that operate the printer and magnetic card reader all contributing to the noise of the machine at idle. The inside of the cabinet is covered with felt sound-deadening material, but even with the cabinetry in place, the machine is still quite noisy when idle. When printing, that noise increases dramatically, with the printer's hammer briskly slapping the ribbon and paper to cause the digits to be printed on the paper. The next steps in validating the machine will be to perform measurements of the power supply voltages and checking of power supply voltage ripple using an oscilloscope.
Until appropriate parts can be found to replace those that have failed or gone missing, little progress can be made toward restoration of either of these Programma 101s. If and when that occurs, this exhibit will be updated to note any progress made. As of this writing, there are not all that many remaining Programma 101s in fully operational condition. It is hoped that eventually, the Old Calculator Museum will be able to add one to that number.