Yet another LONG overdue post

Greetings to all,

With as long as it has been since I’ve posted here, many might think that I’ve fallen off the face of the earth. I’ve also not posted much on the Old Calculator Museum website, which may further add to such speculation. This posting is to say I’m still around, and have been preoccupied by a lot of other stuff in my life that has consumed the vast majority of my time.

I am getting along OK. A lot has gone down over the past couple of years, some of which is not all that great, but it is what it is, and I’m working through the challenges. But, I’m not going to bore my readers with that stuff. The important stuff is old calculators. And, there has been some stuff going on there that is exciting.

The coolest thing is that just two days ago, the museum took delivery of an amazing new addition to the museum. I have been searching for one of these machines for many, many years, and finally, one now makes its home here. The machine is a Wyle Laboratories WS-02 Scientific. I’m extremely excited about this addition, as this is a very uncommon, and also somewhat historical machine due to what its development spawned.

For those that aren’t aware of the story, there is an essay on the Old Calculator Museum website entitled The History of Compucorp that goes into a lot of detail of how Computer Design Corporation was spawned from Wyle Laboratories.

The Wyle WS-02 is the second (and last) generation of Wyle Labs’ calculators. Functionally, the earlier WS-01 is identical to the WS-02, with the difference being the medium used to store the working registers of the calculator. The WS-01 uses a small fixed-head magnetic disk, not unlike the disk drives in computers today, but storing on a tiny fraction of the amount of data that today’s disk drives (or even disk drives of computers in the 1960’s) hold. The disk drive proved to be rather temperamental which led to a lot of problems with WS-01 calculators sold to customers. As a result of the difficulties, the calculator engineering team did some redesign of the WS-01 to utilize a magnetostrictive delay line (a loop of special wire through with torque pulses representing ones and zeroes travel through the wire at sonic speeds resulting in a time delay, or storing of the bits in the wire as they circulate through) to replace the disk drive. The resulting machine was the WS-02.

The museum received the WS-02 calculator in amazingly good physical condition. The main issue is oxidation of the plastic keycaps on the keyboard, which makes a white film over the keycap that makes reading the legends on the keys somewhat difficult. It is expected that this will be able to be remedied, but care must be exercised to make sure that the legends aren’t damaged or the structure of the keycaps is not compromised in the process. Also included in the acquisition was the model PC-01 punched card reader, that plugs into the WS-02 calculator to provide keystroke programming, via codes punched into special cards. The card reader appears to be in good condition physically. Along with the calculator and punched card reader, two original manuals for the machine were included, which is amazing, as documentation is usually lost with time.

The machine was originally purchased sometime in the mid-1960’s by a company that was involved in land development, surveying, and construction. The calculator was used to perform surveying and construction calculations. It is not entirely clear, but the WS-02 and PC-01 may have been part of what is called a WSS-5 or WSS-10 system. The WSS stood for Wyle Scientific System, which was a small desk, with a compartment with electronics in it that the calculator and punched card connected to that provided additional storage registers (8, 16, or 24 registers) and patch boards that could be wired with program steps. If the WSS-5 or WSS-10 was part of the system, it was not retained. The company used the machine as part of its operations until sometime in the early 1970’s, at which time the company suffered tough times, and ended up closing. When the offices were being cleaned out, one of the employees saw the calculator sitting out on a table (which may have been the WSS-5/WSS-10), waiting to be thrown out. He asked his management if he could take the machine, as he thought that it was kind of cool. His manager said that it was fine to take it, and he took it home, and stored it away in his basement. The machine was in full operating condition when it was put away in the basement. The machine remained there all these years.

In early May of this year, I received an EMail from the owner of the machine, saying that he had pulled the calculator out of his basement, and did an Internet search on it, and found the Old Calculator Museum’s WANTED  page for the Wyle WS-01/WS-02 calculators. The EMail asked if the museum would be interested in acquiring his machine, as it was unlikely that he would be doing anything with it, and felt that it should go to a place where it would be preserved and documented. Over the following weeks, and agreement was made, and in early July, the machine was packed up and shipped from Rhode Island. The machine arrived at the museum on July 15th, in an amazing custom-built crate that the owner crafted to assure safe transit for the machine.

The machine made the trip with no problems at all. The packing was incredible, and essentially the crate could have likely survived a drop off the back of a truck with no ill effect to the calculator.

Now begins a slow and methodical process of checking out the electronics in the machine to assure that things like power supply capacitors, edge connector sockets, and wiring harnesses are all in good condition, and if any faults are found, properly repaired. It will likely be some time before the machine will be ready to attempt to power up, but it is hoped that it will be able to be made fully operational.

Of course, a detailed exhibit for the calculator will be created for inclusion in the Old Calculator Museum website.

On other calculator-related topics:

– The Monroe EPIC-3000 calculator that was written about in old postings here has been restored to full operation. It is in the process of being documented for its exhibit in the museum. It is quite exciting to have this calculator working fully, as it is very much a hybrid of electromechanical and electronic technology, and the mechanical aspects of machines like this can be quite difficult to diagnose and repair.

– The museum received a donation of a huge amount of old Friden parts and documentation. Included in the lot was a large number of copies of Friden’s internal magazine, Friden News, which I’ve only begun browsing through and have discovered a lot of very interesting historical information, including introduction dates of Friden calculators, as well as stories about the development and early sales of Friden’s first electronic calculator, the EC-130. There is also a lot of information about Friden’s other products, including the Computypers (small-office billing machines/computers), Flexowriters, Punched tape equipment, Postage Equipment, and in later editions, information about Friden’s computer system, the System 10.

– A number of calculator donations and acquisitions have come in: Addo-X 9958 (essentially a Sharp Compet 32 in beautiful condition), Bohn Omnitrex 12, a Master H-2, a Wang 370 Programmer (fully operational after minor repair work), a Monroe EPIC-2000 (needs some work), and an Wang 360SE that needs some power-supply work. It is just a matter of time until I can get these documented and up on the museum website.

I wish all those who read this posting the best of everything.

17-Feb-09: Friden EC-130 Display System

I recently received an EMail from Mr. Jack Bialik that contained some very interesting information about the development of the CRT-based display system that ended up being used in Friden’s first electronic calculator, the Friden EC-130. All of the information contained in this posting is from Mr. Bialik’s memories of a project he was involved in at Stanford Research Institute in the early 1960’s.

Mr. Bialik obtained his BSEE from University of Michigan in 1950. After graduating, he worked at Consolidated Vultee Aircraft Corp. (CONVAIR), where he was involved in development of a display system utliizing CONVAIR’s Charactron display tube technology. Joseph McNaney of CONVAIR invented the Charactron tube in 1949, but the production operations were later transfered to Stromberg-Carlson (S-C) by General Dynamics, the parent corporation of (among others) CONVAIR and S-C. In late 1955, Mr. Bialik left CONVAIR, and joined Stanford Research Institute (now known as SRI International), a non-profit research and development organization founded in Menlo Park, California, in 1949. The Old Calculator Museum wishes to thank Mr. Bialik for sharing his memories.

In the latter part of 1961, SRI was contacted by Friden Calculating Machine Company’s VP of Research and Development, Mr. Larry Robinson. Robinson requested a proposal from SRI’s Computer Lab to design and develop a prototype transistorized CRT-based numeric display system that could display four lines of 27 digits on a small CRT display. Friden’s stated intention then was to use the SRI’s research efforts as the basis for producing an electronic display for an electronic calculator that Friden was planning to build. Friden’s requirement for such a display for this calculator was defined by the desire for the machine to be able to display the entry register, the result register, and temporary registers used to hold intermediate results of calculations. Existing display methods (Nixie or Pixie tubes) would require way too much space, power, and expense in order to display a similar amount of data. The use of a CRT display would provide a much more compact and efficient means to display this quantity of information.

Mr. Bialik, and his immediate Supervisor, Milton B. Adams, wrote up a proposal for the project that Friden accepted. Work on the project began in late 1961. A five-man design team was put together, with Mr. Bialik as the Project Leader and architect; Dave Condon and Dale Masher performing design work (logic and circuit implementation); Don Ruder to develop a separate testing system to drive the display subsystem; and Bill Stephens to fabricate the designs.

In early 1962 , SRI delivered to Friden three hardware copies (and associated documentation) of an engineering prototype display system that met the requirements established by Friden.. The display system contained four plug-in circuit boards that contained all of the circuitry to implement the display system, including the high voltage drive for the CRT. The prototype units were packaged in an aluminum housing with a viewport that allowed the face of the CRT to be seen, as well as house the electronics and power supply for the display system. Also included in the deliverables was a “calculator simulator”, a device that would allow digits to be entered into a keyboard and displayed on the display subsystem. The simulator device provided a means to test and troubleshoot the display system, and also to demonstrate that it indeed operated. The calculator simulator device was not a calculator — it could not perform any arithmetic. It only provided a means for entry (via a keyboard); storage (via a small magnetic drum); and control logic (transistorized circuitry) that would provide a source of data for the display system to display. Along with the hardware, all of the design information, engineering notebooks, and any other data related to the project were turned over to Friden when the project was completed and signed off.

Along with all of the work on the project itself, a patent (US Patent #3430095) on the principles of the display system and the “calculator simulator” was filed. It isn’t clear if SRI drafted the patent application for the concepts of the display system on its own, or if this was part of the arrangement with Friden. What is known is that because the work done by SRI was an exclusive “CLIENT CONFIDENTIAL” contract with Friden, once the patent was approved (not until February of 1969), SRI assigned all rights to the patent to Friden Calculating Machine Co. The patent lists Mr. Bialik, Mr. Masher, and Mr. Stephens as the inventors, but makes no reference at all to Stanford Research Institute.

The display system worked as required, and Friden appeared pleased with the results. The design of the display system was used pretty much un-modified from its SRI-designed form in various calculator prototypes. An early prototype electronic calculator, patented by Friden (US Patent #3474238), was based on a magnetic drum memory system, very similar to that used in the “calculator simulator” developed by SRI. Diagrams and text in this patent are very similar to those listed in the patent for the display subsystem and “calculator simulator”. Later patents from Friden outlining design prototypes that led to the development of the EC-130 also used much of the material in the original patent with little changes.

Although more research needs to be done, it seems pretty clear that in the early stages of brainstorming their ideas for an electronic calculator, Friden grappled with issues relating to how they were going to display the working registers of the machine that they had envisioned. One of the early prototype calculator patents filed by Friden indicated that it was considered very important that the calculator be able to display all of its working registers for the operator to see. As a result of this requirement, and limitations with existing numeric display technology, Friden had to look outside the company for design expertise in display systems technology. While it’s clear that Friden had internal resources skilled in the art of digital design, perhaps the “analog-ness” of the design requirements to generate a CRT-based numeric display required skillsets that didn’t exist in-house., This is probably why Stanford Research Institute’s Computer Lab was hired to do the design.

While the development of the display technology certainly played a significant role in making the EC-130 an early reality, the display system was only a part of what was needed to make a complete electronic calculator. It appears that much of the display system concept developed at SRI, along with some concepts from the “calculator simulator” (including basic transistorized logic gate designs) were used by Friden in the development of the EC-130. However, clearly the internal design work that went on at Friden to put the “brains” behind the display system was by far a more challenging task.

This commentary is in no way intended to take away any of the significance of Friden’s engineering effort in the development of the EC-130. It is, however, an interesting new tidbit of inforamation to add to the story of the development of Friden’s first electronic calculator.

28-Jan-09: Busicom 161

Apologies for delays in posting to the blog — things have been somewhat chaotic around here.

Some more work has been done on the Wang 370/371, but it needs to be written up and posted.  Hopefully I’ll get to this soon.

This posting is about a recent development relating to a Busicom 161 electronic calculator. For background, the Busicom 161, introduced in July of 1966, is the first electronic calculator sold by Japanese calculator manufacturer Nippon Calculating Machine Co. (NCM).  NCM had made a big market for itself in the Japanese market by making wonderfully compact and reliable hand-operated mechanical calculators that could replace the soroban (abacus).  With the advent of the electronic calculators, first by Sumlock Comptometer/Bell Punch (the Anita C/VIII) , followed by others, NCM felt that their corner on the market for small mechanical calculators could be at risk.   As a result of this concern, NCM embarked on a program to develop an electronic calculator.  The Busicom 161 is the result of their efforts. NCM created a new business subsidiary called Busicom (Business Computer) that would be the trade named by which NCM would market their electronic calculators.

Rumor (not well substantiated at this point) is that the design of the Busicom 161 was based on information gleaned from reverse-engineering a calculator made by Italian calculator maker, IME (Industria Macchine Elettroniche), the IME 84.  The IME 84 was a brilliantly-designed machine, implemented with all-transistor logic, magnetic core memory for register storage, and Nixie tube display. There are claims that the IME 84 was the first electronic calculator to use magnetic core memory, but this assertion is incorrect, as the Mathatronics Mathatron calculator holds this historical distinction.

The Busicom 161 may share the same architecture of the IME 84, but its implementation is much less sophisticated. The IME 84 has a much higher component density, with the entire machine housed in a compact, low-profile desktop package. The Busicom 161, in contrast, is comparatively huge. It takes up about the same amount of square inches of desk space, but it is much taller. The circuit boards in the 161 are very simplistic, using low-component density (the most complex boards have 20 transistors on them), with circuit traces only on the back side of the board, and single-sided edge connector fingers. The 161 uses a total of 42 circuit boards, arranged in three rows of 14 boards each, stacked vertically in the chassis, which is the reason why the 161 is so tall. The machine has a capacity of 16 digits, with fixed decimal of 0, 1, 2, 3, 6 or 9 digits behind the decimal point. The Nixie tubes display the digits zero through nine, and decimal points are indicated by discrete neon lamps situated between the Nixie tubes. The 161 has a single accumulator-style memory register, and one-key automatic square root. Transistors used in the Busicom 161 are made by Mitsubishi, using standard 2S-series silicon transistors.

NCM/Busicom holds a special place in history. In the late 1960’s, Busicom engineers had developed a design for a complex calculator chipset that could be configured to perform the functions of a large number of different calculator applications. Busicom contacted fledgling integrated circuit memory manufacturer Intel, to see if they could fabricate the chips that Busicom had designed. Intel’s engineers looked at the complex nature of the chipset, and decided that it would be more efficient to develop a small programmable computer on a chip that could be programmed to perform whatever operations a calculator designer required. Intel took this idea back to Busicom, and while Busicom management and engineers weren’t initially impressed by the idea, and instead pushed back at Intel to simply fabricate the chips as originally specified, Intel’s engineers had developed a breadboarded prototype of a simple calculator based on the idea, and this convinced Busicom that this may well become the future of electronic calcualtors. Busicom negotiated an exclusive contract with Intel to produce the programmable processor, which historically became known as the first commercial microprocessor, the Intel 4004. Intel also developed support chips that provided read-only memory, random access memory, and peripheral interfacing chips. Busicom and Intel jointly developed a business-oriented desktop printing calculator that Busicom sold as the model 141-PF — the world’s first microprocessor-controlled calculator, introduced in October, 1971. Sadly, the Japanese economy went into a recessionary period during the early 1970’s, and in spite of the technological triumph that Busicom 141-PF represented, NCM/Busicom started to suffer financial problems. Intel saw an opportunity to get out of the exclusive agreement with Busicom for the 4004 microprocessor and support chips, and a deal was negotiated to allow Intel to sell the microprocessor to other customers just a month after the 141-PF was introduced. This development led to Intel further developing its microprocessor business, leading to the Intel 8008(April, 1972), and then, the 8080(April, 1974), the microprocessor chip that truly ushered in the beginning of the personal computer revolution. By February, 1974, Busicom’s financial woes prompted it to file for bankruptcy. In November of ’74, NCM/Busicom ceased operations, making the company the first major Japanese calculator manufacturer to succumb to ongoing shakeup in the calculator industry that resulted in many calculator makers getting out of the calculator business, or going out of business altogether.

Fast-forward to more recent times. A while back, the Old Calculator Museum managed to acquire a complete Busicom 161. It has not yet been documented in the museum because the machine needs restoration before it can be exhibited. The main issue with the machine was that the edge connector sockets in the hand-wire backplane were disintegrating. It isn’t known if this is endemic to the material used in the connector sockets in the Busicom 161 in general, or if this is a malady suffered by this particular machine due to some kind of environmental conditions during the machine’s lifetime. Appropriate replacement sockets were found, and a number of the original sockets that were in the worst condition were replaced. This operation was extremely tedious as the wire density in the backplane is high. The power supply was checked out to be OK, and once the connectors were replaced, the machine was powered up, but alas, it doesn’t function properly. Since then, more edge connector sockets have started to fail, meaning yet more connector replacement must be performed.

Now, to the present. About a week ago, a frequent donor to, and good friend of the Old Calculator Museum, sent an EMail indicating that there was an interesting looking calculator chassis up for auction on eBay. I checked it out, and realized right away that it looked very familiar. The machine at auction was a complete Busicom 161, minus it’s cabinet. In further EMail dialog, this friend indicated that they would be willing to bid to win the calculator, and then donate it to the Old Calculator Museum. The bidding went well, and now the Busicom 161 chassis is on its way to the museum.

It is hoped that the soon to be received chassis will have backplane connectors that are in good shape. The goal will be to try to get a fully operational machine, either by getting the chassis up and running, or by mixing and matching parts to get a working machine going.

Watch this blog for more information on this project.