21

My mom is an accountant and when I was a child I remember that she had at her work some odd accounting machines. I wish to identify what those were, but unfortunately the bits and pieces she remembers aren't much. Here's what I found out:

  • Time: Somewhere between 1988-1993.
  • Location: Latvia. This is important because in 1991 we regained independence from the Soviet Union and the Western market became much more accessible to us and the machines in question were of Western make. However it's also possible that they were acquired before the fall of the Soviet Union. In any case, since at that time Latvia was a fairly poor country, I wouldn't be surprised if these machines were bought already used and/or dated.
  • Count: there were 2 machines. One was simpler, the other one more advanced.
  • Neither machine had a display. I believe the output was printed.
  • I do not know how the input was done.
  • One (the more sophisticated?) machine could take cards with magnetic stripes on them. Similar to modern credit cards, but larger. I believe those cards contained or selected the program the machine ran.
  • The other one used (from what I gather from my mother's explanation) a sort of rail, where she affixed triangular and/or square pieces to, and then inserted this rail into the machine, and that then selected the program.
  • As for the functions - again, I couldn't really understand from her what they did, but it sounded like some sort of calculators on steroids or perhaps spreadsheet processors. You could feed them a list of numbers and they did math on them. For example, you gave a base salary and it then calculated the tax amount and the total amount.
  • She is adamant that one of the machines had the word "Ascona" on it and they used to call it "The Ascona". I do not know if it's a model or manufacturer name, and searching the internet for "Ascona" connected with anything computer-related yields nothing.
  • The devices were fairly large, taking up most of a decently sized table. One of them was perhaps a whole meter wide.
  • I think the machines were electronical in nature (at least the one taking the magnetic stripe cards), not mechanical, but I could be wrong.

Do these descriptions ring a bell? Perhaps, if the exact machines cannot be identified, at least something similar or a class of machines? These devices are unlike anything I've ever heard about.

2
  • Could you post some pictures.
    – cup
    Commented Feb 25, 2023 at 22:02
  • 2
    @cup No, of course not, since this is all from memory.
    – Vilx-
    Commented Feb 25, 2023 at 23:14

1 Answer 1

47

TL;DR:

The simple one was most likely an

  • Ascota 170 or one of its variation. The definitive clue here is the plug board with triangular and square pins used for programming.

enter image description here

Ascota 170-LB (LB standing for Lochband, English Punch tape).
(Picture taken from robotrontechnik.de)

The more sophisticated one was most likely a

  • daro 1750 or one of its variation. The clue here are the magnet cards, hinting it was a Magnetkontencomputer. Although not as definitive, together with time and place it's quite likely.

daro 1750 Magnetkontencomputer

daro (Ascota) 1750 also known as robotron 1750
(Picture taken from robotrontechnik.de)


[Sorry, most links are to German language pages, as there are no or only misleading ones available in English]


Origin in All Likelihood

Going by name and country, I would think at least one being an ASCOTA. In fact, I would think that both were Ascota, as they had essentially a monopoly on those. Within COMECON East Germany was designated to build this class of bookkeeping devices. There was virtually no accounting office in the Sowjet Union not using Ascota machinery.


Company

Ascote was the brand name chosen by the East German government for products of the former Astra- and Wanderer-Werke. Both located in Chemnitz (Karl-Marx-Stadt). After the war both(*1) were nationalized (*2) and turned into VEB Buchungsmaschinenwerk Karl-Marx-Stadt which continued production. By 1969 they were made part of VEB Kombinat Zentronik, centralizing all office and small computing machinery. Since new names are the cheapest way of modernisation, the machines were now as listed as 'daro' or 'daro Ascota'.

Nine years later, in 1979 Zentronik was merged with most electronics/computing companies into the VEB Kombinat Robotron, stylised as 'robotron', so naming changed again.

Ascota/Robotron were the main suppliers for small computing/office machinery in the SU as well as most of COMECON. But especially the SU.


Machines

The 'Simpler' One

Due the way of programming the (almost) only candidate would be a member of the Ascota 170 family. They were build for very long time and spread all around the world - SU states (like Latvia) being the main market.

Ascota 170

Ascota170 first version

(Picture taken from robotrontechnik.de)

The Ascota 170 might quite well fit what you implied as 'less sophisticated'. it was a purely mechanical computer. During

  • production from 1955 until 1983
  • about 332,742 units were made and
  • exported to 101 countries,
  • the SU being the biggest target market, including
  • many produced in SU only variants.

[There is a very detailed page (in German) about the series - as there were countless variants and extensions - at robotrontechnik.de.]

While the term mechanical computer may sound not much, the 170 was an unusual beast and maybe the peak of mechanical computing. In 1955 it for sure outclassed all electronic computers of similar or even lager size.

Here a some video showing a 170 calculating and printing a Fibonacci sequence and the square root of two. Babbage might have sold his soul to see something like this :))

It features one or more 12 digit (decimal) wide ALU and up to 5 result registers per ALU all connected over a mechanical parallel bus.

Ascota ALU

ALU With attached Result Registers
(Picture taken from robotrontechnik.de)

Essentially each of those ALU/register units were the equivalent of one Ascota 110 desktop calculator(*3). In addition up to 5 memory units with 10 storage registers could be added to the bus.

Storage drum with its 10 registers _Storage drum with its 10 registers
(Picture taken from robotrontechnik.de)

Programming of the 170 was done by fitting pins into a plug-board as seen here:

Man programming an Ascota 170 using a hammer :))

Man programming an Ascota 170 using a hammer :))
(Picture taken from robotrontechnik.de)

Using such plug boards is a distinct feature of (electro-) mechanical devices. Later electronic were fitted with different technologies (*4) Using triangular and square for different categories was, to my knowledge, unique to Ascota. No other machine used them (*5).

Detailed picture of a plug board showing the square and triangular programming pins

(Picture taken from robotrontechnik.de)

A program could have up to 53 steps made up one (or sometimes more) of the 90 instructions.

The system could be extended with various peripherals including punch tape output, so data could be recorded and used for further processing. While the earliest of those did simply record output as directed by the 170, a later unit was programmable computer in itself. Here a program tape did control which data was taken when from the 170 and optionally formatted before output onto tape.

enter image description here

(Picture taken from robotrontechnik.de)

Other remarkable extensions wer multipliers, mechanical or electronic. Most historic notable maybe the R12, a tube based multiplier that could be shared by up to three 170s.

enter image description here

(Picture taken from robotrontechnik.de)

The TTL based MD24 multiplier/divider was developed as late as 1980 aans almost exclusive for SU customers .

Ascota 170 variants

There were as well several variants (130/140/150/160/etc) of the Ascota 170 with reduced functionality (less ALUs and/or less storage drums and/or less registers per drum), so it could have been one of them.

Ascota 071

The Ascota 071, also known as 'daro Ascota 1350', 'robotron 1350' or 'KB', was a smaller, reduced function version without a full keyboard and only up to 3 ALU. More then 50k units were produced.

Ascota 071 base unit

(Picture taken from robotrontechnik.de)

It might not fit the search as programming was done by sticking pins onto a drum.

Ascota 071 programming drum

(Picture taken from robotrontechnik.de)

Program execution was also very limited due being tied to each print position. 42 instructions were available.


The 'More' Advanced One

Here the magnet cards are the most important hint here. Magnetkontencomputer (roughly: Magnetic Account Computers) were all the hype in accounting from ca. 1968 up to the late 1970s. They brought computer based accounting to many mid and small companies. The basic idea was to store all data for a client (or employee in payroll) onto one or more magnetic cards. Access was done by reading that card, processing as needed, and storing updated data on the same card again - in addition information like account balance could be printed on the card as well. Special ledger sheets could be produced holding the usual human readable information as well as the everything computer readable in a magnetic stripe along the border. All very handy and perfect for integration into existing office procedures.

These systems became the core computing families of the 1970s. Well known names are

and many others. They all were small but fully digital computers. while the very first were still transistor based, integrated circuits soon replaced them - only to be replaced by microprocessors in the second half of the 1970s.

Eastern Germany's answer was the

daro (Ascota) 1750 Magnetkontencomputer

Introduced in 1974 it soon became a major seller to replace Ascota 170 and gaining new customers (*6). The base unit featured 5 KiB of wire ROM holding the system software plus whole 8 KiB of core based RAM for data and loadable programs.

enter image description here

(Picture taken from robotrontechnik.de)

A magnet card reader "Magnetkontenaufsatz daro 1294" was available as add on. It was to be put instead and replacing the paper feeder

enter image description here

Magnet card reader
(Picture taken from robotrontechnik.de)

Magnet stripes were located along the border of regular sized paper sheets. Sheets pulled from above and read while being pulled in all the way. For payroll or ledger work the program would use the data read to advance it toward the next usable line any do whatever necessary. when finished the sheet was pulled in another time to write whatever booking data was added. Thereafter the sheet was dropped into a bin below (or for manual handling to the top again) and the next sheet pulled from above.

In addition the sheets could be used to load programs. The magnet strip did hold three tracks with a total capacity of 256 or 384 Bytes, depending on sheet size.


*1 - To be correct, only the parts on East German soil :)

*2 - Astra was in fact first put under direct Sowjet control for several years, partly dismantled and shipped to Russia, while the remaining factory were producing exclusive for SU demand.

*3 - Except the 110 was in no way a classic special to type mechanical calculator, but as mentioned made up from a dedicated ALU that gets feed data and operations. IMHO the pinnacle of mechanical computing, never reached again.

*4 - After all it would be rather stupid to add a motor and large mechanical levers just to read commands that could be way better and more compact stored using wires, punch tape or core.

*5 - As usual there is a but in form of an inbetween unit, the Ascota 700 Elektronische Buchungsanlage (Electronic Accounting Facility). Designed in 1971 as fully electronic replacement for the Ascota 170 it would have been a fully digital computer with 128 words of core based RAM and 2.5 KiB of program ROM, made up from paper cards with wires (each card 2 words). Except not all programming was done so, as the print unit was still programmed a mechanical plug board much like the 170.

Since production ended after only 50 units - most of them being test installations in Germany, I would believe it being extrem unlikely that one unit not only ended up in Latvia, but as well being operational for the next 20 years.

Levers reading the programming

(Picture of the sensing levers taken from robotrontechnik.de)

*6 - At least outside the SU, as Sowjet companies still continued buying the 170 way into the 1980s. Maybe one of the lesser great results of long term planing.

27
  • 6
    Astonishing that these things existed and I'm absolutely flabbergasted they worked and were delivered in production in such large quantities! Mechanical engineers of the last two centuries absolutely rule!
    – davidbak
    Commented Feb 25, 2023 at 23:40
  • 6
    I would travel long distances to see one of these operate - probably in some museum setting. A video showing the internals rattling around while doing a computation would be awesome!
    – davidbak
    Commented Feb 25, 2023 at 23:41
  • 4
    @Vilx- That does not add up. Machines that used programming as described (pins) are always and only mechanical ones, never electronic. Maybe there's a confusion between electric and electronic? The 170 was of course powered by electric moors (running at 360 rpm . essentially their clock :)). So from an operators perspective no levers are to be pulled. Further, the Ascota 170 is the only one I know that used square and triangular programming pins.
    – Raffzahn
    Commented Feb 25, 2023 at 23:44
  • 3
    @davidbak See the storage drums in action here. The ALU(s) are behind on the other side of the bus rods connecting all of them. Looking at the keyboard shows this being one delivered to USSR.
    – Raffzahn
    Commented Feb 25, 2023 at 23:50
  • 4
    That was a great read! Commented Feb 26, 2023 at 17:55

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .