Predecessors of PLC – K1-20, MS2102 – history of creation of the first domestic automated workplaces

Dear colleagues, let me share an excursion into the past of domestic automation, namely into the history of the creation of some of the first domestic hardware and software for automated process control systems. My name is Mikhail Nepomnin, I previously worked as the head of the automation design bureau of the EPO Signal and, as they say, I am ready to tell you first-hand how some of the first programmable controllers and automated workplaces were developed.

In the distant 1988, after graduating from university and working for 5 years in an elite workshop as a REA adjuster, I moved to OMA (the department of mechanization and automation), to the new bureau of automation of verification work (KB APR for short). Some of the engineers from the local polytechnic institute were moonlighting, introducing automation of verification of manufactured products to us. And we were on hand to help them and learned the ropes or comprehended the intricacies of automation.

The products were special pressure sensors manufactured for the military – they were produced in two types. Here I will finish the story about the sensors themselves, since they are secret products. But the automated workstations (AWS) for checking them during release are not a secret. I can talk about them freely. Moreover, the AWS are no longer used for release, they have been replaced by a new modification. So I am not even giving away a commercial secret.

And so, at the end of 1987, the design bureau was created, and in April 1988 I transferred there as a software engineer. I think that was my job title. The design bureau was predominantly youth-oriented. It was headed by a competent comrade, who six months later became deputy chief engineer, and I became the head of the design bureau.

The Polytechnic developed for us automated workplaces for checking and one automated workplace “central processor” for process control and maintaining archives, both on paper and in electronic form. The female part of the design bureau dealt with the CD for checking and helped the Polytechnic students to compose a program for the central processor in the high-level language TurboBasic. Or rather, the girls developed the program under constant supervision and prompting from the Polytechnic employees. And the male part of the design bureau, under the same conditions, developed a program in assembler for the K1-20 controller of the 580-series microprocessor microcircuits.

And don't laugh at us. There were no Windows back then, TurboBasic was a cool language and it wasn't a shame to write a program in assembler. Let me remind you that it was back in 1988. We planned to write a program for the central processor on a domestic DVK computer, but then we got hold of IBM computers. And if on the DVK computers we played the computer game Tetris in our rare moments of rest, then on the IBM computers we began to seriously study high-level languages ​​like Turbo and QuickBasic in addition to games.

Since I attended a course of lectures on programming in Algol and Fortran at the university, and also completed an internship in programming, I also began programming in BASIC together with the ladies of our design bureau. And I did not master one assembler for K1-20. By the way, studying BASIC was very useful to me later. Well, more about that later. But for now I will tell you about the composition and characteristics of the K1-20 controller. This is its typical name, I think. In reality, we had two controller options with the codes MC2702 and MC2721.

The pictures show how they differed from each other, but overall the controllers were almost identical.

The 5V power supply was powerful – 3A. The controller was heating up, but it worked. I will say that we managed to install three ARMs in one workshop and one in another. The work was in full swing. We were going to implement the “central processor” program and increase the number of ARMs and the coverage range. But (how often does this damn “But” interfere with life) the conversion occurred and sensors became needed in much smaller quantities, and with them our automation became unnecessary.

I wanted to give brief technical characteristics of the controllers, but I have not had the documentation for a long time. And I did not find anything on the Internet. Therefore, I am writing what I remember.

These controllers contained full-fledged microprocessor sets of the 580 series. With a direct memory access controller, interrupt controllers and a serial interface. The difference was in the design of the console and the number of connectors.

The MC2702 had 3 connectors, but they were large. The MC2721 has 5, of which 4 are smaller, and the fifth one is, I think, the same as the MC2702.

In the box in foil is a ROM chip for flashing user programs. The spare parts kit is wrapped in capacitor paper, except for the cardboard packaging. A small screwdriver is present. In general – a thrill or a dream of an automation engineer of Soviet times. So, let's move on to the main part.

Workstation composition:

1. Controller MC2702 (2721)

2. Power supply for controller GN 09-01

3. Voltmeter V7-28 (V7-34)

4. Interface block

5. Harnesses

One type of controllers for potentiometric sensors, another for inductive ones. Due to the passage of time and the onset of sclerosis, I don’t remember which one was used for which sensors. But I do remember which voltmeter was used for what. Since the B7-28 could measure voltage in fractions of the reference voltage, but could not work with alternating voltage, it calculated potentiometric sensors.

And the V7-34 voltmeter worked with inductive sensors. They were powered by voltages with a frequency of 400 Hz or 2000 Hz, and were easily calculated by those who knew how to measure alternating current V7-34. For potentiometric sensors, the reference voltage was taken from the standard power supply GN 09-01. Inductive ones were powered from a specialized AC unit of factory design.

The interface unit consisted of a separate case and a switching board and a board for communication with a voltmeter. The switching board performed a sorting of sensors on the collector. The voltmeter board programmed the voltmeters and sent the “Start” command when reading data from each tested sensor. The case served to connect the boards and connect the connectors of external harnesses.

It's a pity that I don't have a photo of the entire ARM, a photo of the interface unit and ARM diagrams. I can only draw a functional or structural diagram of the ARM and interface unit from memory.

Everything disappeared during the period of perestroika and wild capitalism. The industrial structure in our country was not very strong anyway. And during the period of late Gorbachev and Yeltsin it became completely useless.

Documentation was thrown away, equipment was stolen, and it was believed that those who do not produce material values ​​should be paid the minimum or even fired from work.

Well, now the process is slowly getting better.