/reference/

My biography

 
Hardware designer biography

    When a man begins to feel his age as substantial he wants to write a book about his life, at least most of men. An approaching of my fifty stimutates this attempt fo me. I try to do it in a modern way- to create web page about may way.
    I have been working with Institute of Nuclear Physics since 1976. Most of my time was spent for automation hardware designs. Each designed device should be supported by test software and a lot of time was spend for application and test software designing. We used exotic computers- ICL-1900. In 1984 my colleges designed CAMAC microcomputer which replaced ICL-1900. It was very uncomfortable to work with this rare computer and I wrote a few system component software (editor, text processor, print utilities an so on) to increase a comfort of my job. In some years years this side of my activity consumpted a lot of time. But my main activity was hardware design. This page describes most of my hardware designs.


1976
    I graduated Novosibirsk ElectroTechical Institute in 1976 and came in Institute of Nuclear Physics. The first year of my activity was spent for designing conveyer ADC. As normal graduated student I was learning hard, but most of ideas came from my teacher of those days- Mr. Batrakov. That ADC had not bad parameters for those days- 8 bits and 20 Msamples per second. We assembled two devices with CAMAC interface and memory in 1977, gave them to customers and that direction was closed (mostly by my efforts).

1977
    I had learned to something in that year, probably. A few ideas appeared which contradicted to official policy in this field. A personal design appeared. My institute bought first arithmetic-logical unit IC and I wanted to use it in ADC. The idea was very simple- to make tracking ADC with parallel comparator section. I assembled this ADC on prototype board during one week-end. The device had 8 bits and could process signals 80 KHz. It turned out that aparameters of this TTL device exceeded pameters ECL tracking ADC which my colleges were designing at that time. It was fine result and my device was recognized as a perspective one. This ADC board was used in different models of our digital oscilloscopes (8100, 8100.2, 8500). I returned from my first vacation with three schematics (tracking ADC board, successive-approximation  ADC board (101) and CAMAC interface board). These schematics were basement for digital oscilloscopes family which we began to produce from 1978. 
1978-1979
    It is very difficult to recover a chronology these years. There were designed and produced first digital oscilloscopes, first fast analog multiplexer for them and auxiliary devices. It was time of appearance of russian middle integration ICs and first memory ICs. That was a reason for designing many versions of our digital oscilloscopes. A low level of russian ICs allowed to implement a digital oscilloscopes on at least 3 PCBs. These version were produced from a pair ADC-boards (tracking ADC and successive-approximation  ADC boards), three memory boards (fast memory on TTL and ECL ICs and low speed board on DRAM chips) and the single CAMAC interface board.
 
Conveyer ADC. There were produced an prototype device and a pair of oscilloscopes shown on photo. An ADC board consist of four comparator sections which fixed 2 bits each section. A delay of analog signal between sections was implemented by a coaxial cable (total delay was 150 ns). This cable defined dimensions of device. There were used digital ICs 100LP16 as fast analog comparators.
    The device provided 8 bits resolution, 256 words of memory, 50, 200, 1000, 4000 ns  time of measurements, four input ranges.
    ADC 8100 - 4M size, tracking ADC, TTL memory.
Resolution - 8 bits.
Input ranges - from 0,08V to 10,24V.
Input resistance - 160 KOm.
Conversion rate - from 125 nsec to 50 msec / point.
Memory length - 256 words.
Maximal frequency of processed signal (full amplitude) - 70 KHz.
    A few devices were produced. Then we bought ECL memory with capacity 256 bit/package. It allowed us to modify the device.
    The ADC board was based on 8-bit DAC implemented on KT342 transistors, diodes switches and R-2R ladder and a parallel comparator section implemented on 554SA1 (15 comparators). The parallel section measured a difference between input signal and comparator voltage then an error code was added to DAC code by ALU.
    ADC 8100M -  4M size, tracking ADC, ECL memory. 
Resolution - 8 bits.
Input ranges - from 0,08V to 10,24V.
Input resistance - 160 KOm.
Conversion rate - from 125 nsec to 50 msec / point.
Memory length - 256 words.
Maximal frequency of processed signal (full amplitude) - 70 KHz.
    This device consumpted less energy than previous version.
    ADC 8100.2 - 4M size, tracking ADC, DRAM memory (memory board was developed by Batrakov A.). 
Resolution - 8 bits.
Input ranges - from 0,08V to 10,24V.
Input resistance - 160 KOm.
Conversion rate - from 1,25 mcsec to 50 msec / point.
Memory length - 4096 words.
Maximal frequency of processed signal (full amplitude) - 70 KHz.
    These modules were produced not many. A reason of creating this version were difficulties with manufacturing ADC101 boards. We produced a few ADC8100.2 devices to satisfy urgent needs. This device was able to work with our multiplexer KAS-4 in multichannel systems.
    ADC 101 - 4M size, successive-approximation  ADC, DRAM memory. 
Resolution - 10 bits.
Input ranges - from 0,1V to 8,192V.
Input resistance - 15 KOm.
Conversion rate - from 1 mcsec to 2 msec / point.
Memory length - 4096 words.
Maximal frequency of processed signal (full amplitude) - 200 KHz.
    This device had a very complicated biography. Main components of ADC were crated and tested by Batrakov and me, jointly. Drawing schematic, routing PCB, assembling prototype board was done by me. Debugging first ADC board and next modifications were done by Batrakov. A memory board was designed by Batrakov and CAMAC-interface board (and tests) were designed by me. And who is the author then?
    KAS-4 - 2M size, 4-channel fast analog multiplexer with input programmable gain amplifiers. It was designed to work with our digital oscilloscopes ADC101 and ADC8100.2 in multichannel systems.
   Manual CAMAC controller. Mr. Batrakov and me were pioneers in developing CAMAC devices. After assembling our first CAMAC device we discovered that all our Institute had the only manual controller and no computer controller at all. We didn't want to wait and made a pair of that devices. 
  Dataway display. We had a necessity of CAMAC dataway display. I had to design it. This device was manufactured in significant volume (hundreds modules). Here is shown a photo of second version (a first version included two PCBs). 
   Generator. We had a shortage of different test equipment. We had no convenient pulse generators. Most of russian industrial generators loved to die after occasional short-circuit of output. Traditionally, I assembled prototype during week end, then assistants made a serial version. The device was implemented in CAMAC standard to avoid making power supply. The device was not controlled by CAMAC. It only used power suply of CAMAC crate. This device is used now.

1980-1981
    USSR electronics industry was developing these years. When we bought 1Kbit MOS memory, we immediately decided to make our digital oscilloscope set more technological and reliable. Newest ICs allowed to design a single interface/memory (2K*12bit) board. We used with this board the same ADC boards (tracking ADC and successive-approximation  ADC). In these years was designed a new fast multiplexor (KAS-8). There was introduced in our product line a new device type- a pulse voltmeter. And now the same but in more detail.
 
    ADC8500 (C0638)-  3M size, tracking ADC.
Resolution - 8 bits.
Input ranges - from 0,08V to 10,24V.
Input resistance - 160 KOm.
Conversion rate - from 500 nsec to 50 msec / point.
Memory length - 256 words.
Maximal frequency of processed signal (full amplitude) - 70 KHz.
It's possible using with KAS-8 in multichannel systems.
  ADC101M (C0615)- 3M size, successive-approximation  ADC.
Resolution - 10 bits.
Input ranges - from 0,1V to 8,192V.
Input resistance - 15 KOm.
Conversion rate - from 1 mcsec to 2 msec / point.
Memory length - 2048 words.
Maximal frequency of processed signal (full amplitude) - 200 KHz.
It's possible using with KAS-8 in multichannel systems.
   KAS-8 (A0634)- 2M size, 8-channel fast analog multiplexer with input programmable gain amplifiers. It was designed to work with our digital oscilloscopes ADC101M and ADC8500 in multichannel systems.
Input channels - 8.
Settlement time - 500 ns.
Gain 1,0 or 0,25.
Input range - 8 V.
Output swing - 2 V.
Input resistance - 40 Kohm.
  ZIIS-4 (C0639)- a pulse voltmeter. It consists of four sample-and-hold and ADC.
Resolution - 10 bits.
Accuracy - 0,1%.
Input range - 2046 mV.
Input resistance - 1 KOhm.
Settlement time - 2 mcsec.
Maximal frequency of processed signal (full amplitude) - 20 KHz.

1982
This year I developed next version of CAMAC dataway display. It was compact version, 1M size. But most users preferred old version (2M size).
    This year I participated in developing a digital oscilloscope based on storage tube ("Magnolia"). This device was developing by all of us (Batrakov, Chukanov). I dealed with digital part of device. Reading, digital processing (detecting screen defects, removing defect areas, determining curve of signal) and transferring data to CAMAC. Here was used a pair of interacting microcontrollers. This experience allowed me using arbitrary microcontrollers in single system without any problems in their interaction. It's pity but I haven't photo of this device.
    Parameters of "Magnolia" digital oscilloscope
Resolution - 7 bits
Iput range - from 0,128 V to 16,384 V
Input resistance - 50 Ohm
Maximal frequency of processed signal (0,7 level) - 130 MHz.
Horizontal resolution - 128 points
paasas - from 64 ns to 8192 ns
Horizontal nonlinearity - 2%


1983-1984


    That time I decided to make a CAMAC controller. We used K0601 controller. It suited users from installations, but it didn't suit me as CAMAC designer. To check a presence or absence of X/L signals I should waste a lot of time and efforts. Checking Q signal reuired about 1 second. It was easily to check reaction of device on F8 command by manual controller than by computer test. It was reason for my design. My crate-controller (K0612) was compatible with K0601 (old program could work with new controller without modification) but provided a number of new opportunities. It was possible to read X and Q state by single command, it was possible read all LAMs by single command. There was used a pair of microcontrollers.
    We produced about 50 controllers K0612.
USSR electronics industry was developing these years. When we bought ADC chip with 10 bits resolution and 1mcs time conversion we decided to renew our digital oscilloscope set. Newest ICs allowed to reduce hardware and improve performance of devices. It was to be expected a lot of new digital oscilloscopes. To reduce software expenses of mine and our customers I composed a standard of interface. It was a "S" series, that means a standartizied device. All new designed digital oscilloscopes had to have an identical set of commands, registers and ranges. I had participation in designing a pair models of devices- ADC-101S and ADC-850S. After this our young designers, directed by Batrakov, developed about five models of devices. All of them were testing by single test set. After appearance of next model I added a few constants in tests (real resolution, size of memory, using ranges from standard set, presence of analog multiplexer and etc.). 
        ADC-101S
    This design was made with Batrakov (he designed sample-and-hold for device).
        Parameters of digital oscilloscopes C0616 (ADC-101S)
Resolution - 10 bits.
Input ranges - from 0,1V to 8,192V.
Input resistance:
    for 1.28-10.24 ranges - 100 KOm
    for 0.08-0.64 ranges - 425 KOm
Conversion rate - from 1 mcsec to 2 msec / point (it is possible to use an external generator)
Memory length - 4096 words.
Maximal frequency of processed signal (full amplitude):
    for 1.28-10.24 ranges - 400 KHz
    for 0.08-0.64 ranges - 100 KHz
It's possible using with KAS-8 in multichannel systems.
        ADC-850S
        Parameters of digital oscilloscopes C0621 (ADC-850S)
Resolution - 8 bits.
Input ranges - from 0,1V to 8,192V.
Input resistance - 50 Ohm
Maximal frequency of processed signal (full amplitude):
    for 1.28-10.24 ranges - 4 MHz
    for 0.08-0.64 ranges - 400 KHz
Conversion rate - from 50 nsec to 2 msec / point (it is possible to use an external generator)
Memory length - 1024 words.

1985
That year I designed a new version of ZIIS using a modern ICs. A new version was compatible with previous design but it was more simple and cheap.
    Parameters of C0643 (ZIIS-4M)
Resolution - 10 bits.
Accuracy - 0,1%.
Input range - 2046 mV.
Input resistance - 1 KOhm.
Settlement time - 2 mcsec.
Maximal frequency of processed signal (full amplitude) - 20 KHz.
    These years our designers began to use microcomputer Odrenok instead of minicomputer ODRA-1325. That microcomputer provided a fast and convenient access to CAMAC dataway. An user programm spent only 25 mcsec for execution of CAMAC command. An access to remote crate via existing serial link required a few millisecond. To provide a fast access to remote CAMAC crate I designed a pair of devices (DS-24S and CC-24S) and wrote a pair software packages for user programs and a set of test programmes of general purpose (a manual controller emulator, system test and etc). A new software (and developed hardware) allowed to make application programs which was able to work with hardware in different crates (in central crate and in remote crates) without recompilation. We produced hundreds of these devices. Photos of developed devices are shown below.


1986-1992
    That period I was dealing with programming and small jobs. I designed a few VME devices together with my colleges. It is pity but I haven't Photos of these devices. I participated in designing VME processor module based on 1801WM3 chip, 4*RS-232 interface, arbiter module. I designed also memory board a few Mbytes. Then perestrojka came and my colleges went away for best life and this theme was closed. 
    I designed also a pair of programmers. One of them was intended to programm PALs 1556 series (PAL16L8, PAL16Rx). The second device was able to program all russian MOS EPROMs. Photos of these devices are shown below.

 
I was codesigner of programmer 556RT1,2 (82S100, 101). At first, this job was doing by my college Kolya Uwarov. The device wanted to have two PCBs. He asked me to help. I minimized a few schematics, so device didn't require another board. After that my participation in this design was completed.
At the end of this period I designed together with my postgraduated student Andrey Akulov PC-version of my DS-24S. We produced a few boards of this device, but users tryed this CAMAC interface didn't want to change it for widely spreading PPI. The reason was that user received with board a lot of software (library, manual controller emulator, system test, CAMAC dataway test, software for PLD and EPROM programmers). Users of PPI cannot get this software up to now.

1993-1995
    These years I designed electronics for Micro-HPLC System (EnviroChrom, Milichrom-A02). It was quite heavy job. I dealed with 3 board with microprocessors (one PCB contained 2 microprocessors). Embedded software was written in assembler. These programs contains above 5000 lines (32-bits mathematics, logarithmes and etc). Besides embedded software there were written a few test programs in PC to debug each part of all device. This device is prodused now and is selling by a few traders. 
Official parameters (from Econics catalogue) 

    “MiliChrom A-02”

The device is intended for using in stationary, mobile or field laboratories performing analizes for industry and science.

  • A typical sensitivity of anylize:
- in concentration 0,1 - 1 mg/L
- in amount 1 - 10 ng
  • A number of substacies detected in sigle sample may be up to 25-30.
  • Time of analize procedure is 3-30 minutes depending from sample.
A chromatography information processing system includes the following functions: automatic peak labelling, calibration, identification and concentration calculation with print output. The software run on IBM PC with  WINDOWS 3.1 or WINDOWS-95.

Technical parameters:

Consumption  power 130W
Dimensions, mm 535х210х320
Weight (without computer), kg 17 

Detector - dual beam UV spectrophotometer
Detects 8 wavelengths simultaneously

Wavelength range,nm 190 - 360
Cell volume, µL 1,2
Range of measurements, AU -10 -+ 18
Reproducibility, nm 0,01
Time of anylize procedure, s 0,03 - 20

Autosampler:

Samples 46
Vial volume, mcL 100
Injection volume, mcL 1 - 99

Pump- dual syringe , gradient

Volume, mL 2 ґ 2,5
Working flow, mcL/min 3 - 999
Filling flow, mL/min 2
Max. working pressure, MPa 7
Mixing chamber volume mcL 40
Unlimited number of gradient programs  

HPLC-system:

Columns:-size, mm
- column heating, оС
- column heating error, оС
Ж 2ґ 75

35 - 90 
0,3 


1998-1999
    These years I returned to hardware designing. VME interface for CC-24S was designed together with Vlad Shilo. A pair of VME devices were designed together with Vladimir Repkov. They are a VME interface for wire beam sensor and a specialized VME device for wiggler automation. These device aren't general purpose devices and I will not describe them in detail. Device photos are shown below.
 

My latest designs have a special page, look at it. I list them briefly here.


 2000

I have designed the following devices: CANDAC16, CANADC40, интерфейс CAMAC-CANbus(the latest one was designed with my student Alexey Fisenko).

2001

I have designed the CDAC20 device and have created the My Designs section in my WEB-site.

2002

I have designed the following devices: VMEADC16, CEDAC20, CGVI8.

2003

I have designed the following devices: CPKS8, CANIPP (the latest one was designed with Vladimir Repkov).

2004

I have designed the CURVV (with my student Michail Romah) and I have created WEB-site of our laboratory and the Application Notes section in my WEB-site.

 

Dr. Victor R.Kozak,Novosibirsk,  25-jul-2002г.
email: kozak (at) inp.nsk.su