Optron #18 18 сентября 1998

История

Iron - The birth of the modem in Lviv.

(C) Sergey Filimonov, 1998


             BIRTH OF THE MODEM


               Prehistory

             (First Kisses)


  How Lviv appeared the first two C-DOS modem, and how much 
with them was troubles, our readers are aware - in general - 
from an article by Andrei Stegnitskogo in the first issue of 
"Optron. 

  Having ordered from Moscow, where C-DOS
modems, Andrei and I hooked them - and
what? Connect with each other and not
able.

  Had tried various exchanges and communication conditions. In 
the best case skonnektitsya succeeded, but the quality of 
communication left much to be desired ... 

  In the end, we came to the conclusion that
Moscow modems will not fit
for Lviv conditions.



          Experiments with the scheme

               (Defloration)


  What could I do? Money paid, and to
developer purchased modems - far
as the sky.

  To good not to waste, we had to, you want - do not want to 
get in a principled scheme - and suddenly be something in it

improved?

  After analyzing the circuit in front of me has opened a wide 
field of activities. 

  Feature of the Moscow version is
that the line is busy primary winding of the transformer load 
it in the variable component at all times work modem. But the 
resistance of the coil - very small. As a result - is 
extinguished approximately 50% of the signal.


  Further, the linear chain of nodes and transmission
Dialing absolutely identical - and at the
same time, connected to a separate coupler
each, which, in my quite complete, it is redundant. Therefore, 
I first joined to "OR" both channels into one, than saved 
scarce optocoupler and a couple of smaller parts. 

  Then began the long and often unsuccessful experiments to 
improve the reception / transmission. 

  When I tried to reduce the implantation at
transfer (put a resistor in series
primary coil, replaced the transistor,
transformer to the connector line,
another - with a large saturation voltage)
- Degraded reception. And when he wanted to improve
reception, changing this transistor current
Key KR1014KT1 - then became an almost perfect reception, but 
the transfer was quenched by about 80%. 

  Any increase in the amplitude of the signal at all "eaten up" 
transformer. 



            The main revision

                (Conception)


  Conclusion suggests itself - to share
channels control the reception and transmission. So
I did.

  The scheme has been changed so that
transformer primary current key connected to the line only
at the time of admission. For this purpose, digital
part of the modem was introduced by a trigger that handles the 
key through the opto-coupler. 

  This flip-flop - the main distinguishing
sign of the upgraded scheme.

  Thus, the new version, as in
Moscow initially, the remaining three
coupler, but they have to perform perfectly
other functions - as shown in comparative table.


 NN serve in version:
optocoupler Moscow modernized

 Sending a set of numbers and

                     transfer

 2 Dial transformerless

                     class lines

 3 sessions through inclusion transformation
                     mator in line with

                     admission



                Recognition

          (Introduction to the test)


  Having defined the structure, I've posted
using a converted Modem his letter
Moscow distributors, confessing them
in all my deed. They have a personal
"Download" failed. Later, about their reaction, I learned from 
the corresponding FIDOshnoy conference.


  "Have nothing against the modernization and
subsequent spread of your modem,
- I was told. - On the contrary, we only
pleased if ZhNet will join another city. "

  Thus, I have received the coveted
"Carte blanche".

  And done the same dismay this correspondence in
ZxNet-noble family!

  "C-DOS modem connection, set a record for
long distance "- reported FIDOshniki
each other.

  Then Andrew and indicated in the "Optocouplers" on
Soon the birth of the Party C-DOS modems adapted to conditions 
of Lvov. 



                When?

           (A long brooding)


  Time passed. In "Optocouplers" an article
about modems and ZxNet, optimistic messages like "remained 
unsolved literally a couple of questions, "" Work in full 
swing! " "That's it!", Built a far-reaching plans

- And promised there was still no modem.

  - When at last? - Ask me
all and sundry.

  Modernisation schemes turned out, contrary to
initial rosy expectations, the long
and painful process of implementation of various improvements: 
some changes caused the need for others.


  Solutions were different: something
transpired after the analysis of the scheme, something
found in the ZX-press, something else was determined "at 
random". 

  All this and I discussed with Andrew, and rasp. Especially 
helped me with advice and ideas Taras Nikolaychuk (ham

Nadvornaya of his articles about ZxNet you read
in "Optocouplers") - for which I am very grateful to him.



             Other improvements

          (Intensive Care)


  The linear part


  For signal transmission output transistor has been replaced 
by composite included in the scheme of Darlington - because

now increase as not to impede!

  To improve reception was raised
(Instead of the transistor KT940) above
Key KR1014KT1, having very small resistance to the transition 
to the open position. 

  Then had to solve the problem of selection
load for a specific line. Not very
wanted to put the pot: in line -
large currents, therefore, potentiometer
need a fairly powerful, and it will lead to
increase in size modem.

  - What maeshsya? Pick up each time the desired fixed 
resistors - and all! - advised me to Taras.


  After listening to him, I'm in the load circuit line
put a few resistors with different
resistances and added a small switching field. This allowed 
using Jumper to choose a combination of resistors,

optimal for the load.


  Bell of


  First of all, linear and bell parts
were electrically isolated. Thereby
digital and linear "land" added
third - bell.

  Further, it appeared that used in the Moscow version of the 
chain ("+" line - Zener at 90-100 V - LED optocoupler - "-" 
Line) when working with many of Lviv ATS makes no sense: at the 
moment call on the line disappears constant component, and 
there is only variable. 

  So that the bell had to completely redo a part.

  Was taken as a basis for the corresponding node
telephone ELETON, commercially available in the Ivano-Frankivsk.

  In contrast, in a modernized
Modem current load is not "melodic" chip and LED coupler.


  Pairing with the computer


  In the Moscow version of the computer was asking
address of the modem is only one discharge address bus - A4. 
Because of this, there were conflicts with other ports.


  - It sometimes happens - listen Mouzon -
Andrew told me. - Suddenly I hear -
phone jingled. Shoot up - and
it turns out, the modem is driving music
computer in the line!

  Even before I was a general modem connector
removed the wire line and led them through a single connector - 
the way it should be. So that the open contacts on the 
connector was missing. 

  "Decoding" at LL1 has been replaced by this decoder to ID7, 
to the door of which, except A4, were also connected to A0, A1, 
A2, A7. 

  I left the filter, which was only added a potentiometer in a 
circuit of constant displacement. 



        Constructive realization

              (Heavy labor)


  Left, it would seem a little - total
only breed, produce, assemble and
debug a new board modem.

  Razodka was made in the bureau LPI. (This
glorious institution, the walls of which
born on the first home
Spectrum, commented on the pages
"OPTRON).

  The first layout designer (as it always is, if not the 
divorce) has made as it was convenient to him. As a result of 
"land" were not separated, and "blended in a common coil. A 
because this could cause not only vzaimonavodki and 
vzaimoproboy and - because the potential between digital and 
linear "land" could reach 400 or more volts.


  Had to develop P-CAD. (As it happened, and what, ultimately, 
resulted - quite another story.) 

  When wiring board size was
(Compared with the Moscow version) is somewhat reduced.

  Then, in the wiring to make further adjustments.

  As boards were fabricated, as extracted components for them, 
as they are mounted - also another story. 

  As a result, the light appeared Modem
Though completely in software and hardware compatible with its 
"ancestor", but with markedly different from it "stuffing."




               Finally!

             (The first U-aa! ")


  Here it is - a fully assembled and debugged final version of 
the new C-DOS modem! 

  It is time to final inspection.
I called Rasp - send him an article for
"OPTRON" (the same one on "FunTop'98", you
read the previous issue), with a sinking
include heart, and ...

  - Do not be! - I can not help me
from involuntary ejaculation.

  His first premise - the number, the second - the name
Filename: Communications - perfect!

  The modem was tested on different exchanges. Communications,
Basically, a stable and reliable. (Not
talking about the special "humpback" ATS: they are not
HAYES-working and modems).

  Now, finally, it is possible to calm
soul release my baby "on a large
the road. "


                  Contact author's address:

                         serge@ipg.lviv.ua