ATM-Turbo and everything, everything, everything - and a little history of the Profi computer

ATM-Turbo and everything, everything, everything
 by Alone Coder
 (this article was already published in Russian, but since then
new sources were found, which made it possible to significantly
 add article)

During the Soviet microcomputer revolution, mass
was provided, of course, not by radio amateurs, but by factories
(sometimes by the military - for example, documentation for the Impulse computer
has decimal number 3.098.018 D15 with stamps and sheet
registration of changes),as well as various companies.

One such firm was the manufacturing division
"ATM-COMPUTER" of the Moscow company"ATM" ("Association of Creative 
youth"), according to some sources - later namedJSC "ATM" 
("Association of Technology and Microelectronics"). They were located in
VDNKh and the House of Culture andTechniques "Creator". In 1990 
("Radio" 11-1990 p.30) they sold books, circuits, flashed ROM
- and produced a computer compatible with the ZX Spectrum 128, with
disk drive and printer port. Moreover, they specifically indicated that there
128K RAM, probably because The Hobbit was advertised as 128K
(RAM+ROM). This computer is easily identified as "Pentagon 128"
aka "Pentagon 2+", aka "ATM 128"(MULTIFUNCTIONAL
ATM-TURBO COMPUTER. - ATM-PRESS, 1992 p.4). Then the company"ATM"
mentioned it in advertising as her development("Radio" 9-1991,
p.75). The Pentagon 128 board with the marking"ATM 1991" is also known:
http://sblive.narod.ru/ZX-Spectrum/Pentagon128k/Pentagon128k.htm

Of course, a private shop could not reach the mass
circulation of factory computers, especially with the advent of clone
Spectrum ULA - TZChVG1. Their number by the end of the era can be safely
estimated at a million, more than the official circulation of BK-0010, DVK and
UK-NC - and it was a million 48K game cassettes. But the task 
there was no circulation. The guys wanted to sell Spectrumhow
replacing the IBM PC, targeting the "serious" minority
consumer.

ATM's main competitor in Moscow was - no, not IBM - but a branch
Russian-Spanish-Lebanese joint venture"Kramis",
named"Pro"("Radio" 8-1992, p. 63, 8-1994). In more
later advertising and materials feature another name:LLP
"Condor" or"Condor (Kramis)", with a different address and telephone number. And 
this new address was "Youth Library 214". Compare with
company address"ATM" or, for example, with the phone number specified in the demo
for C-64 "Meaning of Death"(1993) - they were located in
organization "Moscow City Palace of Children's (Youth)
Creativity." 

The investigation of these names is like a detective story."Pros", throughout
apparently, was an unregistered trademarkSP
"Kramis" (which was engaged in woodworking and production
wallpaper, notcomputers), and"Condor" could be the name of the same
persons after registration. The only"Condor" whose traces
are on the network - this isCondor LLP, headed by Tadeusz
Radusz.Tadeusz Radusz is a famous Polish publisher who 
started in the computer business(see his interview in the newspaper
Wyborcza http://wyborcza.pl/1,75248,140079.html ). This"Condor"
no longer exists - its address and phone number are occupied by political
organization"Black Hundred" (its leader Alexander Shtilmark
- wife's brotherRadusha). If Radyush was the same entrepreneur,
who created "Profi", then first of all it is worth remembering him
work in the Soviet-Polish magazine "Computer"(1990-1991).

! ----- quote from FIDONET.HISTORY FAQ -----
!There was a family there: Polish husband Tadeusz Radusz [as he was published in
!magazine in Russian transcription - Stanislav Efimov,
!2:5020/525.1024] and his Russian wife Lena Radyush. Actually the node and
!Elena was engaged in BBS.
!All this equipment stood at their house in the area
!m. "Krasnopresnenskaya".
!And there was also - and maybe there still is - Soviet-Polish
!The magazine "Computer", published here despite the fact that the layouts
!made in Poland.
!Elena collaborated with this magazine. Russian articles - approximately
!half of the log volume - transferred from here to there by mailer
!direct calls, and the already assembled options were downloaded back.
!Well, the BBS was sort of like a computer magazine for good measure
!all the same. The whole thing was running on 386/25, modem 2400/MNP,
!BT 2.20/2.30 and RA 0.04. Tadeusz himself, although he appeared in
!nodelist and in screensavers, I did not approach the node, but was engaged
!commerce.
! ----- end of quote -----

This node was the first Moscow FidoNet node.
In the magazine "Computer"Radyush once published an article about
Betadisk. But there was no advertising for "Profi". There were rumors that 
Radyush brought the Beta 128 disk interface circuit to the USSR, but, in
in any case, she was first cloned at the beginning of 1988 completely
other people -Sergey Patsyuk and Vyacheslav Bogomyatov from NTK Plus
(Moscow), seems to be according to the scheme from an unknown Czech magazine.
It was Beta 128 that became the standard for"serious" clones in the USSR,
although he was practically unknown abroad - no traces of him.

Ahead"ATM" in terms of terms by six months, "Kramis"/"Profi"/"Condor"
introduced the Profi computer in the spring of 1991(we use the data
from MikroArt, because you can’t find “Profi” advertisements from those times
successful) - perhaps the first serial clone of the Spectrum co
built-in music coprocessor AY-8910/12, as well as
with 256-512 kilobytes of RAM, high-resolution graphics and
turbo mode. Also in 1991, the Brest computer "Byte-01"
also had high-resolution graphics and a disk drive, and in 1992 already
CP/M(copyright costs 1990, but “Byte” does not have CP/M in the table 
"Radio" 7-92 p.26). In St. Petersburg,Zonova was being developed at this time
"Scorpio" with 256K RAM and(further influenced by "The Hobbit" - see. 
interview in ZX-Format #2 and #6) with a shadow service monitor. And everywhere
a printer connection was provided.

It should be noted here that incorrect dating of the issue is common.
"Scorpio"because of the interviewZonova in ZX-Format #2. Zonov couldn't 
divide the circuit in the summer of 1991. If we collect all the sources
put together, it turns out that he showed the Scorpio prototype at
modified Leningrad board in October 1991(when it came out
book by Larchenko and Rodionov "ZX Spectrum for users and
programmers - 10.16.91). Coincidentally, the first Russian
musical demonstration "MEGA SOUND 128" just came out
10.10.91, and its authorAnton Vasiliev worked for Zonov - made
for him the Song in Lines patch in 1993. In the ROM of "Scorpio" there is
copyright 1992-1994, in Shadow Monitor 1993-1994, in CP/M for
"Scorpio" - 1992-1994, and the first advertisement for "Scorpio" 
published 15.8.93("new development" in the book "System
programs for ZX Spectrum" published by "Peter", in this advertisement
CP/M is already mentioned). In addition, one of the authors of iS-DOS,Alexey
Leontyev (also from St. Petersburg) believed that "Scorpio"
came out later "ZX-Next 512" -see. "Open Letters", 28.7.99:

"I was very happy for you when I learned that
Your ZXNEXT has as much as 512 kB of RAM. Also
you can be happy for many others
owners of this computer.   File
nextS12.hlp is dated back to 10/11/93, i.e.
Muscovites received a half-megabyte machine
almost earlier than they did it in St. Petersburg
256-kilobyte."

ATM could not do less. And the necessary computer "ATM-Turbo" was
developed for"ATM" by the creative team "MicroArt", based on
their version of the amateur radio "Pentagon" - mentioned above
"Pentagon 128". The original "Pentagon" they are, of course, not 
developed, its author(according to Konstantin Sviridov) -
ham radioVladimirDrozdovRAZAO. "Pentagon 128" they too
continued to produce, and the company"Profi" also in parallel
produced 128K computer "Nadezhda"("Radio" 8-94).

"ATM" also collaborated with the company"INTER-LINK", which 
produced connectors and television adapters("Radio" 8-1992
p.63).

The ATM-Turbo computer, the result of this joint development, was
launched on the market in the same 1991 under the name "ATM-SR"
("Radio" 9-1991 p. 75).

It was only laterMicroArt separated and produced its own computers
itself, under the names (ATM)Turbo 2 and Turbo 2(+) - differences from
ATM brand ATM-Turbo 2 are insignificant. Disagreements in
were originally authored: the ATM 4.20 circuit has the sign(c)
MICROART, but 4.50 and 6.2A (in some album) are not. Of course
The ATM-Turbo 2 circuit was developed byMicroArt.

MicroArt also produced modems andROM programmers, published
literature, sold IBM PC-compatible computers, and today
switched to the production of electrical equipment.

                             * * *

What exactly did they do? For starters, memory.

128K RAM in the Pentagon was quite mechanically replaced by 512K
(on RU7). But the issue of addressing was acute. Programs could
climb to all, even crazy, port numbers, could address
port#FD, and even with non-standard bits in it, and any
a non-working program could cause an outburst of indignation among
"serious" client. Manufacturers chose each number
port, ATM-Turbo also had its own.

But they still didn’t guess - in 1992, in “ATM-Turbo 2”, port
had to be hidden in the area of TR-DOS ports (in 2+ they also expanded
memory up to megabyte). Only after such hiding
compatibility has become more or less complete. At the same time we made three more
port for switching pages in other windows. True, for this
I immediately had to redo everything specially written for
computer programs.

Ideas for a rapidly growing business were in the air. Everyone
the manufacturer thought that programmers would support it
option, but everyone has reserved compatibility with
ZX Spectrum 128K, despite the fact that there were still programs for it 
not enough. Each manufacturer at firstwas ready
sponsor the development of programs for yourself - like the sameZones.
"Condor" and"MicroArt" had their own staff of programmers. But 
an operating system was needed.

While the commercial superproject iS-DOS has not yet entered the market,
there was only one answer and no options - CP/M.
This operating system is standard for the Z80, it had a mountain
software, and among Spectrums it was successfully used for many years on Timex
Sinclair 2048/2068, ZX Spectrum +3, as well as in Polish and 
Romanian clones, in Czech devices fromLEC and Sinsoft, on Aerco
FD-68 Floppy Disc Drive and who knows where else. 

WhileAndrey Larchenko adapted CP/M for Scorpio, and Kamil
Karimov - for Quorum, Profi and ATM-Turbo (as well as Belarusian
"Byte") presented ready-made CP/M solutions. In case of ATM-Turbo 
- even in ROM and with markings from TR-DOS.

Judging by the lack of mention in the advertisement, CP/M for ATM-Turbo is not
was ready in 1991, like CP/M for Profi.

It’s interesting that when developing their version of CP/MMikroArtovtsy
we used the documentation for the CM1800 computer from 1979 :)

And especially for CP/M, advanced graphics helped.Unlike Timex, Profi and Byte, ATM-Turbo decided
apply permissions in extended graphics that correspond to
IBM PC:320x200 and 640x200 (including because the standard for 
CP/M - 80 characters per line). And unlike them, already with 
The first versions of the high-resolution graphics were colorized.
At the same time, a second additional mode was made - 16 colors per
point. His absence was blamed on Profi in 1993
parody Best 0 Super Demo (as a response to the Profi advertisement,
which the company gradually carried outTHD), but it was too late -
"Condor" only colored its high-resolution graphics and did not
was ready to invest further in development - switched to
resale of "Amiga"(see X-Ray #1). A"MicroArt" took a risk - and
added text mode to Turbo 2. It turned out that this
it’s not complicated, and everything didn’t go beyond one board (Profi has two).
ATM-Turbo was installed in the Mikrosha PC case. (Scorpio - in 
Corvette body. Byte and Profi had their own cases.)

How is the color per point mode made? Very simple.
1. We implement hardware multicolor (just differently
multiplexing addresses). 
2. Raise itfrequencies doubled (the authors decided to do this
two lines of RAM, so as not to freeze the processor) and 
We use another address line. We get the hires mode. 
3. Instead of reading into attribute and raster registers, we read only in
attribute register. We switch the meander into the raster. We get the same thing 
bit arrangement in byte: %RLRRRLLL. Subsequently color scheme 
The point for the Pentagon was done the same way. 
4. For a snack, unbind the high bits of the address counter and change
the logic for forming the screen frame to get 320x200. The right one 
logic was driven into the most primitive PLM, like Scorpio (in Profi 
and ZX-Next this role was played by a ROM chip, if you don’t go into 
details). This was still necessary to save cases 
when correcting the scan (exotic Pentagon screen, child 
saving on microcircuits was a sore point at that time, 
discussed in ZX-Review) and at the same time protect the board from copying. 
(It was only later that all the documentation was transferred to the public 
property.) A funny feature of this logic turned out to be 
undocumented possibility of hardware scrolling by 
vertical (see Catdemo and Info Guide #10), but it cannot be achieved 
easier than on Atari ST :)[http://www.pouet.net/topic.php?which=6844&page=1]

The graphics output included a palette scheme and a SECAM encoder. This encoder
was, apparently, from the companyINTER-LINK, and subsequently his
removed.
The palette is organized as an external device - cell number and
rendering, and recording is determined by what is coming from the computer
RGBI signal, and the data for the cell is taken from the data bus
processor when accessing the port. Palette scheme for 4096 colors
fromDDp(originally an add-on for Pentagon 102ХSL 2.x
http://realddp.narod.ru/zx/palЧЧ4/index.htm, then included
Pentagon 2.6bbLE) is made in the same way, only additional data
are also taken from the most significant bits of the address bus (with a short
addressing they correspond to the data bus, so the circuit
compatible). But that's in the future. And here only 64 were sold
colors like on EGA. This is all for converting from PC
games.

As with Profi, the first games for ATM-Turbo were copies of games
from IBM PC.
But without stopping at Color Lines and sappers,MicroArt took
high bar - ported Prince of Persia andGobliiins. Ports
were so accurate that even the passwords matched :)
In addition toMicroArt, she also wrote commercial games for ATM-Turbo
companyMacroSystems Laboratory, and maybe someone else.
Commercial development ceased around 1996, but since
the emergence in the 21st century of new incarnations of ATM-Turbo (first
queue fromNedoPC ) the number of games with ATM features has reached
up to about 40, and a total of about 300 programs:
http://alonecoder.nedopc.com/atmsoft.txt
This is excluding programs that only support DACs.

Games, in addition to graphics, also needed sound.

Music coprocessor AY-8910/12 was the only imported one
part that needed to be installed on the ATM-Turbo board and for
for which there was no replacement. (Expensive imported ROM could be
replace with RF2 and load the operating system from disk or even
from tape.) But the music coprocessor could not be installed.
In addition to it, there was also an 8-bit DAC on the printer port (the same#FB
named ZX Lprint), digital music could be played through it
from the IBM PC - such a program was sold under the name "Scream
Tracker Player" or simply "COVOX". For Profi this was also the case, 
and Profi decidedsupport stereo sound due to
non-standard ports.

ATM-Turbo initially contained a modem based on this DAC (with 
support primarily in the Caller ID-Secretary program), but
Later it turned out that it was better to just leave the DAC and ADC, and
modem sold separately.

And a sea of digital music had to be stored, for which purpose was needed
hard drive.

ATM-Turbo 2 contained, apparently, the first IDE controller 
for Spectrum in history, this time ahead of Profi.
It was not cut down in any way (for MFM disks, or for the IDE half)
- it was a full-fledged IDE interface.
You could also connect a CD-ROM to it, but in those days it was
very expensive (when it became cheap, they started playing on ATM-Turbo
films from it - not bad for equipment based on Soviet elements
:)).
Hard drive support was provided in CP/M itself (that is,
directly into ROM), which made the computer even closer in capabilities to
The IBM PC, which was rapidly falling in price and ruining business. 
(SubsequentlyMikroArtu, like its competitors in the workshop, also
had to trade them...)

The first advertisement for ATM-Turbo 2 was published in Radio 10-1992
p.63(that is, the circuit was not developed in 1993, as was previously thought, 
and about the break of "MicroArt" with "ATM" in 1992 is written in the text of
program"Honey Commander") Unfortunately, not mentioned in the advertisement
IDE controller, so there are suspicions that it started working later,
when the diagram was published(this is also hinted at by the fact that
incomplete ATM 6.00 diagram in this part contains designations
elements N11,N12 in place of D32,D34 in future schemes). If
suspicions are correct, then the first HDD controller for the ZX Spectrum
turns out to be the controllerNikolay Tyrsin(1992, see ZX-Format
#5/iS-DOS/IDE HDD). But nevertheless, the controller in ATM-Turbo 2
was the first IDE controller for the ZX Spectrum.

By the way, in the meantime, the production of the famous
58-key film keyboard without letterY (who
developed?) and the time has come to switch to IBM consumer goods.
And for this we needed a keyboard controller.

ATM-Turbo 2+ was equipped with a 181bVEZ1 microcontroller with a witty
firmware that implemented the serial port and
XT keyboard support. It was supposed to connect to
serial port and a mouse, but there are no programs for it yet
discovered.But the mentionedKamil Karimov was able to do it in 2005
connect a modem there and ping the Internet :)

By the way, the first versions of ATM-Turbo had access to a toggle switch
turbo mode, which could be installed in the Mikroshi case
or in an old IBM PC case. But the cases are from Mikroshi, apparently
became a rarity, and on new PC cases the button is already
disappeared, so the toggle switch was replaced with a port - fortunately the ports are now
were hidden. With these hidden ports at the same time it disappeared
the need to introduce a “port blocking button” (which
Profi had it), there was also nowhere to install it.

There was no way to have an edge system connector, like on the Profi.
install in an IBM PC case. ATM-Turbo did not contain this
connector, only a couple of ports were provided for connection
programmer As a result, after the appearance of General Sound, his
I had to hang it on wires. Such a copy is available at
Maxagora.Maxagora also found out that Zorel (which already 
developed a new ATM-Turbo board) wants in the next version
install the system connector on the board as on Scorpio (ZX-BUS).
On a new element base (with FPGA emulation) such implementations
ATM-Turbo already exists - ZX Evo and Pentagon 2.6bbLE. In computers on 
Making an FPGA ATM-Turbo 2+ from its ancestor - the Pentagon - is easy 
change 200 lines in the program.