My Speccy #01 02 февраля 2004

История

With the world on a thread - a story Spectrum clone ATM Turbo.

 There is such a clone ...: 



As you know, after the establishment of model ZX-Spectrum 128 
process sinklerostroeniya passed into the hands 

Masters of single and small firmochek (at least in ex-USSR). 
Most of them 

not simply repeating ZX-128, but sought to make their 
innovations in order to increase the chances of 

Spectrum in a given area. And since these firmeshki practically 
does not relate to each 

other, then did it each in his own way, as the head vzbredet. 
As a result, everything in the Soup 

and bells and whistles, including the same purpose, was formed 
as standards that 

programmers grabbed his head at the prospect of writing 
software for them. Fortunately, most 

of these standards did not survive, but they have only a small 
amount of what many 

facilitated by the development of the Internet and reducing the 
number of spektrumistov in general, when the remaining 

their small number could be fairly rapid exchange of 
information about all the innovations and 

align program standards for them. But as a result of 
bacchanalia standards of previous years 

still managed to form several stable branches of the Spectrum, 
a common root which 

originates from the ZX-Spectrum 128, and then goes its own 
unique set of bells and whistles, very 

often not compatible with a set of bells and whistles of the 
other branch. Each branch is represented by its 

independent clone of the Spectrum, as Profi, Scorpion, KAY et 
al. On one of these clones, and 

ATM is about and will be discussed. 




Why ATM? For several reasons. The most important of them is 
that I am a 

this clone, and therefore, better than anyone else, and studied 
his features, and is prone to say 

it is about him. But there is another reason, no less 
important. It lies in the fact that ATM-turbo 

is one of the least documented of the Clones, and this lack of 
information often leads to 

that its standards are not supported in modern software, or not 
fully supported. 

Just over a year ago to correct this situation by me (Timonin 
Maxim aka Maksagor) was 

created a site specifically devoted to this branch of the 
Spectrum-compatible machines -  
http://atmturbo.narod.ru , 

which has already begun to bear fruit. Therefore, if anyone 
needs the technical details, so please 

on my page! And here I would like to briefly discuss the 
history of the creation of ATM-branch and its main differences 

from other Spectrum. 




It all began in the late already distant 80's, when the company 
ATM together with the creative team MikroART started working 
together on the creation of Spectrum in the domestic element 
base. The first computers from them were called "Pentagon" 
(yes, that!) Because of the special form of wiring diagrams. At 
first it was a Pentagon-48. Then - the Pentagon-128 different 
versions - (version 1, 2, 2 +, 3, and maybe even what). The 
difference between these versions, though I kill, I do not 
know. Apparently, the software point of view, no difference, 
but it all comes down to improving and optimizing the circuit 
and removed from her mistakes. Although, according to 
unconfirmed rumors, the very first version of the Pentagon in 
1948 and 128 were non-TR-DOS and AY-coprocessor. But I have no 
confirmation. Well, be that as it may, the Pentagon in its 
various versions saw the light. Here we must recall that it 
represents during the late 80's - early 90's. And it was, on 
the one hand, the "golden" time of the Spectrum in Russia in 
general, but on the other hand - rapid development of all 
cooperatives, and firmochek firmeshek, each in his own way 
worked radiodelom and, of course, the Spectrum. And while they 
have the vaguest idea of ​​any out there, copyright, copyright 
and other like Biden, and sometimes vague idea was not 
available. And what it led to: the Pentagon's scheme was 
relatively simple, designed for the domestic element base, and 
it was very easy to repeat. What to and enjoyed all of these 
companies and cooperatives, replicated fee without compunction, 
and even without mentioning the ATM and MikroARTa throughout 
the former Soviet Union. And the scheme proved so successful 
that the replicability she left behind all other options 
ZX-128.Konechno, and it was not without its flaws with the 
technical side of the issue. Namely, the main drawback is the 
non-standard INT and incomplete decoding port # xxFD. But these 
defects are soon turned into its exact opposite, and it is 
thanks to the wide dissemination of the Pentagon, very soon, 
which has become the de facto standard Spectrum-128 on 
post-Soviet expanse. Non-standard compared with the firm Speccy 
INT allowed in one interrupt accommodate slightly more strokes 
than were quick to take advantage of the programmers in their 
time on critical programs and demos, after which their programs 
were normally only at the Pentagon, and on machines with 
proprietary INT inhibited or Weasley. Features incomplete 
address allowed to choose the port of pages # 7FFD, using only 
the younger half of the port through the command OUT (# FD), A, 
which saves a few cycles, and also had a significant importance 
in time-critical programs, however, again, often resulting in 
unusable these programs on machines with more competent 
addressed. 
 



But all of the above turned out not once, especially as the 
Pentagon's non-clones in the total number of machines initially 
ninetieth was not very much. The creative team also MikroART 
ATM with the firm decided to continue to develop their project 
"spektrumostroeniya" and more. What was not surprising, because 
in those tumultuous years, the popularity of Speccy all wanted 
to improve it further, to keep up with the "big" cars, the good 
and the gap between them and the Spectrum was still very small. 
However, experience has lost because of pirates pilferers, 
publicized scheme Pentagon across the country, profits were 
taken into account. And on the next development it was decided 
to protect. This protection has firmware programmable logic 
array PLA 1556HL8 on which the main part of the video 
controller, and formed some of the other system signals. 
Firmware is kept in the strictest confidence. In 1991, the new 
ZX-development of ATM and MikroART with the PLA on the board, 
saw the light. Thus was born the new clone ATM-turbo. It took 
into account the shortcomings of the past the Pentagon, as it 
seemed then, and it made a standard INT and implemented a more 
complete decoding # xxFD, and a subset of these ports has been 
extended - a decoding was added to address A9, which addressed 
several new ports for the sample of new bells and whistles. And 
there were many. Namely - were implemented two new graphics 
mode. This is a hardware multicolor (attribute in bytes) at a 
resolution of 640x200 and EGA-screen 320h200 every point of its 
color, 16 colors. In all modes, you can customize the color 
palette, with the ability to display 16 colors simultaneously 
from all 64. True, the extended screens began to occupy each of 
32,000 bytes, which caused the brakes to work with them. 
Therefore, the board has been implemented as turbo and 7.0 MHz, 
includes hardware at the touch of a button. True, and it helped 
not fully, but still operate in this mode, the screens become 
easier. More of the bells and whistles have been implemented 
ADC and COVOX, on which there is on the board was reconstituted 
primitive modem, but without electrical isolation, interface is 
implemented CENTRONIX (LPT) for the printer. Also, on the basis 
of this interface was divorced ryadm other special plug 
external devices can be connected to their own development, and 
under that was designed programmer UNIPROG (produced and today, 
but used on PCs, although the possibility remains connected to 
the ZX), and later a special HAYES-compatible can at 1200 baud. 
Memory, too, has undergone changes: the RAM can be left 
unchanged - 128K or 512K to expand, replacing the chip on 
565RU5 565RU7. ROM has been expanded to 128K in chips 271,000. 
One half of it to be used for actual system (BASIC, TR-DOS, 
etc.), and the second to realize the ROM-disk, which would have 
pierced the files of various utilities and used for quick 
access. However, this idea was not communicated to the end and 
really in the ATM-turbo1 used PZUshka 27512 with 64KB of 
memory, which sewed the standard firmware without the ROM-disk. 
But was implemented option to disable ROM and inclusion instead 
of zero page RAM, and at the same time (and only once!) At # 
4000 instead of the inclusion of the fifth page, ZX-screen - 
the fourth without a screen. And this was done on purpose. As 
I've said in the early 90's still very much cherished hopes for 
a more serious than a hobby, the use of the Spectrum. For this 
seemed to be just a little bit - add a bit of speed, graphics, 
memory and peripherals to expand and write more professional 
software (databases, spreadsheets, editors, communications 
software, etc.). Something, of course, you could write on their 
own (and, looking ahead, saying that it was written), but it 
will take time, and the computer must be ready and fully 
staffed as soon as possible for obvious commercial and other 
reasons. Yes, and it was already clear that, in future, on the 
primitive TR-DOS is very much does not go far if guided not 
only toys, but for professional use. And so the question of 
software developers have gone the way of borrowing, namely, 
adapt, and ATM-turbo universal, in those days, for eight-bit 
computers based on processors i8080 and Z80 Professional 
operating system CP / M v2.2, writing under her own BIOS by 
implementing the Extended screens, which allowed to realize the 
output of 80 characters per line. This system requires a lower 
16K (# 0000 - # 3FFF) is RAM and ROM can be switched off 
immediately necessary just by the way, but rather was conceived 
precisely for this system. Well, as it were, and from borrowing 
CP / M can now use a lot of professional software written at 
the time under this OS. In addition, it was decided not to stop 
there, and under this version of CP / M was organized by 
writing your own programs (primarily working with graphics, 
COVOX, etc.) and games (for example, were transferred to a 
PC-to-one and PRINCE GOBLINS, operating in a 320h200h16). Was 
released a few versions of BIOS CP / M for ATM, as far as 
catching bugs and optimization. The latest version of BIOS for 
the ATM-turbo was 1.03, continue to do even for ATM-turbo 2 
(+), but more on that later. Not bypassed small changes and 
standard ZX128-firmware (bas48 + bas128 + TR-DOS). In free 
field TR-DOS was posted part of the BIOS, and rather colorful 
start menu with options, which is loaded into RAM when you 
start work or on the button RESET. Incidentally, in recent 
versions of ROM RESET button has also not been overlooked. In 
the ROM BIOS on the zero address and the TR-DOS program was put 
the interceptor on the basis of which was implemented by a 
resident: he was hiding in the memory above 128K and RESET or 
zeroing TR-DOS interceptor on a special header checking the 
presence of this resident, and if it confirms a software 
control passed to this resident program. This feature is very 
useful in many situations, but in practice, in TR-DOS used it 
HONEY-commander, also written by request MikroART. He remained 
at the top when you run out of his program, and on or RESET 
output to zero address TR-DOS, we immediately returned back 
into the shell. 
 



Thus, having lots of iron and program bloat, in 1991, ATM-turbo 
1 

(Or, as it was called then -

ATM-turbo 512k) saw the light. Although his scheme was 
"genetically" a continuation and development schemes Pentagon 

(Experienced eye, looking at the concept, finds much in 
common), yet innovations were 

so numerous that ATM was an independent clone, retaining, 
however, the numbering of the versions to pay even with 

Pentagon. ATM-1 fees were 4.xx version numbers, apparently, 
version 1, 2 and 3 were the most by the Pentagon 

2 and the Pentagon Pentagon 3. Although, maybe I am wrong. 
Well, it's not crucial. As 

use ATM-1, on its board detected minor glitches and 
shortcomings that were eliminated in 

future versions. So see the light board with the numbers 4.10, 
4.20, 4.40. Final basic version 

was 4.50, on which the development of iron in the ATM-1 has 
been completed. True, later was released on payment 

5.20, but it did not differ from 4.50, except that on one board 
and the keyboard was located, and 

were also separated all the connectors, so that did not have to 
solder them manually on the computer housings. 
 



However, to repeat the success of the Pentagon ATMke was not 
meant to be. This happened because, thanks to 

Protection based on PLA MikroARTu and ATM have managed to 
prevent the copying fee pirates. This helped 

keep company profits, but hindered the "people" spread clone 
across this vast country. AND 

Although MikroART at the time of its peak had branches in 
several major cities in Russia 

For example, in Novosibirsk, catch up with the Pentagon for the 
mass was impossible, so often, and others 

clones that did not have such powerful protection. Therefore, 
ATM and its subsequent modifications were 

"Thing in itself", and many users, getting ATM is not in the 
firm, and will buy from someone with hands 

without documentation, are often not aware of the many hidden 
features and bells and whistles of this clone. To 

eliminate hunger for information was a portal  
http://atmturbo.narod.ru , and even this article 

written with the same purpose. But we have strayed from the 
topic. Continue. 
 



As sales of ATM and its getting used in a variety of other 
clones ZX, began to identify not only the dignity, of whom we 
have mentioned, but also the shortcomings of this model. And 
these flaws manifested themselves from an unexpected quarter, 
and were associated with the previous MikroART brainchild - the 
Pentagon. An early disadvantage. Not associated with errors: 
the ATM was not implemented Kempston joystick, as the authors 
honestly all warned, referring to the space savings on the 
board (of which there to save?). Apparently, something they did 
not like it, just not implemented it in the ATM-2 (+). And 
users had to solder on its second floor. And now the more 
serious shortcomings. As already mentioned, it is because of 
its prevalence, has become the standard, and therefore become 
the standard all its defects and abnormalities, such as INT or 
incomplete addressing # xxFD. And unexpectedly, developers who 
sought to approximate the characteristics of the ATM to 
corporate Speccy for the sake of better compatibility with the 
software, faced with an increasingly growing incompatibility 
with new software, began to emerge in large numbers and with 
good quality even in Russia. At the ATM because it violates INT 
Pentagon "floated" the majority of multicolor (a small part of 
the programs due to a smaller number of ticks on INT brake or 
hung). But this was not scared. Another problem was worse. With 
it, users are familiar with all scorpions, and KAY, where as in 
the ATM complemented by addressing # xxFD and introduced a new 
subset of these ports on the base address of A9. It is 
primarily the port of additional memory and the port # FDFD 
palette # 7DFD. Therefore, the use of incomplete address # xxFD 
on the ATM often leads either to freeze the program, or to the 
fact that "flew" palette and everything on the screen, randomly 
painted in unimaginable colors for the standard Spectrum. There 
was another disadvantage: selection-screen mode and disable 
ROM, were hanged at the address lines A7-A5 (through their 
zero) port curb # FE. It would seem that everything is OK, 
well, who will only use part of the so eight-port? It was found 
that use of steel and often, and this led to the inclusion or 
non-standard screens or disables ROM and the fifth page, which 
naturally led to a freeze in all cases. Why if you disable ROM 
- understandable, but when you change screens it was this: 
circuit switching graphics has been implemented in a strange 
way. In the transition from ZX-screen in any extended or 
conversely, the address space to mix up: bits 5,6,7 and 8,9,10 
in the address space are interchanged. It is still unknown why 
the authors have done so. They themselves said that in order to 
facilitate the scheme. But in the ATM-1 to change the graphics 
without entanglement have been only slightly modified scheme 
absolutely not complicating it, and the ATM-2, which later, 
only to replace one of the signals on the inverse, which is 
also present on the board. Apparently overlooked the 
"zamylennym" eye. It happens. There. Hence, and because of the 
graphics prepared iron hanging. Well, the users, of course, 
with this fight, they hung a toggle switch, blocking when 
necessary, all of these expanded ports to Pentagon was soft and 
came up with other ways. On the whole, successful. But the 
developers realized that this is not the case. After all, they 
wanted to make a professional machine, which would not required 
any toggle switch, and could be included without any trouble 
and work ... Moreover, the technology in the world continued to 
evolve, in particular, became cheaper IDE-drives. And in the 
plans of the creative team MikroART were maturing plans of the 
new model - ATM-turbo 2. 
 


ATM-turbo 2 was released in 1993. This machine was also done 
using 

protection of PLA 1556HL8, but radically different from its 
predecessor. If the scheme is still possible to find common 

parts and even the rudiments of the ATM-1, with the software 
point of view, there was only a common mode ZX-128, without 

which, without losing the proud title of the Spectrum, will not 
do ... Many of the decisions and the bells and whistles with 
the ATM-1 remained in 

new models, such as enlarged screen (with the same structure 
and resolution), but change the port to 

Besides, most of them, except, of course, the standard ports 
ZX-128, were hidden in the area of ​​port 

TR-DOS, then there were "shady". Some ports, however, remained 
the same - it's external port connector 

Device # FA, and reading the port A / D - # 7DFD. The truth, 
apparently, the developers responding to reports of problems 

with new ports, not analyzed the cause - the ports were 
removed, and the address decoder in the A9 was left in 

consequence of that problem with the halves # FD remained. But 
this is easily cured simply cutting of 

corresponding track on the decoder, if the ADC was not needed 
or collecting unpretentious circuit on a single 

logical element "and" to the ports with A9 only worked on 
reading, which does not interfere with compatibility. 

By the way, COVOX and the ADC has been improved and implemented 
on the new chips. 
 



But do not just change ports typical ATM-turbo 2.

She also tried to make the most modern in those days. And 
therefore have added new bells and whistles. But something that 
was removed, namely - SECAM-coder, since it is rarely used by 
anyone, the picture gave the bad, and the seats on the board 
held a lot. Now for the bells and whistles: first of all this - 
the controller IDE, allowing to connect the IDE HDD in any 
capacity, and theoretically, and CD-ROM, provided by him in 
writing. Respectively, was modified BIOS CP / M, which was 
added to the hard disk driver, and takes into account changes 
in the ports and architecture, about which later. Also, the 
developers were unhappy with the brakes when working with 
extended screens. Since the work in them occurred mainly CP / 
M, and this axis runs standard in console mode, 99% of the time 
graphic screens are only used for character output, which means 
lots of computer time is spent irrationally. Therefore, on the 
basis of a converted mode 640x200 was introduced yet another - 
a text - a display mode that allows the hardware from a special 
ROM with decoder, output symbols with a resolution 80h25 signs 
colorable in 16 colors and 16 color INK PAPER. As a result, to 
display a single letter on the screen, it was necessary to take 
only one byte (+ one byte of color, if necessary). And the 
screen is occupied 4000 bytes (2000 bytes of the screen itself, 
and 2000 bytes - attributes). So the brakes have been solved 
with text programs. Another on-board controller was divorced 
XT-keyboard. True, the first versions of it were flawed, 
because the controller only remembered obtained SCAN-code, and 
then gave the NMI in the ROM BIOS, where there was a treatment 
program emulated ZX-keyboard. This led to the brakes and 
partial incompatibility with programs that have had little 
place in memory, which may not have had a stack that is used by 
NMI. As well as ATM-1, ATM-2 also improved gradually. Were 
released several versions of the boards: 6.10, 6.15, 6.20, 6.30 
and 6.40. They were caught minor bugs and improved the keyboard 
controller. Simultaneously with the continued opportunity to 
use standard mechanical Spectrum keyboard. , The possibility to 
enable or disable the software TURBO-mode. 
 



But that's not all. Radically changed and the computer 
architecture that 

lets talk about the ATM 1 and ATM-2 as the different clones. 
Principle was applied away from rigidly defined 

paging architecture, which is predetermined by what page where 
included. Where they can be changed, and where - 

no. Was introduced a special memory manager, which allows any 
page of RAM or ROM to put in any 

of the four windows of the address space. In this case, if we 
use standard ports Spectrum-128, the program 

did not notice, as in the mode of ZX-128 controller has 
pre-programmed to emulate the Spectrum 

architecture. Naturally, the controller ports were also hidden 
in the port TR-DOS. ATM-2 could address up to 

512kb RZU and 512K of ROM. As for RAM, then, as in the ATM-1 
could deliver only 128KB on chips 

565RU5 or 512K on chips 565RU7. With regard to ROM, the board 
has been divorced socket for ROM 

with a maximum capacity of 128Kb - 271 000. Although, as in the 
ATM-1, often used only 27,512 with 64K, 

since a long time for larger volume there was no policy 
support. True, unlike the ATM-1 

She eventually appeared, but about it later. Nevertheless, 
despite this "iron" restriction 

Manager can address up to 512K ROM and the user can put more 
volume chip, just 

soldered wiring to connect the respective output controller for 
additional address inputs of ROM chips. 
 



Before we continue talking about the models of ATM-turbo, 
should tell about it: somewhere in 1993, the creative team 
MikroART quarreled with the firm ATM because of the fact that 
they are new, just preparing to enter then sell cards ATM-turbo 
2 does not put copyright MikroARTa and refused to produce 
settlements. After this MikroART refused to cooperate with ATM 
and began to develop their own computer, as they have entitled 
to do so. ATM firm yet released their version of a computer 
with its own version of the firmware PLA 1556HL8 and main ROM, 
where instead of CP / M was laced MSX-DOS. As a result of ATM 
machines and MikroART were only partially compatible. But the 
company has its version of the ATM machine let out a short time 
and in small quantities and not being able to organize, in 
contrast to MikroARTa, massive software support for its 
creation. As a result, the production of these machines have 
been phased out, and soon the name of the company's ATM was 
lost in the abyss of a dashing time. Whatever it was, and 
subsequent history of the clone ATM-turbo connected exclusively 
with the creative team MikroART transforming itself into a 
proper company. For some time he continued to enhance fee 
versions 6.hh even jointly developed with the ATM. As already 
mentioned, small defects are eliminated and optimized some 
sites. Increasingly software was being developed, primarily 
under CP / M. Despite the fact that the ATM-1 and ATM-2 were 
incompatible with each other on ports and architecture, it was 
not a big problem, because many of the programs used standard 
calls CP / M and just did not notice the change. Those few 
system programs, as well as some games that used the access to 
ports and screens directly, were promptly rebuilt. Often turned 
out to be quite a bit to alter the basic COM-file in those 
places where the appeal to the ports. Therefore, these programs 
comes with two executable files: one for ATM-1 and the other 
under the ATM-2. Later, there were games and programs with auto 
computer model. But gradually the support of ATM-1 was stopped, 
and the latest programs and games worked only with ATM-2. A 
software for CP / M by this time accumulated a lot. Of the 
games - it's PRINCE, GOBLINS, BOMB TETRIS, LINES, COLOR FIGHT, 
MINER and lots of others. From the application software, except 
standard utilities CP / M and the same standard databases, 
languages, etc., were designed corresponding text and graphics 
editors, the editor of sprites, player tracker STM-music 
"nortonoobraznaya" shell X-COMANDER and much- much more that 
can be downloaded from  
http://atmturbo.narod.ru/atmload.htm . Improved and 
optimized, and BIOS CP / M. For ATM-2 have been released in 
version 1.06 and 1.07. It has been fully transferred to the 
ATM-1, and continued support of residents. 
 



But plans to radically improve the car did not leave 
MikroARTovtsev. And, somewhere in 1994, this resulted in new 
development. As its title "dropped" the first part of «ATM» and 
the new model was called just «TURBO 2" and later «TURBO 2 +, 
although the habit is still (and in this text, too) this 
machine is called" ATMkoy " , «ATM-2 +" or «ATM-turbo 2 +. 
Version of the new cards were from a subset of 7.hh. The first 
of these was 7.00, and later released the latest version - 
7.10, with minor "cosmetic" corrections. The new model not only 
radically different from the ATM-2, than the ATM-2 differs from 
the ATM-1. Remained virtually complete software compatibility 
between ATM-turbo 2 TURBO 2 +. From the cosmetic changes: the 
memory manager has been finalized, which could now address up 
to 1024KB of RAM and 1024KB ROM. On the motherboard still has 
been divorced socket just under 128K of ROM, but as mentioned, 
it was not a hindrance. Left and the new version BIOS - 1.07.10 
and 07.01.1912. A little later, was built to support the work 
with the hard drive iS-DOS, which could StartUp from the HDD 
when the computer starts. Was somewhere in 1996 went firmware 
ROM 272000 (256k), which differ primarily surround ROM-drive 
full of CP / M-utilities. It is true this happened at sunset 
popularity Speccy, so even in low prevalence, for reasons 
already mentioned, cars from MikroART general. Therefore, this 
"big" firmware has not received wide distribution, and only a 
miracle has helped her do not disappear into obscurity. As for 
RAM, then double that to 1024KB could, soldered the second 
floor an additional 16 chips 565RU7. Incidentally, this 
revision manager was so simple that it can be done at ATM-turbo 
2 versions 6.hh that many did. But there were more fundamental 
differences that are easy on the older models were not 
implemented. First of all, was finally removed from the modem 
board who settled there in the days of ATM-1 and has not used. 
Instead use an external HAYES-modem that connects to the 
connector external devices, like programming UNIPROG. A seat 
has been redesigned support for the ADC, which, with full 
compatibility with the old, became the eight-channel. And also, 
and most importantly, was divorced entirely new universal 
controller IBM-Keyboard and RS-232. Unlike his predecessor, he 
did not use the main memory and not downloaded Z80, because it 
was based on his own OEVM 1816VE31 (i8031) and has its own ROM 
with software support. And it was universal because it could 
support many different types of keyboards. Only need to be sewn 
into its ROM related support. So, initially, the controller 
worked with XT_klaviaturoy. But gradually it became 
increasingly difficult to find on sale, and was written by 
firmware to work with more modern AT-keyboards are sold 
everywhere so far. As already mentioned in passing, this 
controller is embedded in itself, and RS-232 interface, but 
support it had not been done. In 1995-1996, the sale of 
Spectrum, and hence ATMok in the country went sharply on the 
decline MikroART stopped work on the improvement of their 
offspring, and soon turned to the production and sale of what 
was already attained. For some time made the development and 
sale of software, but 1998-1999 years was terminated and she, 
though still firm MikroART provides information support to 
interested users to their offspring via the Internet portal  
http://www.microart.ru . Besides MikroARTovtsy gave the 
people everything that relates to the ATM, including the secret 
earlier firmware PLA 1556HL8. 
 



But the story ATM-turbo is not over. Despite the low prevalence 
of and something which has not eliminated gaps (address A9, 
pereputyvaemost addresses when switching schedules, lack of 
Kempston joystick), which had to remove manually (albeit 
without much difficulty), we can safely say that until the 
appearance of a clone SPRINTER of Peters-Plus, ATM-turbo 2 + 
was the most powerful features ZX-clone. Even today he keeps on 
his second place. Furthermore Sprinter'a he can compete unless 
SCORPION connected with both GS, SMUC and GMX. But only just. 
Why did the story has not ended? Until recently, due to its 
non-proliferation, ATM quietly faded away from the attention of 
other spektrumistov compounded by complete absence of 
descriptions of all its features. But in 2001-2002, was 
initiated steps to "break the information blockade" and unite 
those who are still using these clones. For this purpose, was 
organized by several times mentioned the site  
http://atmturbo.narod.ru , which was declared search 
former and current users of the ATM of all types (of which 
(users) has already gained a few dozen), and the search of 
software that was developed and / or developed so far. It must 
be noted that this undertaking was successful. Range of users 
began to grow, there were experts both in programming and in 
radiodele and gradually ripened idea of ​​creating a new 
generation of ATM - ATM-turbo 3, which would be made, as well 
as Sprinter, the new cell-based modern microcircuits, would the 
ability to connect modern devices and have their support, which 
is planned to develop a new OS, and, of course, would remove 
all the shortcomings the previous model. One could speak at 
length about what is planned to implement, but it is a separate 
big topic of the article. Whatever it was, but the work begun 
by firms ATM and MikroART lives efforts of its users. Whether 
it will continue successfully, you will see a light new 
ATM-turbo 3, time will tell. But I hope that all goes well and 
ZX-SPECTRUM, in the face of one of its clones ATM, will receive 
a new impetus to the development. 
 



Timonin Maxim A. aka MAKSAGOR 


 max_timonin@mail.ru