Atari 8-bit family

Atari built a series of 8-bit home computers based on the MOS Technology 6502 CPU, starting in 1979. Over the next decade several versions of the same basic design would be released, but the models remained largely identical internally. They were, for their era, one of the most technically advanced machines on the market, but a combination of factors, largely business related, meant they did not have a major market when the Commodore 64 shipped a few years later and took over most of the market.

History

Origins

As soon as the Atari 2600 was released the engineering team, calling themselves Cyan, started work on its eventual replacement. They felt that the 2600 would have about a three year lifespan, and tried to limit themselves to those features that could be perfected by that time. What they ended up with was essentially a "corrected" version of the 2600, fixing its more obvious flaws.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

The newer design would be faster than the 2600, have better graphics, and include much better sound hardware. Work on the chips for the new system continued throughout 1978, primarily focusing on the much-improved video hardware known as the CTIA (the 2600 used a chip known as the TIA).

Related Topics:
CTIA - TIA

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

However, at this point, the home computer revolution took off in the form of the Apple II family, Commodore PET and TRS-80. Atari management saw this as a golden opportunity to re-purpose the machines, and started research on what would be needed to produce a workable home computer of their own. This included support for character graphics (something the 2600 didn't support), some form of expansion for peripherals, the BASIC programming language, and a keyboard.

Related Topics:
Apple II family - Commodore PET - TRS-80 - Peripheral - BASIC programming language

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Design

All aspects of the machine were considered open to new solutions, and since the current state of the art was extremely basic, the Atari engineering teams often designed their own replacements to common problems. For instance, they did away with the traditional set of single-purpose I/O ports like RS-232 and the Centronics printer port, and replaced them all with a single daisy-chainable system known as SIO (serial input/output). Of course this also meant the machines were "non-standard", which may have affected their acceptance in the market.

Related Topics:
State of the art - I/O - RS-232 - Centronics printer port

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Cyan's main area of work during this period was the creation of a very powerful character/display driver known as ANTIC. Unlike the 2600 where the video was controlled solely by the TIA moving sprites (known as player/missile graphics in Atari lingo) around a colored background, in the new machines the ANTIC did most of the work drawing the screen (including the ATASCII characters). ANTIC then wrote the resulting "pattern" to RAM, which the GTIA then read, colored and added sprites. This separation of duties allowed both chips to be as powerful as possible in an era of expensive silicon, and the machine's graphics were the best on the market until the release of the C64 in 1982.

Related Topics:
ANTIC - Sprite - ATASCII - RAM - 1982

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

One of the main reasons the 6502 was so widely used in early machines was cost, but the other was a unique feature that made it ideally suited to graphics. During normal processing of instructions, there was a single cycle of every four that it was guaranteed to not be using the bus. During this time the display hardware, in this case the CTIA, would quickly read out the graphics from memory to the screen.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

In the Atari design this was not, of itself, a complete solution. Since the ANTIC and 6502 would both be writing to memory, some system needed to be in place to ensure they both did not do so at the same time. Fairly tight timing constraints limited the modes the ANTIC could support without stepping on the 6502's toes. A more convincing solution was offered later in the form of a custom version of the 6502, originally nicknamed SALLY but later officially known as the 6502C, that could be halted when the ANTIC needed memory access.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Another custom support chip, named POKEY, was responsible for reading the keyboard, generating sound and serial communication (The latter in conjunction with the PIA). The same POKEY chip was also a very common solution for sound effects and music in arcade games in the 1980s, producing a distinctive square wave flavor that is popular among chip tune aficionados.

Related Topics:
POKEY - Arcade game - 1980 - Square wave - Chip tune

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Atari had originally intended to port Microsoft BASIC to the machine, as had most other vendors, intending to supply it on an 8k ROM cartridge. However the existing 6502 version from Microsoft was 12k, and all of Atari's attempts to pare it down failed. Eventually they farmed out the work to a local consulting firm, who appears to have little better luck. Instead they recommended writing their own version from scratch, which was eventually delivered as ATARI BASIC, only weeks before the machines finally shipped.

Related Topics:
Microsoft BASIC - ATARI BASIC

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Based on the 6502B, a faster version of the more common 6502A model, the new design ran at 1.77 MHz (PAL version, European) or 1.79 MHz (NTSC version, US). The odd-sounding speed was a result of the timing, which required it to sync to the video signal at some multiple of a TV's 3.48 MHz (NTSC). Other machines intended to work with TV's had similar speeds, but tended to be much slower; the Apple II and Commodore PET ran at .96 MHz, while the TRS-80 was at 2.03 MHz but was actually about 1/2 of the speed due to its processor, the Zilog Z80's, design.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

The early machines: 400 and 800

Management identified two sweet spots for the new machine, a low-end version known as Candy, and a higher-end machine known as Colleen. The primary difference between the two models was expandability; Colleen would include a number of slots, a second 8k cartridge slot, monitor output and a full keyboard, while Candy used a plastic "membrane keyboard" and didn't include any slots. Both machines were built like tanks with huge internal aluminum shields, a side effect of meeting a FCC specification that was soon to be removed anyway (the first model of the TRS-80 actually never met that FCC spec).

Related Topics:
Sweet spot - Monitor output - Membrane keyboard - FCC - TRS-80

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

The machines were brought to market in February 1979 as the 400 and 800, although they weren't widely available until late 1979. The names originally referred to the amount of memory, 4K RAM in the 400 and 8K in the 800. However by the time they were released the prices on RAM had started to fall, so the machines were instead released with 8K and 16K respectively, making the naming somewhat superfluous.

Related Topics:
1979 - K - RAM

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

For some unknown reason, the 800's slots did not offer all of the pins of the motherboard, as they did on the Apple II for instance. Instead they consisted primarily of the memory and memory signalling pins, allowing it to be used for memory expansion only. Originally the machines shipped with 16k, but as prices continued to fall Atari eventually supplied the machines fully expanded to 48k, using up all the slots. This made them almost useless as a true expansion system.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Nevertheless a few enterprising companies managed to produce expansion cards for the machine by copying data into and out of memory. The simple operating system in the machine made it particularily easy to write device drivers, so this was not as difficult a task as it might have been on other machines. One such system added an 80-column display system and 16k of RAM so you could remove one of the Atari-supplied cards without losing memory.

Related Topics:
Operating system - Device driver

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Nevertheless, even these developers were limited to a small market, as consumers discovered that the two machines were largely identical and bought the lower-priced 400. The same problem meant that only a very small number of "right cartridges" were ever made. In retrospect the design would have been much better off with a single 16k slot instead. This is particularily frustrating given that a 16k slot would have made Microsoft BASIC easy to support, and instead this was supplied on disk for the few users that were willing to pay for it.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Interestingly, the four "joystick" ports on the front of the machines were actually serial I/O ports. For example, one could run a modem off of a joystick port. Ostensibly these ports accepted joysticks, keypads, paddles, drawing pads, etc. {{mn|AtariMuseum|1}}

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

The second generation: XL series

The 800 was rather complex and expensive to build, consisting of a number of circuit boards in various locations inside or outside the massive aluminum shield. Additionally the machine was designed to add RAM only through cards, which required expensive connectors and packaging even though it was now fully expanded right from the factory. At the same time 400 didn't compete technically with some of the newer machines appearing in the early 1980s, which tended to ship with much more RAM and a real keyboard.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

In 1982 Atari started the Sweet 16 project to address these issues. The result was an upgraded set of machines otherwise similar to the 400 and 800, but much cheaper to produce. Newer fabs allowed a number of chips in the original systems to be condensed into one. For comparison the original 800 used seven separate circuit boards (many of them small), while the new machines used only one. Sweet 16 also addressed problems with the 800 by adding a new expansion chassis as well, although it was to be external. Like the earlier machines, the Sweet 16 was intended to be released in two versions as the 1000 with 16KB and the 1000XL with 64KB. RAM was still expensive enough to make this distinction worthwhile.

Related Topics:
1982 - Fab

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

But when the machines were actually released there was only one version, the 1200XL, an odd hybrid of features from the Sweet 16 project. For instance, the 1200XL included the expansion chassis connector, but the case did not include a hole for it to be used. A new video chip provided more chroma for a more colorful image, but then someone decided to disconnect that pin in the monitor port, reversing the effect. It is not clear, in retrospect, why these odd changes were made so late in the project.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

A number of problems in this machine, including a change in its operating system which made many programs written for the 400/800 computers incompatible, made the machine a flop. While Atari digested the problem, Commodore released the Commodore 64. While the 1200XL was technically a match, it was shunned by consumers as its problems, often completely overblown, spread by word of mouth.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Atari addressed the 1200XL's problems in the hastily-completed 600XL and 800XL, which were largely identical to the original Sweet 16 specifications. These machines also had Atari BASIC built into the ROM of the computer. By the time the new machines replaced it the Commodore 64 had already become the market leader, and Atari was unable to address this.

Related Topics:
ROM - Commodore 64

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

By 1983 Commodore had started the price war that would end with all of the home computer companies in the US being killed off during the Videogame crash of 1983.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Tramiel Era: XE series and XEGS

The final machines in the series were there 130XE and 65XE. These were really just cut-down versions of the 600XL and 800XL in much cheaper cases, a result of Jack Tramiel's efforts to wring every dollar out of the platform before finally killing it. The X in XE stood for XL-Compatible.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

A theory for why the number 65 was used for the first machine in the XE series is because Atari wanted their machine-numbers to correspond with the amount of RAM they came with, but as Commodore already used the numbers 64 and 128, Atari decided to add one to 64 and chose 65. All subsequent model-numbers were multiples of 65 instead of 64. This numbering-scheme was used in the Atari ST line of computers as well.

Related Topics:
Commodore - Atari ST

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

An additional 800XE was available in Europe (mostly Eastern Europe), it was basically a 130XE with half the memory. Almost as an afterthought, there was also the XE Game System (XEGS), released in 1987. Launched at the very end of the family's market life, the XEGS was sold bundled with its detachable keyboard, a joystick and a lightgun, and a couple of game cartridges.

Related Topics:
Europe - Eastern Europe

~ ~ ~ ~ ~ ~ ~ ~ ~ ~