Apple II History Part 12: Peripherals & the Apple II Abroad

@SteveJobs

Published in

Apple II History

· 5 years ago

Compiled and written by Steven Weyhrich
(C) Copyright 1992, Zonker Software

(PART 12 -- PERIPHERALS & THE APPLE II ABROAD) [v1.1 :: 11 Jul 92]

THE APPLE II ABROAD

Early on, Apple got involved in selling the AppleII in Europe and the Far East. To function in those parts of the world called for a change to handle a different voltage (240V instead of the 120V we use in the U.S.). Also, the language differences had to be overcome. It was easiest in Europe where, for the most part, the standard Roman alphabet was used. The primary differences were in symbols used together with letters for certain specific uses. Apple's Europlus][ had a modified ROM, and certain ESC key sequences could generate the German umlaut symbol to go with certain vowels.<1>

When the IIe was released there were some other differences. The German version was built with a an external switch below the keyboard, allowing the user to change between a standard U.S. layout and a German layout. (American versions of the IIe lacked the switch, but had a place on the motherboard that could be modified to allow a Dvorak keyboard layout to be switched in instead of the standard keyboard). The IIe auxiliary slot, which was placed in line with the old slot 0 on American versions (but moved forward on the motherboard) was placed in front of slot 3 on German versions. This was because the European AppleIIe's also had added circuitry to follow the PAL protocol for video output used for televisions and computer monitors in Europe (in the U.S. the NTSC protocol is followed). Because of the extra space needed on the IIe motherboard for the PAL circuits, the auxiliary slot had to be moved to be in line with slot 3. Because the 80-column firmware was mapped to slot 3, if an 80-column card was installed in the auxiliary slot it was not possible to use any other card in slot 3. Versions of the IIe made for other European countries had similar modifications to account for regional differences.<1>,<2>

When the AppleIIc came along, it was designed from the start to take the foreign market into account. If you recall, the U.S. version of the IIc had a standard layout when the keyboard switch was up, and a Dvorak layout when the switch was down. European versions were similar to the American layout with the switch up, and had regional versions that could be swapped in with the switch down. The British version only substituted the British pound sign for the American pound sign on the "3" key, but the French, German, Italian, and Spanish versions had several different symbols available. A Canadian version of the IIc was the same as the American with the switch up, and had some other special symbols with the switch down. This version was unique because each keycap had the symbols for both switched versions. For example, the "3" key had the "3" and "#" symbols, plus the British pound symbol, making it a bit more crowded than a typical keycap.

The AppleIIGS continued the practice of making international versions available, but improved on the design by making the various keyboard layouts all built-in. On the IIGS it was selectable via the control panel, as was the screen display of the special characters for each type of keyboard.

APPLE II PERIPHERALS

Moving on, we will now take a look at hardware items that extend the capability of the AppleII. The ability to add an external hardware device to a computer has been there from the earliest days of the first Altair to the present. In fact, the success of a computer has inevitably led to hackers designing something to make it do things it couldn't do before. The more popular the computer, the more variety you will find in hardware add-ons. The AppleII, designed by a hacker to be as expandable as possible, was once a leader as a platform for launching new and unique hardware gadgets. Today, in 1991, the AppleII unfortunately no longer holds the front position; it has been supplanted by the Macintosh and IBM crowd. However, the AppleII still benefits from the "trickle-down" of some of the best new devices from other computers (SCSI disk devices and hand scanners, for example). This is due partly to emerging standards that make it easier to design a single hardware device that will work on multiple computers, and in the case of the Macintosh, because of Apple's decision to make peripherals somewhat compatible between the two computer lines.

Trying to sort out all the peripheral devices ever designed for the AppleII series of computers into a sensible order is not easy. In this segment of the AppleII History I'll try to give an overview of hardware devices that were either significant in the advancement of the II, or unique, one-of-a-kind devices. Obviously, this cannot be a comprehensive list; I am limited to those peripherals about which I can find information or have had personal experience.

WHAT IS A PERIPHERAL?

A basic definition of a peripheral would be, "Something attached to a computer that makes it possible to do more than it could previously do." It is called a "peripheral" because it usually is connected to the computer after it leaves the factory. An argument could be made that something built-in is not a peripheral, but as things have changed over time there are some devices still called "peripherals" from force of habit, though they are now built-in (hard disks come to mind). Quite probably, in time manydevices that were once considered optional accessories will become so essential that they will always be built-in.

Recall that the earliest computers came with almost nothing built-in. They had a microprocessor, a little memory, some means of data input and display of results, the ability to access some or all of the signals from the microprocessor, and that was all. For those computers, the first things that users added were keyboards and TV monitors to make it easier to use them. Recognizing that the earliest hardware peripherals were keyboards and monitors highlights one fact: Nearly everything that is sold as a peripheral for a computer is either an input device, and output device, or an interface to make it possible to connect input and output devices. Exceptions are cards to add memory, co-processor cards to allow it to run software from another computer, and accelerators to make the computer run faster.

EARLY PERIPHERALS

When we come to the release of the first AppleII, two important "peripherals" were built-in: A keyboard, and the circuitry to allow easy connection of a TV monitor. It had, of course, the slots for inserting expansion cards (none were available), a game port (for attaching the game paddles that were included), a pin that could be used to connect an RF modulator (so a standard television could be used instead of a computer monitor), and a cassette interface. Since there were no cards available to plug into the slots, you would imagine that the AppleII couldn't make use of any other hardware. However, those early users who had a need usually found a way around these limits.

To get a printed copy of a program listing, for example, was no trivial matter. First, there were very few printers available. Those who could, obtained old used teletypes salvaged from mainframe computers. These noisy, massive clunkers often had no lowercase letters (not a big problem, since the AppleII didn't have it either), and printed at the blazing speed of 10 cps (characters per second). To use these printers when there were yet no printer interface cards to make it easy to connect, hackers used a teletype driver written by Wozniak and distributed in the original AppleII Reference Manual (the "red book"). This driver sent characters to the printer through a connection to the game paddle port. One part of being a hacker, you can see, is improvising with what you have.<3>

Another of the earliest devices designed for the AppleII came from Apple Pugetsound Program Library Exchange (A.P.P.L.E.). They were involved in distributing Integer BASIC programs on cassette to members of the group. To make it easier to send those programs to the person responsible for duplicating the cassette, Darrell Aldrich designed a means of sending the programs over the telephone lines. There were no modems available at the time, so his "Apple Box" was attached to the phone line with alligator clips and then plugged into the cassette port on the AppleII. To send a program, you first called up the person who was to receive it and got the computers on each end connected to the Apple Box. The sender then used the SAVE command in BASIC to tell the computer to save a program to tape. In actuality, the program was being "saved" through the cassette "out" port to the Apple Box, and onto the phone line connected. At the other end of that phone line, the data went into the other Apple Box, which was connected to the cassette "in" port on the other AppleII. That computer was executing the LOAD command in BASIC to "load" the program from the Apple Box. A.P.P.L.E. sold about twenty of these Apple Boxes at $10 apiece.<3>

INTERFACE CARDS

One of the first interface cards made for the AppleII was released, naturally, by Apple. The AppleII Parallel Interface Card was released in 1977 and sold for $180.<4> Wozniak wrote the firmware ROM, and managed to make it fit entirely in only 256 bytes. As a parallel device, it used eight wires to connect the computer with a printer, one line for each data bit in a byte. Various parallel devices also used one or more extra wires as control lines, including a "busy" line (so the receiving device could tell the sending device to stop until it was ready for more), and a "ready" line (so the receiving device could tell the sending device to resume transmission). Because each of the eight bits needed a separate wire, the cables for parallel devices looked like ribbons and were not very compact. Most of the early printers available required this type of interface.<5> A problem noticed with Apple's card, however, was an inability to properly handle these "busy" and "ready" signals (a process known as "handshaking"). One solution offered by a reader of Call-A.P.P.L.E. magazine in 1979 was to add a couple of chips to the card. If that was not done, however, the only way to do printouts that were very long was to either buy a 2K print buffer that could be used with some early printers, or use the "SPEED=" statement in Applesoft to slow down the speed at which data was sent to the printer.<6>,<7>

Apple released the Centronics parallel printer card in 1978. Selling for $225, it was specifically designed to work with Centronics brand printers.<4> It was similar to the Parallel Printer Interface, but had fewer control codes. The "Centronics standard" used seven data bits and three handshaking bits.<8> It would automatically send certain control codes to the printer when a program sent the proper command (such as a change in line width). As such, it was limited to properly working only with a Centronics printer, but many companies made printers that used the same control codes and would work with it.<5>

In April 1978 the AppleII Communications Card came out, selling for $225.<4> It was intended for use with a modem, and worked for speeds from 110 to 300 baud. The low speed (by today's standards) was for several reasons. One was that most modems of the time were acoustic. With an acoustic modem you dialed up the number yourself, and when you made a connection you put the handset (that's the part you talk and listen with, for you non-technical folks) into rubber sockets to seal out extraneous sound. A tiny speaker and microphone in the modem were then used to send and receive signals. This leads to a second reason for the low speeds of the time, which was that greater than 300 baud communications was not considered possible. In fact, the Phone Company was quite certain that speeds over 300 baud were not possible with any modem, although they would be glad to lease you a special data-quality phone line so you could get the best possible connection at 300 baud.

The AppleII Serial Interface Card ($195) appeared in August of 1978.<4> Serial devices required fewer data transmission lines, and so could work with more compact cables. Instead of sending each byte as eight simultaneous bits as was done in parallel devices, serial interfaces send each byte as a series of eight bits, which only took two wires; one to send and one to receive data. Like the parallel cards, there were a couple of other wires that went with the data lines to control handshaking. Also, serial cards needed a means of letting the sending and receiving devices identify when a byte began and ended, and the speed at which data was being transmitted. This meant that some additional information, such as "start" bits, "stop" bits, and "parity" bits, was needed.

The original version of the Serial Interface Card had a ROM that was called the P8 ROM. It contained the on-card program that allowed a user to print or otherwise communicate with the card without having to know much on the hardware level. The P8 ROM didn't support handshaking that used two ASCII control characters named ETX (Control-C) and ACK (Control-F), so a later revision called the P8A ROM was released. (ASCII stands for American Standard Code for Information Interchange). This worked better with some printers, but unfortunately the P8A ROM was not compatible with some serial printers that had worked with the earlier P8 ROM.

The Apple Super Serial Card firmware was finished in January 1981. It was called "super" because it replaced both the older Serial Interface Card and the Communications Card. To change from one type of mode to another, however, called for switching a block on the card from one position to another (from printer position to modem position). The Super Serial Card was also able to emulate both the P8 and P8A Serial Cards, making it compatible with most older software written specifically for those cards.<9>

VIDEO CARDS

After getting a printer interface card (and printer), the next variety of peripheral cards popular for the AppleII and IIPlus were ones that allowed display of 80 columns of text (which was rapidly becoming a standard outside the AppleII world). An early entry into this market was the Sup'R'Terminal card made by M&R Enterprises, the same company that made the Sup'R'Mod RF modulator for the AppleII. One of the most popular of the 80-column cards was the Videx Videoterm. Videx even made a display card that would display 132 columns card for the AppleII, but it never made much headway in the computer world (being supplanted by bit-mapped graphics displays, ala Macintosh).<3>

Many other companies made 80-column cards, but for the most part they were not very compatible with each other. One problem was deciding on a method to place the characters on the 80-column screen. With the standard Apple 40-column display, you could use either the standard routines in the Monitor, or directly "poke" characters to the screen. With these 80-column cards, they often used a standard from the non-Apple world, that of using special character sequences to indicate a screen position or other functions. For example, to put a character at row 12, column 2, a program needed to send an ESC, followed by a letter, followed by 12 and 02. Similar ESC sequences were used to clear the screen, scroll it up or down, or do other things that Apple's built-in screen routines could do.

When the AppleIIe was released, with its RAM-based method of displaying 80 columns of text, nearly all the older 80-column cards disappeared from the market. As of 1991, only Applied Engineering still makes one for those remaining II and IIPlus users that don't yet have an 80-column display.

One unique video product was made by Synetix, Inc. around 1983. Their SuperSprite board plugged into slot 7 (which had access to some video signals not available on other slots), and was promoted as a graphics enhancement system. It worked by overlaying the hi-res screen with animated "sprite" graphics (programmable characters that moved independently on any screen background). Since each sprite was on its own "plane" on the screen, they didn't interfere with each other. Also, it didn't take extra effort bythe 6502 microprocessor to manipulate the sprites; once the programmer placed the sprite on the screen and started it moving, it would continue until told to change. This was much easier than trying to program a hi-res game using standard Apple graphics. Unfortunately, at the price of $395 it never took off. (It was hard for developers to justify writing programs for only a few users that might have this card). Another company later made a similar card called the StarSprite, but it suffered the same fate. Even Apple's own double hi-res graphics, introduced on the IIe, had the same problem with a small supply of supporting software until the IIc and IIGS market got large enough to guarantee that enough owners had the capability of displaying double hi-res.<10>

ROM/RAM EXPANSION CARDS

All peripheral cards released for the AppleII up to the time of the AppleIIPlus were usable only in slots 1 through 7. Slot 0 was designed differently, and until the release of the Applesoft Firmware Card ($200) in 1979 nothing had been built to make use of it. The Firmware Card contained ROM that paralleled the upper 12K of AppleII memory. If you recall from the discussion in Part 3 of this History, Integer BASIC and the ROM version of Applesoft covered the same space in memory, and so could not co-exist. When it was clear that a floating-point BASIC (Applesoft) was what many people wanted, the IIPlus came out with Applesoft in ROM. To make sure that the previous AppleII owners were not left out, Apple released the Applesoft Firmware Card to plug into slot 0. It had a switch that allowed the user to select which BASIC should be active. In one position, the motherboard ROM would be selected, and in the other position the Applesoft and Autostart ROM was selected. Because there were quite a few Integer BASIC programs that AppleIIPlus users wanted to run, the Firmware Card also came out in an Integer BASIC version with the old Monitor ROM, that allowed IIPlus users to simulate owning a standard II.<4>

One of the benefits of the Integer BASIC ROM was the lack of something known as a "RESET vector" in the Autostart ROM. The Autostart Monitor was called that because it would automatically try to boot the DiskII drive when the power was turned on, and jumped to a known memory location when the RESET key was pressed. This allowed the disk operating system to reconnect itself, but more importantly made it possible to create copy-protected software. Since the Autostart ROM made it possible for a programmer to do something on RESET that prevented a user from examining his program, it was popular with companies producing programs that they didn't want copied and freely given away. Usually, a RESET on a protected program would restart the program, erase the program from memory, or re-boot the disk. The Integer BASIC and Old Monitor ROM lacked this feature; a RESET would just drop the user into the Monitor. This, of course, was just what hackers and those who liked to break copy-protection wanted. The users with non-Plus AppleII's or with the Integer BASIC Firmware Card on a IIPlus could prevent a RESET from restarting anything, allowing them to hack a program as much as they wanted.

The next card Apple released for slot 0 was called the Language Card. It was released in 1979 with Pascal, and expanded a 48K AppleII into a full 64K memory computer. It did not remove the upper 16K of ROM, but the card contained 16K of RAM that was electronically parallel to the ROM. Using "soft switches" (recall that these are memory locations that, when read or written to, caused something internally to change) one could switch out the ROM and switch in RAM memory. This extra memory was used to load the Pascal disk system, and under DOS 3.2 and 3.3, to load into RAM the version of BASIC that was not in the ROM. This was a more flexible alternative to the Firmware Card, and opened the way to other languages beyond BASIC for AppleII users.

Since the only way to get Apple's Language Card was to buy the entire Pascal system ($495), it was too expensive for many users. Other companies eventually came out with similar cards that did not require purchasing Pascal, and some of them designed the cards with more "banks" of memory, making 256K or more of extra memory available. Saturn Systems was one early suppliers of the large RAM cards. Typically, each 16K bank on the card would be switched in to the same memory space occupied by the Language Card RAM through the use of a special softswitch.<11>

CO-PROCESSORS

Although it did not go into slot 0, another significant card for the AppleII was the Microsoft Z-80 Softcard, which sold for around $300. It was a co-processor card, allowing the AppleII to run software written for the Z-80 microprocessor. The most popular operating system for the Z-80/8080 processors was the CP/M (Control Program for Microcomputers) system. Although the DiskII use a different method of recording data than was used by Z-80 computers, AppleII users managed to get programs such as the WordStar word processor transferred to the Apple CP/M system. Microsoft worked to make it compatible with the 80-column cards that were coming out at the time, since most CP/M software expected a screen of that size.<3>,<12>

After the arrival of the IBM Personal Computer and its wide acceptance by the business world, there was interest in a co-processor for the AppleII that would run IBM software. A company called Rana, which had been producing disk drives for the AppleII for several years, came out with the Rana 8086/2 sometime in 1984. This was a system that plugged into slots on a IIPlus or IIe, and would allow the user to run programs written for the IBM PC. It would also read disks formatted for that computer (which also used a completely different data recording system than the one used by the AppleII). One Rana owner, John Russ, wrote to A2-Central (then called Open-Apple) to tell of his experience with it: "We also have one of the Rana 8086/2 boxes, with two [Rana] Elite II compatible drives and a more-or-less (mostly less) IBM-PC compatible computer inside it. Nice idea. Terrible execution. The drives are half-high instead of the full height drives used in the normal Elite II, and are very unreliable for reading or writing in either the Apple or IBM format... And this product again shows that Rana has no knowledgeable technical folks (or they lock them up very well). We have identified several fatal incompatibilities with IBM programs, such as the system crashing totally if any attempt to generate any sound (even a beep) occurs in a program, or if inverse characters are sent to the display... The response from Rana has been no response at all, except that we can return the system if we want to. Curious attitude for a company, isn't it?"<13> By August 1985 Rana was trying to reorganize under Chapter 11, and the product was never upgraded or fixed.

A co-processor called the ALF 8088 had limited distribution. It worked with the CPM86 operating system (a predecessor to MS-DOS) was used by some newer computers just before the release of the IBM PC.<14>

Even the Motorola 68000 processor used in the Macintosh came as a co-processor for the AppleII. The Gnome Card worked on the IIPlus and IIe, but like other 68000 cards for the II, it didn't make a major impact, with the exception of those who wanted to do cross development (create programs for a computer using a microprocessor other than the one you are using).

The most successful device in this category was the PC Transporter, produced by Applied Engineering. It was originally designed by a company in the San Jose area called The Engineering Department (TED). The founder was Wendell Sanders, a hardware engineer who formerly had worked at Apple and was involved in the design of the Apple III and parts of the SWIM chip (Super Wozniak Integrated Machine) used in the IIc and IIGS. Around 1986 Applied Engineering began discussions with TED about buying the PC Transporter to sell and market it. At that time, the board was about four times the size it eventually became. AE's people were able to shrink a lot of the components down to just a few custom ASIC chips. The software that helped manage the board originally came from TED also.<15> It was finally released in November 1987, and included a card that plugged into any of the motherboard slots (except slot 3) and one or more IBM-style disk drives. The PC Transporter used an 8086 processor and ran about three times as fast as the original IBM PC. It used its own RAM memory, up to a maximum of 768K, which could be used as a RAMdisk by ProDOS (when not in PC-mode). It used some of the main Apple memory for the interface code that lets the PC Transporter communicate with the hardware.

The PC Transporter has undergone some minor hardware changes and several sets of software changes (mostly bug fixes but a few new features). The major reasons for hardware changes came about because of the availability of cheaper RAM (the original RAM was quite expensive and difficult to obtain). Additionally, changes were made to make the onboard "ROM" software-based, which made it easier to distribute system upgrades that enhanced hardware performance.<16>,<17>,<18> The major limitation for this product has been a reluctance by Applied Engineering to match the changes that have happened in the MS-DOS world and come out with a version of the Transporter that used a more advanced microprocessor (80286, 386, or 486). As of 1991 this is slowly beginning to become more of a limitation for those who wish to use both MS-DOS and AppleII software on the same AppleII computer, since advanced software needing those more powerful processors is beginning to be released for MS-DOS.

ACCELERATORS

The two things that all computer users eventually need (or at least want) are more storage and faster speed. The 1 MHz speed of the 6502 and 65c02 chips is somewhat deceiving, when compared with computers that have processors running at a speed of 20 to 40 MHz. To put things into perspective: Since the 6502 does more than one thing with a single cycle of the clock on the microprocessor, a 1 MHz 6502 is equivalent to a 4 MHz 8086 chip. Therefore, an AppleII with an accelerator board or chip running at 8 MHz is equivalent to an MS-DOS computer running at 32 MHz.

One of the first accelerators for the AppleII was the SpeedDemon, made by MCT. This board used a faster 65c02 chip, with some high-speed internal memory that was used to actually execute the programs (since the internal AppleII memory chips were not fast enough). In essence, it put a second AppleII inside the one you could see, using the original one for input and output. Another speedup board was the Accelerator IIe by Titan Technologies (formerly Saturn Systems; they had to change their name because it was already in use by someone else). This board worked in a similar fashion to the SpeedDemon. Some users felt this product ran faster than the SpeedDemon, but it depended on the application being tested. Both boards were attached to the computer by plugging them into a slot other than slot 0 on the motherboard.

In 1986 Applied Engineering introduced the TransWarp accelerator board. This product has lasted in the marketplace longer than any of the other ones, possibly because AE did far more advertising than the companies producing the older boards. The TransWarp did the acceleration using a different method. Instead of trying to duplicate all of the AppleII RAM within the accelerator, they used a cache. (If you recall from the segment on hard disk drives, a cache is a piece of memory holding frequently accessed information). Because they used the cache, the TransWarp did not require any high-speed RAM on the motherboard. Instead, any memory access was also stored in the cache RAM, which was high-speed RAM. The next time a byte was requested from RAM, the accelerator looked first into the cache memory to see if it was there. If so, it took it (far more quickly) from there; if not, it got it from motherboard RAM and put it into the cache. Early TransWarp boards ran at 2.5 MHz; later versions pushed this speed to 7 MHz (this was the top speed used by the TransWarp GS, released in November 1988 for the AppleIIGS).

The next step in accelerator technology was to put all the components of an accelerator board into a single chip. This happened when two rivals, the Zip Chip and the Rocket Chip, were released. The Zip Chip was introduced at AppleFest in May 1988, and the Rocket Chip soon after. Running at 4 MHz, the Zip Chip was a direct replacement for the 6502 or 65c02 on the AppleII motherboard. It contained its caching RAM within the housing for the processor, the difference being mostly in height (or thickness) of the integrated circuit. Installing it was a bit more tricky than simply putting a board into a slot; the 6502 had to be removed from the motherboard with a chip puller, and the Zip Chip installed (in the correct orientation) in its place. Software to control the speed of the chip was included, and allowed about ten different speeds, including the standard 1 MHz speed (some games simply were too fast to play at 4 MHz, and software that depended on timing loops to produce music had to be slowed down to sound right). The controlling software also let the user determine which (if any) of the peripheral cards should be accelerated. Disk controller cards, since they used tight timing loops to read and write data, usually could not be accelerated, where many serial and parallel printer and modem cards would work at the faster speed. The Zip Chip even allowed the user to decide whether to run all sound at standard speed or at the fast speed.

The Rocket Chip, made by Bits And Pieces Technologies, was almost exactly the same as the Zip Chip, with a few minor exceptions. It was sold with the ability to run programs at 5 MHz, and could be slowed down below the 1 MHz speed (down to 0.05 MHz). Later, when Zip came out with an 8 MHz version of their Zip chip, a 10 MHz Rocket Chip was introduced.

The rivalry between Zip Technologies and Bits And Pieces Technologies came from a mutual blaming of theft of technical information. The Bits & Pieces people insisted that they had done the original work on a single chip accelerator with the Zip people, but had all the plans and specifications taken away without their permission. Consequently, they had to form their own company and start from scratch to design their own chip. Zip, on the other hand, insisted that Bits & Pieces had stolen the technology from them. The problem eventually came to court, and it was decided that Zip Technologies was the originator of the technique and the Rocket Chip had to stop production.

NEXT INSTALLMENT: Peripherals, cont.