By December 1977, the Apple II had been available for about six months. Most customers used their television as an inexpensive color monitor, and used a cassette recorder to store and retrieve their programs and data. Apple’s major competitors were the Radio Shack TRS-80 and the Commodore PET. The products made by these two companies, together with Apple, could be considered as the second generation of microcomputers; they all came fully assembled and ready to use out of the box, with a keyboard and cassette interface. The TRS-80 and the PET even came with a monitor and cassette recorder. The strength of the Apple was expandability and graphics, while the strength of the others was cost (both the TRS-80 and the PET sold for around $600, half the price of the Apple II).
In the latter part of 1977, Apple had introduced some enhancements to the II, including their first version of a floating point BASIC (called “Applesoft“) on cassette, and a printer interface card to plug into one of the slots on the motherboard. But the Apple II still needed something to make it more attractive to buyers, to stand out above the TRS-80 and the PET. One area that needed improvement was its program and data storage and retrieval system on cassette; it was a continued source of frustration for many users. The cassette system used on the TRS-80, though slower, was more sophisticated than that of the Apple II, allowing named files and easier storage of files and data on the same tape. On the Apple II it took very careful adjustment of the volume and tone controls on the cassette recorder to get programs or data to successfully load. The Apple cassette system also needed careful attention to the location on the tape where a program was stored, and was no more accurate than the number on the recorder’s mechanical tape counter (if it had one).
IBM engineers had invented the 8-inch floppy disk in 1971, and over the next two years gradually increased its capacity from 80K to nearly 240K. Alan Shugart, an IBM manager, left that company and formed his own in 1973.
An engineer from Shugart Associates was challenged in 1976 by a company making an S-100 computer to make a smaller size floppy disk, more appropriate to the size of microcomputers. The disk size suggested was about the same as a table napkin, 5 1/4 inches. It was small enough to be a convenient size, but large enough to prevent it being put in a pocket (and therefore be at risk of being bent).
The company went on to design and market the SA400 “minifloppy” drive that same year, with a formatted capacity of 90K. These devices became popular with users of the various microcomputers that were being marketed and sold in the mid 1970s. Though much more expensive than cassette tape storage, diskettes offered greater convenience and flexibility.
Apple president Mike Markkula was one Apple II user that was dissatisfied with cassette tape storage. He had a favorite checkbook program, but it took two minutes to read in the program from the tape, and another two minutes to read in the check files. Consequently, at the executive board meeting held in December 1977 he made a list of company goals. At the top of the list was “floppy disk”. Although Wozniak didn’t know much about how floppy disks worked, he had once looked through a manual from Shugart Associates.
As an experiment Woz had [earlier] conceived a circuit that would do much of what the Shugart manual said was needed to control a disk drive. Woz didn’t know how computers actually controlled drives, but his method had seemed to him particularly simple and clever. When Markkula challenged him to put a disk drive on the Apple, he recalled that circuit and began considering its feasibility. He looked at the way other computer companies – including IBM – controlled drives. He also began to examine disk drives – particularly North Star‘s. After reading the North Star manual, Woz knew that his circuit would do what theirs did and more. He knew he really had a clever design.
Other issues that Wozniak had to deal with involved a way to properly time the reading and writing of information to the disk. IBM used a complex hardware-based circuit to achieve this synchronization. Wozniak, after studying how IBM’s drive worked, realized that if the data was written to the disk in a different fashion, all that circuitry was unneeded. Many floppy disks sold at that time were “hard sectored”, meaning that they had a hole punched in the disk near the center ring. This hole was used by the disk drive hardware to identify what section of the disk was passing under the read/write head at any particular time. Wozniak’s technique would allow the drive to do self-synchronization (“soft sectoring”), not have to deal with that little timing hole, and save on hardware.
Steve Jobs had been visiting the Shugart offices regularly, insisting that he needed a cheap $100 disk drive. After Wozniak figured out the details of how to control a disk drive, Jobs came back and said that not only did he want a cheap disk drive, he wanted just the mechanism; no read/write electronics, no head load solenoid, no track zero sensor and no index hole sensor. Shugart engineers were puzzled by the request, but agreed to give Jobs the drives he wanted. Identifying these stripped down drives with the product name “SA390” (being less than a standard SA400), they provided Apple with twenty-five of these “prototype” SA390 drives to use in developing a disk drive solution for the Apple II.
During the design of his disk interface card, Wozniak first worked with an SA400 drive using the standard controlling electronics included with it. He then made further modifications to successfully control the drive without those extra electronics. When Wozniak began to try some of the SA390 drives, he found that most of them worked, though several required adjustments. However, they found that some of the SA390 drives simply didn’t work, and Apple engineer Cliff Huston solved this problem by optimizing the analog electronics and creating a series of procedures to make it possible to do adjustments on the drives on a production line.
When representatives from Apple returned to Shugart to place orders for more of the SA390 drives to sell under the Apple brand, one of Shugart’s engineers admitted to a deception. The prototype drives that had been provided to Apple had actually come from a pile of bad SA400 drives. They had expected that Apple’s engineers would be unable to make the drives work, and out of frustration would have come back and purchased the more expensive SA400 drives.
As the first Disk II drives came off the assembly line, the first drive was given to Cliff Huston in recognition of his work in making the SA390 drives work. The second drive was given to engineer Wendell Sander, who had discovered and corrected a timing error in the controller’s state machine.,
Wozniak asked Randy Wigginton for help in writing some software to control the disk drive. During their week of Christmas vacation in 1977 they worked day and night creating a rudimentary disk operating system, working hard to get the drive ready to demonstrate at the Consumer Electronics Show in the first week of 1978. Their system was to allow entry of single letter commands to read files from fixed locations on the disk. However, even this simple system was not working when Wozniak and Wigginton left for the show.
When they got to Las Vegas they helped to set up the booth, and then returned to working on the disk drive. They stayed up all night, and by six in the morning they had a functioning demonstration disk. Randy suggested making a copy of the disk, so they would have a backup if something went wrong. They copied the disk, track by track. When they were done, they found that they had copied the blank disk on top of their working demo! By 7:30 am they had recovered the lost information and went on to display the new disk drive at the show. ,
Following the Consumer Electronics Show, Wozniak set out to complete the design of the Disk II. His original task on the disk controller was to reduce the chip count from the 40 chips used on the controllers for S-100 machines. Using so many chips was expensive, and it was also much too large to fit inside the Apple II case. In his redesign, he decided to use a single 8-bit ROM for tracking and reacting to the changing states of the disk controller as it decoded the bit stream being read from the disk. This concept eliminated more than a dozen of the chips used on the standard SA400 controller. Beyond that, he made additional design changes that reduced the total chip count to only nine. This eventually reduced further to eight, since two 555 timers were replaced by a single 556 timer.
On the disk drive itself, Wozniak made modifications to the way in which data was recorded on the diskette. He was able increase the FM encoding from nine sectors per track to ten sectors per track, which provided an eleven percent increase in storage on a disk. Then, after coming up with a completely different technique, he abandoned the FM encoding scheme in favor of a group code recording (GCR) method. This further increased the storage density without increasing the number of bits actually written to the disk, and allowed him to increase the number of sectors per track to thirteen, giving nearly 45% greater storage per diskette. To make this work, it was necessary to create a lookup table and use some clever programming techniques, but Wozniak had often done this in the past to reduce chip count (for example, the odd text and graphic video screen addressing system of the Apple II).
For two weeks, he worked late each night to make a satisfactory design. When he was finished, he found that if he moved a connector he could cut down on feedthroughs, making the board more reliable. To make that move, however, he had to start over in his design. This time it only took twenty hours. He then saw another feedthrough that could be eliminated, and again started over on his design. “The final design was generally recognized by computer engineers as brilliant and was by engineering aesthetics beautiful. Woz later said, ‘It’s something you can only do if you’re the engineer and the PC board layout person yourself. That was an artistic layout. The board has virtually no feedthroughs.'”
The Disk II was finally available in July 1978 with the first full version of DOS, 3.1. It had an introductory price of $495 (including the controller card) if you ordered them before Apple had them in stock; otherwise, the price would be $595. Even at that price, however, it was the least expensive floppy disk drive ever sold by a computer company. Early production at Apple was handled by only two people, and they produced about thirty drives a day. , Because of the custom hardware and software Apple created to manage and access the disks, they had a formatted capacity of 113K, 23K more than the capacity offered by Shugart.
The drives originally sold with Woz’s disk controller were from Shugart. Within a couple of years of the introduction of the Disk II, to further cut costs, they later approached Alps Electric Company of Japan and asked them to design a less expensive clone. According to Frank Rose, in his book West Of Eden:
The resulting product, the Disk II, was almost obscenely profitable: For about $140 in parts ($80 after the shift to Alps) [not counting labor costs], Apple could package a disk drive and a disk controller in a single box that sold at retail for upwards of $495. Better yet was the impact the Disk II had on computer sales, for it suddenly transformed the Apple II from a gadget only hard-core hobbyists would want to something all sorts of people could use. Few outsiders realized it, but in strategic terms, Woz’s invention of the disk controller was as important to the company as his invention of the computer itself.