The development of the Commodore 64 started in early 1981 with the project name "VIC-40" which was planned as the successor of the VIC-20. Since the MOS engineers Charles Winterble and Al Charpentier got no support from the marketing for a new VIC-40, these discussed their plans concerning their VIC-40 secret project in the silence in a restaurant Friday at night. This was also the birth of the 6510 processor, a 6502 with I/O functions. It was also the birth of the 6526 CIA "Charlie's Interface Adapter", which was later renamed the "Complex Interface Adapter" and finished by Bob Yannes. However, the planned sound and graphics chip caused the most discussion. Charles and Al wanted to design the world's best gaming chip. After consulting with Jack Tramiel, Charles Winterble and Al Charpentier and the team began investigating all possible computers available at the time, including the Mattel Intellivison, anything from Atari, and the TI 99/4A from Texas Instruments. Al Charpentier was particularly interested in sprites. At the end they decided to use 8 multicolored sprites, a doubling in contrast to Atari's machines, which could only display 4 monochrome sprites. In the course of this, they only got 2 hours of sleep in the 10 days Jack Tramiel gave them to analyze the computers. After these 10 days, Jack Tramiel gives his approval in April 1981 for the development and production of the new 6566 VIC-II graphics chip. During the VIC-II development, Bob Yannes is working in parallel on the new sound chip, whose passion is electronic music generation. The engineers call this new sound chip the "Sound Interface Device" 6581, or SID for short. Bob Yannes wanted something sophisticated in terms of sound and especially amplitude control, where a sound could change its volume as it progressed. In the common sound chips of that time there was only sound on (full volume) or sound off (no volume). The SID was planned with 32 voices, but due to lack of time it became the well known 3 voices with 4 waveforms each: Sawtooth, Square, Triangle and Noise. Both chips were developed in record time, requiring only 9 months development time. The usual development time was twice as long.
During the development of the chips, Commodore first worked on high-end computers as well as the game console Ultimax in 1981, before the engineers talked with Jack Tramiel about a successor for the successful VIC-20. In the course of the discussion of the VIC-40 it also came to the decision of the RAM size, whereby Jack Tramiel decided for 64K byte, because the competition, including the Apple II, offered at that time maximally 48 KByte, and he was sure that the RAM prices will fall up to the sales start, whereby he proved foresight. Jack Tramiel also planned to introduce the new VIC-40 at the January 1982 Consumer Electronics Show (CES) in November 1981. Therefore, in December 1981, Charles Winterble assigned the project to Bob Yannes, Robert Russel and Dave Ziembicki, who thus had less than 2 months to build the complete computer system. In the course of the tight schedule, there is no time for a new case, and it is decided to use the case of the VIC-20 with a slightly different color. So the computer is ready in time for the January CES 1982.
Naming and Presentation
Until shortly after arriving at CES in January 1982, Commodore's marketing department knows nothing about the new VIC-40 computer, still focused on marketing the successful VIC-20 computer. After the engineers demonstrate the complete computer in a suite to Jack Tramiel, he gives the OK for the CES exhibition and informs marketing about it, who now frantically create promotional material for the prototype. In the course of this, and completely in the sense of the previous Commodore computers P128 with 128 KB RAM and B256 with 256 KB RAM, the marketing would like to rename the VIC-40 in C64, which then also happens (C64/C64C Service Manual). Due to the magic memory size of 64 KByte, the impressive sound chip 6581 (SID) and the 8 freely movable multicolored hardware sprites in the VIC-II chip, the visitors at the CES queue up in front of the Commodore booth, and so the C64 becomes a sensation, also because it appeared out of nowhere without any advance notice.
Since Jack Tramiel wants to go into production with the new C64 as quickly as possible, he expects his engineers to complete the production-ready computer in only 3 months. In April 1982 it became clear that this schedule cannot be kept. In the course of the completion, the problem of the color representation is worked on, with which not only Commodore, but also Atari and Apple have to fight, since brightness and color tone overlap. Charpentier therefore made a change by separating the color timer from the refresh rate. This solved the color problem, but the timers no longer ran in phase and the image waved. For this reason, a phase controller had to be installed, which was only available from Texas Instruments, who directly competed with the C64 with their TI-99/4A home computer. Commodore managed to copy the chip and produce it themselves. The complete C64 ended up costing only $130 to produce. It goes on sale for $600, the usual quadrupling of the manufacturing cost, after the principle that the manufacturing price is doubled and given to the dealers, who in turn double the purchase price, which is then the selling price. The C64 goes into production in June 1982.
In the course of the production still numerous problems arise. To get all components into the small case, they were arranged close together, which led to problems. In April 1982 Robert Russel sends ROM code unchecked to the factory, where immediately hundreds of thousands of chips are produced. Only later was a bug discovered that made some upper case letters lower case, rendering hundreds of thousands of ROM chips useless. Delivery problems of chips also caused delays. The VIC-II also has a "sparkle bug", where a hot VIC-II produces light blue dots in the background. Errors are also made during production in the manufacturing halls. The video modulators of the C64 are wrongly calibrated by the assembly line workers, which results in an oversaturation of the picture. In addition, the use of electrical screwdrivers cuts the conductive paths for the 1541's drive connector.
The non-existent "KERNAL"
In the course of the VIC-20 release, Robert Russel, who was involved in the development of the VIC-20 and C64, makes a mistake in using the actual word "Kernel": "I made a mistake in the heading on the first page". The consequence of this can be read in the respective "Programmers Reference Guide" of the VIC-20 and C64. Not only back in 1981, but also until today, most C64 users fall for this mistake, and misspell the word "Kernel". Even on the pages of Wikipedia, the word Kernel is misspelled on countless pages.
The non-existent "IEC" Bus
The serial protocol that Commodore uses for its computers is the IEEE-488 bus standardized in 1975, which Commodore converted to a serial version. The IEEE-488 is an 8-bit parallel bus. There never was an IEC bus, but this term is still used on some sites. On the outside of the C64 case it is called "Serial" (Serial-Bus), and in the manual of the C64 it is also called "Serial". So it is clear that the correct term is "Serial-Bus".
Nationwide sales of the C64 begin in August 1982 and were initially sold through ComputerLand and other chains. At the beginning the C64 and the completely overstrained power supply causes apartment fires, and also the Sparkle bug continues to occur, whereby the error is discovered by Al Charpentier later in a ROM chip. Commodore additionally struggles with 1541 drive problems, as drive heads became misaligned. Despite all these problems, the C64 turns out to be a best seller. In the first two weeks Commodore sells 12,000 units, in the next 3 months until December 1982 another 65,000 units. The following Christmas business exceeds all expectations. Due to various events the interest in computers is stoked, and under these conditions the triumphal procession of the C64 begins in 1983, which is to last until 1994. Exact sales figures for the C64 do not exist, but estimates range around 25 million units sold, making the C64 the best-selling home computer of all time and a cult object of the 80s.
Processor: CPU 6510, 1 Mhz
Graphics: VIC-II 6567 (NTSC), 6569 (PAL)
Memory: 64 KByte DRAM
Sound: SID 6581
UNI64 Computer - The further development of the C64
In contrast to some C64 replica boards or emulators, which only want to preserve the C64 but not fundamentally change it, and thus do not offer new and diverse possibilities, the new developments of the UNI64 computers go new ways with the C64, which inspire and amaze many people. Especially the new forms or the special features make these new developments so interesting for open-minded computer users.
The starting signal was given in October 2019 with the development of the uAX64 Mini, which had a footprint of only 100x100mm, making it the smallest C64 in the world to date that uses the original chips of the C64. This offers the possibility to put 6 cards into 6 existing slots, which together make a C64 expandable in seconds with unimagined possibilities.
The development of the uAX64 Mini prototype turned out to be a bit bumpy. After the first prototype was assembled, nothing worked. After 3 days and numerous error corrections, flying wires and adapters, the power-on screen of the C64 appeared for the first time, but without a blinking cursor. It took another 3-4 board versions until everything worked properly. The difficulty in the development was the limited number of pins, because in order not to exceed the 100x100mm size, only 72 pin slots could be used. 72 pins sound like a lot, but they were not enough to connect everything together. In the end, the Restore key line had to be optionally routed through a connector. On the right side of the uAX64 Mini a possibility was created to plug the Pi1541 externally to the system to connect an original 1541 in miniature format.
Shortly after that followed the uEliteBoard64, which seems to be impossible to implement due to the "short board" format and the amount of features, but nevertheless could be realized. Among the countless possibilities are the 3 SID sockets, I2C connectors, KERNEL ROM switch via RESTORE button, SRAM socket OnBoard, HAT socket for internal pluggable extensions, as well as many measuring points which invite not only to experimenting, but in conjunction with the many possibilities, can also be used very well for the development of new hardware and software. In addition, the video modulator with its two sockets can be easily replaced. A Stripe-Fix option is also built in. WS2812 RGB led strips can be controlled via an adapter board (optionally automatically via the SID).