Last edit Feb 16 2009. (c) 2009 written by Herb Johnson
Dwight Elvey and I debated during 2007 and '08, who was "first" to offer an S-100 compatible floppy disk controller; specificaly, the first with CP/M. I think it's tough to pin down, because "first" needs to be qualified, and actual dates for events are not well documented. Are we talking about development or announcement or delivery or production? The first delivery? It's a challenge to figure all this out, as information is only available in bits and pieces over time.
What has emerged, is that development of floppy controllers was a kind of "horse race", because the "contenders" did their work and production at about the same time, around and during the year 1976. For the S-100 bus, that was the first full year of the IMSAI 8080 and MITS Altair 8800. Meanwhile, floppy controller developments for minicomputers and then-new microprocessor-based systems were underway.
Here are some reference dates. On January 22 1973, IBM announced the 3740 Data Entry System - that initiated a "Diskette" and floppy drive standard. By the end of that year, Shugart and others offered compatible or similar floppy drives. The Altair 8800 was announced in Jan 1975; IMSAI's 8080 was announced late 1975, delivered Dec 1975. The controller "contenders" I'll discuss in this document are: Digital Systems (CP/M license #1) and the FDC-1 ; IMSAI (CP/M license #3) and the FIB and IFM board set; MITS and their 88-DCDD (with disk BASIC) and iCom and their generic floppy controller (with iCOM's own DOS).
Information on events and dates come from ads or product mentions or editorial comments - some of which are unclear. And Digital Systems' start-up history is not clear, and that matters because the FDC was their first product. I welcome any VERIFIABLE date reference for ANYTHING to do with these three products, or with early CP/M.
Icom was "first" (among others) to offer a generic floppy controller, the FD360, for several processors - in Sept 1974! The design seems similar to Torode's design: a smart controller card, built from small TTL chips, and a parallel interface to a card on the microprocessor (or minicomputer) system which would control the controller.
The system was offered at that time with "interfaces" for the Intellec 8 or 80, National Semi's UMP-16, the HP 2100 and Data General Nova. But the bridge between them was described as several "input and output lines", and the intention was for it to be built into other systems.
ICOM's FDOS-II operating system was advertised in Jan 1976, for the Intel MDS-800 (Multibus) with an interface card. Manuals for the FD360 controller date as early as March 1976; schematics for a S-100 interface card for the Icom system have dates of Sept and Oct 1976. I do not know if iCom made that interface card, or when that S-100 option was announced. CP/M was sold with an iCOM system and S-100 card by Synetic Designs. One system is owned by Rich Cini and he and I describe it on this Web page.
A similar product was offered by Sykes: an OEM floppy controller and drive in ads for Jan and Feb of 1976. "OEM" meand you build the computer interface for your company's products (or your own use), no OS mentioned. I believe a few other vendors offered OEM floppy controllers in that period, for either minicomputers or microprocessor systems, with either suggested software methods or software for specific computer processors.
The March 1976 edition of MITS's "Computer Notes" newsletter, probably distributed in February, announced in a several-page artice the "Altair Floppy System". The article shows and describes a Pertec FD400 8-inch floppy drive in a case and "two PC boards that fit in the Altair chassis". It describes a single-density format on an 8-inch hard sectored diskette. Software is "a preliminary version of Disk Extended BASIC" which is described as "already been shipped, and the production version will be ready to ship on March 29, 1976".
The MITS product's hardware, later called the 88-DCDD, consisted of two cross-cabled Altair bus compatible boards, and a third card between the Pertec drive and one of the bus cards. There was an 8080 processor on one Altair board, with ROM and RAM dedicated as the drive controller. Schematics and manuals about the Altair bus boards have dates of March and April of 1976, and later. But the "data card" connected to the Pertec drive electronics have drawings (schematics?) from MITS draftsperson A.C. Rouckus dated June 15 1975. Associated Pertec drive electronics documents include a "rev A" document from Feb 1975. (This information courtesy of Jack Rubin, as culled from his MITS documents.)
The software side of the MITS floppy product is harder to date. Harder still to date, is when either software or hardware ended up in customer's hands. I traced discussion of the 88-DCDD, MITS' (Micro-soft's) Extended Disk BASIC, and MITS' "DOS" in MITS self-published "Computer Notes", distributed to all Altair owners.
The April 75 apparently announced the MITS floppy disk product. The July 75 issue described it as "available in August" and gave a half-page of description. It mentions a "disk operating system" and "Extended BASIC" with file access.. A 1/4 page note "8080 BASIC" appears in the Jan 1976 issue; it lists file access as a feature; but it also notes the product (from Paul Allen of MITS and "Micro-soft") is for use with the Altair or the Intel Intellec 8/MOD 80 or MDS system. There is a three-page description in March 1976, of considerable detail, which claims "a preliminary version of Disk Extended BASIC has already been shipped, and a production version will be ready to ship on March 29 1976".
An April 1976 edition had disk software programs. The May 1976 issue has an article by Paul Allen on Disk Extended BASIC features not documented; and Paul Wasmund's article titled "DOS is near" says that product is "on the way". He describes a relocating assembler, a linker, a "System Monitor of about 7K bytes", a Text Editor, and "programs for copying one disk to another, initiating disks, etc.". The July 1976 edition has an article by Tom Durston on DCDD hardware, and another Paul Allen article on Altair Disk BASIC.
Unconfirmed information on one individual's actual delivery schedule for these products is as follows:
88-DISC (two 8" [Pertec] drive kits), 11-JUN-1976
88-DCDD (disk controller), 31-JUL-1976
Extended Basic (w/disk support), 26-JAN-1977
[Altair] DOS (i.e. MITS DOS, not CP/M), 14-JUL-1977
I discuss MITS disk controller development on this Web page of MITS S-100 boards and manuals on my S-100 Web site.
Dr. John Torode founded Digital Systems in 1967 as a grad student at the University of Washington. In 1974 and 1975, he worked with Dr. Gary Kildall to provide a working floppy disk controller; they were first to run Kildall's CP/M on that controller late in 1974. Torode's Digital Systems, with Kildall, made the first CP/M sales. Torode both sold CP/M and bundled it with his FDC-1 floppy controller and an Altair-compatible bus interface card during 1975. Meanwhile Kildall founded Digital Research to sell and develop CP/M.
The Digital Systems floppy controller board itself is bus-independent. It's referred to in the docs and schematics as "FDC-1" (or -2 or -3). Digital Systems provided a S-100 interface card, named "HB-1.X", where X is -1, -2, -3, -4. The interface between them was a set of data and control lines. The controller did the fast and difficult work of data read and write; the interface allowed the S-100 processor to work at the track and sector level.
Kildall's and Torode's work was first discussed in Dr. Dobbs Journal by the editor, starting in April 1976, as "tidbits..available from Digital Research...tentative pricing" for software, manuals, hardware. Torode and Digital Systems was first mentioned in Aug 1976 in DDJ for the hardware, Digital Research for CP/M. Note that DDJ did not permit ads in 1976, for any company! But there are references to someone who bought a system earlier, perhaps in 1975.
Digital System docs I've seen so far, include a Dec 1976 dated schematic for the FDC-1 DRI's first ad I could find, seems to be in BYTE for Dec 1976; Torode's Digital Systems first ad in BYTE is Jan 1977.
See this Web page for the history of Digital Systems floppy controller development and CP/M.
IMSAI developed a floppy controller during 1976, as a dual s100-board set. IMSAI's first controller was two S-100 cards: an 8080 based card and a data seperator card. They performed most of the same functions as the single controller card, but used a local 8080 processor and its ROM code as the microcontroller.
An IMSAI ad in June 1976, mentions a "floppy controller with on-board processor and DOS". But IMSAI could not get their in-house software to work well. They brought in Gary Kildall himself to fix it. Todd Fischer himself was there, and says this occurred "in the later half of 1976". IMSAI bought a distribution license (#3) from Kildall at some point during or after development.
IMSAI sold CP/M 1.33 with their FIB and IFM controller board pair. Code and docs from the period have dates as early as Oct 1976, that I'm aware of. Todd confirmed "we offered CP/M 1.3 with the original IMSAI IFM/FIB controller from very late 1976", in public correspondence with me in comp.os.cpm. and subsequently offered IMDOS 2.02, 2.05 (1977? 78?), and CP/M 1.4. (This is documented on the imsai.net Web site, but events there are not well dated.)
See this Web page for a discussion of early IMSAI floppy controller development. For IMSAI's general history, read this Web page.
So, from my research and other's, both IMSAI and Digital Systems had floppy controllers in development during 1976. Both "announced" them in various ways in 1976. "Delivery" is a vague concept if you don't consider quantity or "production". One can argue that Torode had SOMETHING in the field, in use, before IMSAI did. But IMSAI was a well established small company and so their announcements were in paid ads - but they were vague. Digital Systems and Digital Research were barely beyond one-person consultancies, and used word of mouth and reviews first, then ran ads later - but they were product-specific.
All that gives you some idea about the issues, and why I call it a "horse race". Here's how I see it:
Before the S-100 race began:, in 1974 companies like iCom had OEM floppy controllers which an engineer could later interface to something like a S-100 system. Software could be just a specification for your system's OS, or you could roll your own OS. But iCOM provided an Intel MDS interface and some kind of 8080 or 8008 support.
Meanwhile, in late 1974, Torode worked with Kildall on a floppy controller for an Intel 8080-based development system, to run a CP/M operating system developed by Kildall. While Kildall tried to build a controller, Torode certainly had his own controller design in 1975, because he co-published a design in an academic journal in 1973! Torode's Digital Systems likely provided at least one hardware system with CP/M to a few private or OEM customers in 1975. Kildall formed his Digital Research company during 1975.
The S-100 race began sometime in 1975. The Altair was announced in Jan 1975, Popular Electronics. The IMSAI 8080 was advertized in the summer of 1975 in PE; supposedly it was not delivered until December of 1975.
To my knowledge so far, the first published information on a floppy controller product to plug into the S-100 bus, was MITS's March 1976 "Computer Notes" newsletter's announcement of the 88-DCDD disk controller, with "Disk Extended BASIC" to operate it. Documents from the period show software and schematics dated March and April of 1976.
The first published information on a floppy controller product to plug into the S-100 bus for CP/M was about what BECAME the product of John Torode's Digital Systems - the FDC-1, with CP/M. That was in April 1976, in DDJ. And you can't change one qualifier of that description. Because..
... the information was about Digital Research, which never produced a controller! Torode and Digital Systems was not mentioned by name until August 1976 in DDJ. And that is the earliest published mention of that company with CP/M that I'm aware of. But...
...that's after IMSAI's ad on their "floppy controller" of June 1976, a product still in development in Sept of that year.
And by September 1976, iCOM (or some other OEM), developed a S-100 card for the iCom system. They offered their own FDOS, not CP/M.at that time. Later on, a CP/M 2.0 became available.
Certainly the "race" ended in late 1976. All three crossed the finish line by then, in what order I don't know - depends on how you draw the "line". But by 1977, others were also offering floppy controllers - I've not researched specifics but I think NorthStar's was out by then, maybe Micropolis. There were OS's for each controller: CP/M did not become standard for some time. CP/M 1.4, the first popular release, was offered in 1978.
It's hard to find the date of first delivery of each of these products, unless you find someone with a receipt in hand. In a letter published by Dr. Dobb's Journal of Nov/Dec 1976, Robert Swartz in Nov 1 1976 said he had a Digital System product a year prior - in late 1975. Unconfirmed delivery of a MITS floppy controller was July 31 1976 (software much later?).
MITS is likely first to announce a S-100 floppy disk controller, in February 1976. But it ran only from BASIC until (much?) after May 1976 when a "DOS" (not CP/M) was announced as "almost here". It's possible Torode had something to a customer earlier than 1976, however.
I conclude that Torode's contributions are probably "first" in important ways. He published a design in 1973. He worked with Kildall to make Kildall's controller work with CP/M late in 1974. In 1975, he first licensed CP/M, bundled it with his controller, and offered it to select customers and clients. By late 1975, he let word get out to others about his products, possibly sold a few that way. In January 1976, he put a Digital Systems ad in the premier hobby computer magazine of the time - BYTE - where he described exactly what he was offering. And, he delivered, by all accounts.
IMSAI was also important. They "delivered", in due course, and within months of Torode's early efforts. By employing Kildall later on, IMSAI produced CP/M 1.33. Additional work with IMSAI and others led Kildall to produce CP/M 1.4, the first popularly distributed version of CP/M. The need to adapt to many kinds of controllers, and many disk formats per controller, led Kildall to seperate all features of disk format into the BIOS. And that opened CP/M for use by anyone, individuals and OEM's alike, and set a standard for microcomputer OS's.
...and the hardware independence of CP/M allowed more OEM's like iCOM to move to CP/M. The competition with CP/M forced other S-100 producers to make their OS's either CP/M like, or to license CP/M.
The emergence of CP/M as a standard, and as end-user adaptable, is why this early floppy controller development is important in microcomputing history.
Here's links to my references. Sorry there are so many pages, it's a lot of history after all! I welcome any corrections and DOCUMENTED dates of events,products, publications.
My IMDOS discussion notes are at this Web page about CP/M and IMDOS.
My MITS and Altair notes and document lists are at this Web page which is part of my S-100 manufacturer's Web page series. A number of people have contributed to that Web page and my MITS archives. That page in turn, links to several other Web pages about the following: was the Altair the "first" personal computer?; and the origins of the S-100 and Altair name.
My history of CP/M outline is at this Web page, which includes links to ads and references to several events mentioned here.
Early CP/M history is accumulated on this Web page which discusses primary references in detail.
My description of Rich Cini's Icom floppy is on this Web page which links to Rich's iCom docs online at his site.
My home page for Digital Research, CP/M and Digital Systems is at this link It refers to all the above Web pages and more.
Bill Degnan alerted me to a Kilobaud Microcomputing magazine article from March 1977 about floppy drives and controllers. Douglas Hogg wrote "Floppy Disks...What's the Real Story?". Here's a Web link to Bill's PDF of the article as of June 2010. He says "It's a good synopsis of the state of micro drive tech in early 1977."
Rich Cini, Dwight Elvey, Stan Sieler, Jeff Shook and Vern Potter have all provided Digital Systems or iCom documents; discussion and review; and/or considerable technical analysis. Dwight Elvey in particular has reconstructed both firmware and software for DS products, and he's given me a lot of encouragements and compliments. Jack Rubin specifically called out to me the MITS floppy story and provided dates and references.
A number of university libraries have provided me with access to engineering magazines and publications of the period. Thank you for keeping these for decades and making them accessable and lendable. I have some magazines and books in my own library.
All I've done is to assemble the work and documents of others, and glued them together with some chronology, references to publications, and some effort at continuity. It's tough to do that. Most of what I reference can be found by others. Please do so and verify what I've done.
Copyright © 2009 Herb Johnson