Close
0%
0%

6809 OS-9 Level-I machine

hacking like it's 1983

Similar projects worth following
Build the pure Motorola MC6809 machine, I thought. It will be good fun, I thought. It's a nice and easy cpu, I thought. An OS shouldn't be a big deal, I thought. And then I met 6883, 6829 and 6844 :-)

Being frustrated from work and missing a challenge with an not too far away goal, I finally decided to tackle an idea I'm hatching since decades, namely to just go and build my own computer. I grew up with Atari ST and Commodore Amiga and such the 68K was my first choice. But then I thought, well, maybe I'll start a little more humble, with an 6809 system. Should be easy and cheap, a few chips, some ROM and RAM - and here you are. I am no professional in the matter, not at all, but a computer enthusiast since first contact with the Commodore PET series (the bigger ones) at school. Of course I professionally got into the PC world right away, but meanwhile, well, PCs... I can solder, and I played around a lot with Arduinos and such in the last years, so I felt fit enough to give it a go.

Sooo... I sticked myself to an internet terminal for a total of some 30 or more hours of research for details, example projects, data sheets, schematics, literature, hardware components, software possibilities and so on and so forth. After a few weeks with almost daily packets via ebay (unbelieveable, those Chinese), I had a generous stock of chips for experimenting with different designs.

Finding the appropriate design, the one which I thought would be the best for my idea, was, well, a real adventure. The point is the operating system. Hacking up a basic 6809 is indeed no big deal at all, after I had scribbled the first design of a very basic BASIC computer, I realised that it was already rather common, I read a lot of pages referencing to Grant Searle's design. But that's not exactly what I had imagined, I wanted to go a little bit bigger, to be able to run a real operating system on it. My idea is not to just get A computer running, but to stay as close as possible to the initial motorola family concept, to build the computer that Motorola DIDN'T build.

Now, 6809 and operating systems, that is a thing. Certainly not for the contemporary witness, but I remind you, those were pre-internet times... a good deal of digital archeology is essential... :-)

Finally I decided to go for the 6883/6847 - option for system design, as the result would be a machine quite close to a Dragon or a TRS CoCo - and such I think I'm having a good chance of getting an OS-9 ported and running on it. To be precise and honest, it's the slightly modified exemplary design from the data sheet of the 6883.

If this project is running well, I might tackle an even "purer" system, employing multiple 6829 and 6844 and try to get some UNIXoid OS running. (Pure megalomania :-) an 68k is SO much easier)

datasheet-MC6883_SAM.pdf

addressing 96k with 16 bits is totally impossible! 6883: hold my beer...

Adobe Portable Document Format - 5.82 MB - 11/18/2019 at 21:19

Preview
Download

  • 1 × MC6809E CPU Microprocessors, Microcontrollers, DSPs / Microprocessors (MPUs)
  • 1 × MC6883 (SN74LS783) SAM Microprocessors, Microcontrollers, DSPs / IO Controllers
  • 1 × MAX707 Power Management ICs / Power Supply Support
  • 1 × MAX232 Non-IC RF, IF, RFID, ZigBee Components / Up-Down Converters
  • 1 × MC1372 Semiconductors and Integrated Circuits / Misc. Semiconductors and Integrated Circuits

View all 15 components

  • no KiCAD

    kleinesee12/04/2019 at 02:40 0 comments

    Nope. I tried. Nope. I DO understand, convertible file format, open source and so on. But. I didn't like it. Went back to EasyEDA. And learned using ports! :-\ Now, seriously, did I mention I was a (bloody) rooky in schematics drawing? Oh, you've noticed already...

    ;-)

    I got the schematics cleaned-up and improved, so they're slowly looking like... schematics. :-) Split up now in several boards, resp. planes, but that's not yet a final design. I mean the layout of the pcb. Still not decided whether to go for an ATX-board or for a proprietary small bus system with multiple boards. At the moment the design shows different boards resp. planes.

    I will sooner or later make the schematics publicly accessible, but don't feel like, yet. Here's another preview:

  • KiCAD

    kleinesee11/28/2019 at 23:15 0 comments

    alright, OK, I got it

    improving and redrawing everything in KiCAD

    CPU, ROM and serial are final, too, will try to PROPERLY split the layout

  • resistors and capacitors

    kleinesee11/25/2019 at 14:46 0 comments

  • OS-9 OMG

    kleinesee11/22/2019 at 21:01 0 comments

    I got fed uo of drawing and dug a bit deeper into OS-9 (by books so far)...
    Folks, has anybody ever dug OS-9? If so, how comes it just... where was it? Why didn't I ever use any OS-9 computer? It is so brilliant, my mind is totally blown. I didn't know. Why didn't I know, the heck! I'm in love. :-)

    What an adventure this 6809 has become! Little did I know. I should have stuck to the 68K, it's so much easier, linear address space is such a relief. :-) 

    Boring almost... :-) my side (research, have good books at hand) project is of course still a 68K machine, a simple but reasonable computer is really lesser challenging due to things like linear SRAM. And if one likes it more sophisticated, an ATARI ST or AMIGA is only the beginning...  (Looking forward to sooner or later tackling the 030, maybe finally skipping the 68K)

  • 6883

    kleinesee11/21/2019 at 01:35 0 comments

    Okay... still a bit... ugly, but better, still.

    6883 with chip selects, DRAM connect and bus buffer for 6847. 6809 data bus connect still missing. One day I will re-draw those chips in a far more reasonable geometry. May the router do the job. 1-1 to PCB footprint is not reasonable. 

  • bizarre design

    kleinesee11/20/2019 at 23:43 0 comments

    it IS a weird system design... bizarre, but ingenious

  • rework in progress

    kleinesee11/20/2019 at 21:31 0 comments

    heck, I need to redraw it on multiple sheets, will take a little

  • Schematic Sciences

    kleinesee11/18/2019 at 21:40 0 comments

    Schematics... schematics... how, what file format, what software...? Have I mentioned that I am a bloody layman? - Another few hours online in the electronics design jungle.

    I ended up with EasyEDA. So far I'm quite fine with it.

    I finally cobbled something together which almost looks like a schematic, but it's very... unoptimized... :-)

    I'm going to optimize it and clean it up a bit first, it's too embarrassing for publishing now, but confident it is basically working, though. Will be back asap, but I'm not too fast, as it is a pure hobby project happening besides life, work, cooking, pets and such.

View all 8 project logs

  • 1
    Finding the right system architecture

    The first step when thinking about designing a single-board (or modular, doesn't matter actually) retro computer system is of course to think about the appropriate operating system. Disconsidering the Intel world, there's basically two options: simple BASIC (or similar, maybe assembler or C) or "real" operating system, meaning disk operating system.

    I went for "operating system" option, because BASIC is all too limited for the envisaged especially "nice" system.

    Through time there popped up several alternative operating systems on 6809-based systems, each of which has its own advantages - and disadvantages. There's FLEX and there's OS-9  and there's NitrOS9. There's no CP/M and no DOS.

    I think I'll end up with a more or less original OS-9, although I'll give NitrOS9 a try. We'll see, I think this will be challenging.

    Anyhow, a "real" OS requires a bit more of an environment than everything, ROM, RAM and periphery cramped into 64KB. With the 6809, there's two options: 

    First, the ingenious 6883, which makes most sense together with the 6847 video display controller. It's a "synchronous address multiplexer" that enables extending of the address space to 96K, meaning 64K or more of free RAM. It is designed for use with the 6847, so it makes most sense to me. I am not interested - at this stage - in sophisticated video output. That's why I'll hard-wire text-only mode with green-or-amber selection.

    Second, there's a real MMU available, which even supports multiple instances, and shamelessly extends the addressable space to 2M. Dynamic RAM, DMA components and interrupt management (chips available in the family as well) can add up to a real machine, including hard disk and networking. Things get complex, but it would represent a real computer being able to run a multi-user multi-tasking operating system. The real challenge.

    My formula: 6809+6883+6847 = OS-9 Level-I

    If the winter is long :-) I'll checkout the formula 6809E+(2x?)6829+6850+6821+6844+6854+6845+6828+6840+... = OS-9 Level-2.  Would be a hell of a retro machine! I don't know of anything alike ever built.

View all instructions

Enjoy this project?

Share

Discussions

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates