6809 Single Board Computer project

[jbeeksma]

Hi Ron,

The schematic in your post just contains a small part of the project. Is the full schematic avaiable somewhere? Did you produce PCB's, and are these available?

I'd love to build one!

Best regards,

Jacob

[digsys]

Add one more for the - ahhhhhh wonderful memories !! And I agree with many on this thread, one of the best (or the best) 8bitters of it's time.
Just checked, still have some 6809s and peripheral chips. Had the $12,000? emulator for it, and many other expensive development tools.
Gave the whole lot away (to someone on this forum) a few years ago, including 100s of original data books. Made many industrial controllers using the 6809.
I do remember the last (short lived) generation of TRS80s (?) switched to 6809 (from Z80 I think). I got straight onto dis-assembling the source code :-)
I even worked on a 1st (short lived as well) version of an IBM compatible - before the NEC APC3 mess. Yep, old memories.
Maybe even have all my old circuits / pcbs layouts somewhere, but I think they are bishop graphics tapes on vellum !!!

[grantb5]

Not sure it would be helpful, but I have a 6809 ROM monitor somewhere around here that I wrote long ago.  I think I used the PTM on the NMI pin for single-stepping. It's just a basic monitor but it's (text) menu driven and has breakpoints as well (for code in ram).  I ported it to the 68HC11 ( yes such a big step down CPU-wise) and then later to the Hitachi 6303 for a synthesizer I'm working on (yet another step down ... I think I'm at the 6502 level by now).  Anyway it looks more or less like this and hopefully I can dig it up.  Not sure what assembler I used.

http://musictechnologiesgroup.com/blog/?p=293

[intabits]

I only became aware of this thread when the recent post brought it into view, where the phrase "6809 Single Board Computer" caught my eye.

Some time in the 1980's (judging by the latest date code, about 1985) I bought and assembled this "Uniboard" by "Compacta Incorporated". They supplied just the PCB and a couple of pre-programmed PALs, and the user had to source all the other parts and assemble it. 

It's been sitting in my junk room since then, and I'm struggling to remember if I even ever got it going. I think I did, but I can't picture it actually running, so maybe not. I do remember that U15, the ceramic chip with the badly corroded lid, an MC6844 DMA controller, was hard to get and expensive. The board has some corrosion and general scum, but it does look a little worse in the photo than it actually is.

I don't know where the documentation for this is, I think I've lost it. It was only a few sheets of paper, not a proper manual. A few months ago, I tried searching online for some documentation, but couldn't find anything about it.

Does this thing jog anybody's memory? Any clues as to where to find some doco on it?

I'd like to make a video about it, but that would be pretty dry unless I could also show it operating, and the docs would be a big help for that.
In the meantime, I'd better try to copy those ROMs and PALs...

[grantb5]

You may need to get more eyes on this in order to turn up something. Reddit has a 6809 subreddit and a retrobattlestations reddit, though hard to say who visits those. Maybe the Vintage Computing or Vintage CPU forums would be better.  I don't have links for those last two handy.

https://www.reddit.com/r/6809/

https://www.reddit.com/r/retrobattlestations/

[intabits]

Thanks for those suggestions, I will try places like that.

[grantb5]

That's a really interesting board.  Huge effort one someone's part.  Do you have an EPROM programmer than can archive that old ROM? Looks like maybe a EPROM-microcontroller below that too. Can't see what that is.  Bunnie Huang m(sp?) ight be interested in this  https://www.bunniestudios.com/blog/ .

Anyway, it is mentioned here:
http://www.vcfed.org/forum/archive/index.php/t-54578.html

[grantb5]

There is an ad for it here.  Maybe they did an article for it at one point in 68 Micro Journal

https://ia800605.us.archive.org/35/items/68micro-vol-07-num-04/V07N04_Apr1985.pdf

It's also briefly mentioned in an article here:

https://ia800709.us.archive.org/10/items/the-computer-journal-63/tcj_63_September-October_1993.pdf

[intabits]

Thanks for the links.

Looks like maybe a EPROM-microcontroller below that too. Can't see what that is. 

Nothing like that on it. If you mean U23 (above U14), its an FD1797 floppy disk controller.

I managed to read the monitor ROM (U14, a 2732), and the video character generator ROM (U59, a 2716).
But the two PALs (U29 PAL12L60J) and U40 (PAL10L8CN), both from MMI are unknown to my programmer. Looks like a job for Arduino...
(update: U29 might be a PAL12L6CN)

The only ASCII in the ROM is a few error messages:-

[grantb5]

Sorry, I meant U15.

Excellent work so far.  That's a great photo too.

[grantb5]

Not sure it would be helpful, but I have a 6809 ROM monitor somewhere around here that I wrote long ago.  I think I used the PTM on the NMI pin for single-stepping. It's just a basic monitor but it's (text) menu driven and has breakpoints as well (for code in ram).  I ported it to the 68HC11 ( yes such a big step down CPU-wise) and then later to the Hitachi 6303 for a synthesizer I'm working on (yet another step down ... I think I'm at the 6502 level by now).  Anyway it looks more or less like this and hopefully I can dig it up.  Not sure what assembler I used.

http://musictechnologiesgroup.com/blog/?p=293

I found some of my stuff. 

"6809 & 68HC11 Monitors, Assemblers, C Compilers, RTOS … from the 90s"

http://musictechnologiesgroup.com/HC11/

By the way, the ASxxxx Baldwin Assemblers are still supported and you can download then from (including source):

http://shop-pdp.net/ashtml/asxxxx.htm

[digsys]

The 6809 was one of my favorite processors, I even made severa "industrial" motherboards for it. Unfortunately, I threw out 100s of PCBs a couple years ago,
at a local hamfest. Even all the emulators and dev tools :-( Still have heaps of peripheral chips for it, but they too will go at the next purge .. siigh

[tkamiya]

Wow....

Brings back (found) memories!  6809 was my favorite CPU when it was called a "newly introduced CPU" and "the latest".  What I did was rather small scale.  I've designed an industrial single board and used it to read XYZ signal from a HUGE digitizing machine.  Then spit out the coordinate to a paper puncher.  Only thing was, the speed of which the signal came out was faster than any cpu can catch, so we/I wrote to a memory space physically via DMA and then flipped that memory space to control under cpu to read.

Ported a simple monitor program (firmware) and trig functions.  Fun stuff. 

I liked 6809 over 8080 or Z80 for its simplicity.  Compared to it, intel/zilog series were cumbersome.  I understand 6809 was a reimagined PDP11 where as 8080 was a supersized calculator? 

It was so long ago at near beginning of my carrier.  I don't remember much.

[grantb5]

Byte Magazine
A MICROPROCESSOR FOR THE REVOLUTION: THE 6809
Volume 04 Number 01

https://cdn.hackaday.io/files/460001968064000/byte_6809_articlesx3.pdf

[innkeeper]

trs-80 color computer ... did a lotta programming and hardware hacking on that back in the day.
also did some custom com boards using the 6809.. love that processor.

[coppice]

I liked 6809 over 8080 or Z80 for its simplicity.  Compared to it, intel/zilog series were cumbersome.  I understand 6809 was a reimagined PDP11 where as 8080 was a supersized calculator? 

Its unreasonable to compare the 6809 with the Z80. Its a much later design, with a much higher transistor budget. The 6809 is nothing like a PDP11, but if you want to see a current core which is, look at the MSP430.

[iMo]

I understand 6809 was a reimagined PDP11..

68000 was the PDP11, the asm mnemonics was almost the same.. 68k being my favorite one

[ned_head]

Hi Ron, I just registered on EEVblog after reading your post and all the replies. It's great to know so many people still love this CPU. Maybe it's too late for changes but one thing no one mentioned is the Hitachi HD6309E. They're still available on eBay but I see you're using the on-chip oscillator 6809. I'm not sure if Hitachi made a 63C09P. I've seen one ad on eBay but several for the 63C09E. I mention it not so much for the 2 or even 3MHz capability but for the presence of the illegal instruction trap: something I rued the absence of in the old days when my TCB-108 "Flexi-Plus" board would just hang without me knowing why but strongly suspecting I'd done something stupid on a von-Neumann machine.. I wish Hitachi had been able to announce that (along with the other awesome features you probably know about - 16b*16b signed multiply and 32b / 16b signed divide in particular) but I read somewhere that making it known clashed with their licensing agreement with Motorola. More accumulators, 32bit concatenated registers, bit and block transfers, a bunch of additional instructions and shorter execution times in native mode..
PS. In my opinion, the best thing about the 6502 was the fact that the family of peripherals, often superior to Motorola, were bus-compatible with 68xx: The Flexi-Plus has a 6551 ACIA (with on-chip baud-rate generator unlike a 6850) and 6522 VIA's (with serial hardware and timers).
http://www.6809.org.uk/dragon/6309.txt

[grantb5]

Thanks for posting that here. I was aware that there were improvements but had no idea they were that vast. I'm working on a HD6303 at the moment and it looks to only have a few bit manipulation instructions added over the 6801 family.  Very interesting about the ACIA. The external baud rate chip was always a pain in the butt for the 6850.

I would kill to have a modern flash micro based on a 6502 or 6800 or 6809 or anything similar.  The HC11 is close I guess, but it's not flash as far as I know. HC12 is too oddball.  Maybe an HC908 or something could fill that roll. Something with a decent amount of flash and ram...  oh well, I can use an flash 8051 or ARM I guess. I did that once for an 8021 drop-in.

[rhodges]

I would kill to have a modern flash micro based on a 6502 or 6800 or 6809 or anything similar. 

I programmed the 6809 in the early 1990's and I liked it. That is coming from earlier assembly with z80 and 8088.

Now I am programming for STM8 (notably stm8s103 and stm8s105). I like STM8 architecture a lot. It has two general purpose pointers (X and Y) and a LOT of useful addressing modes, including easy offsets from SP. Want a cheap counter? "label: push #8; ...loop... dec (1, sp); jrne label". The addressing modes are very compiler friendly. There are some very useful bit operations, including branch on bit.

Even though the clock is "only" 16mhz, many or most of the instructions can execute in one cycle. There are multiply and divide instructions. Most of the internal peripherals have one or two dedicated interrupt vectors. Even though the cheapest chips have only three timers, timer1 and timer2 have four and three capture or compare channels for measuring or creating pulses or as general purpose timer interrupts.

An stm8s103 has 8K program flash, 1k of RAM, 640 of EEPROM. They go on ebay, mounted on a breakout board, for just over a dollar. An stm8s105 has 16k flash, 2k RAM, and 1K EEPROM, and you can buy one on a board and with a crystal, for two and a half. A USB programming adapter can cost two and a half. A USB serial adapter can cost from $1 to $1.5 The SDCC compiler is free and works great. I can even compile and program the devices from a Raspberry PI.

You can do a lot of useful things with 8k, 16k, or 32k of program flash and 1k or 2k of RAM.

If you are curious for more, I invite you  to browse my personal and github project pages:
http://www.hodges.org/rh/stm8/
https://github.com/unfrozen

Update: Here is the STM8 programmer's manual:
https://www.st.com/resource/en/programming_manual/cd00161709.pdf

[grantb5]

Wow, that's interesting. I didn't realize other manufacturers made micros with similar programming models. I remember Cosmic from somewhere and would be tempted to try that out versus SDCC, especially if ST has an IDE of their own with compiling and debugging.  With this ST-Link thing for $2+, can you set breakpoints and single-step and such? I see you mentioned simulator too. Thanks for the lead!

[rhodges]

With this ST-Link thing for $2+, can you set breakpoints and single-step and such?

I usually just throw debugging text out the UART, but I have read that the ST-Link can do debugging. I just don't know how deep. OpenOCD might be a good place to start.

I have used the Cosmic simulator, and it was handy for counting cycles. The free version was for Windows, but it ran fine on Wine

[ron.owens]

Hi Ron, I just registered on EEVblog after reading your post and all the replies. It's great to know so many people still love this CPU. Maybe it's too late for changes but one thing no one mentioned is the Hitachi HD6309E. They're still available on eBay but I see you're using the on-chip oscillator 6809. I'm not sure if Hitachi made a 63C09P. I've seen one ad on eBay but several for the 63C09E. I mention it not so much for the 2 or even 3MHz capability but for the presence of the illegal instruction trap: something I rued the absence of in the old days when my TCB-108 "Flexi-Plus" board would just hang without me knowing why but strongly suspecting I'd done something stupid on a von-Neumann machine.. I wish Hitachi had been able to announce that (along with the other awesome features you probably know about - 16b*16b signed multiply and 32b / 16b signed divide in particular) but I read somewhere that making it known clashed with their licensing agreement with Motorola. More accumulators, 32bit concatenated registers, bit and block transfers, a bunch of additional instructions and shorter execution times in native mode..
PS. In my opinion, the best thing about the 6502 was the fact that the family of peripherals, often superior to Motorola, were bus-compatible with 68xx: The Flexi-Plus has a 6551 ACIA (with on-chip baud-rate generator unlike a 6850) and 6522 VIA's (with serial hardware and timers).
http://www.6809.org.uk/dragon/6309.txt

I'm glad to hear this thread was interesting enough to you to sign up. I have been absent for a while working on a Z80 design, of all things. Your post has inspired me to get back to work  on the 6809 project again. I didn't get a chance to work with the Hitachi 6809 parts. I was  Motorola man, back in the day. In fact, I worked at Motorola when they where manufacturing 6809s. That's how I got interested in them. I worked at the plant in Mesa, AZ and the 6809 was manufactured in Austin TX. There was an in-plant class on both the 6800 and 6809 I took. They were mostly hardware, but we got to do some assembly language programming on a D2 development board they gave us. I wish I had it now. It was fairly simple with a keypad and 7-segment displays. I wrote a four function calculator program for it during the class. I'll try to do better at posting now. I should have an updated schematic and board layout soon that i will post.

[grantb5]

I think I used the 6800 D2 board too (in school). I've seen stuff like this for sale on eBay or VintageComputers.com or CPU-World.com.  Our second level class moved onto the 6809 and we wire-wrapped it. I think I still have it.

I discovered a Japanese site, well, group of sites/individuals, that are making SBC's for these old beasts. If you Google and then Google translate SBC6809 you can find them.  I just ordered 10 blank PCBs since that was the minimum, so I'll have those in the USA in a few weeks. Here is one such page:


https://translate.google.com/translate?hl=en&sl=ja&u=https://vintagechips.wordpress.com/2017/12/05/sbc6809%25E3%2583%25AB%25E3%2583%25BC%25E3%2582%25BA%25E3%2582%25AD%25E3%2583%2583%25E3%2583%2588/&prev=search


Also, I discovered a nice 68xx simulator.  I've used several DOS ones in the past, and my own monitor is pretty comprehensive, but I tested the 6303 one and it emulated pretty well.  Anyway, nice in that it actually works, not cute though. Unfortunately NO 6809, but the source code is available. I haven't used the 6809 much recently so I'm probably overlooking some obvious tools. I use the Baldwin Assemblers for a few different CPUs.

[Squiddaddy]

Ahh the 6809!!
I still have a modified Color Computer 3, that supports 8MB ram, and has real serial and parallel ports.
It was one of those things that they called 'NoCan' back in the day, where you removed the RF modulator box, and plugged a board into the memory expansion. The one I have is built around a Lattice chip.
That and a hard drive controller, I have a pretty decent OS9 box. It's all boxed up at the moment, but man, I had a lot of fun with that thing. Built an EPROM programmer for it back in the day to modify the ROMS.
Cool stuff!!!