• Please review our updated Terms and Rules here

Renovating my 6809 Flex system

DaveH

Experienced Member
Joined
Sep 1, 2012
Messages
173
Location
Scotland
During the 1980's my job involved embedded control using the Motorola 6800 and 6809 series microprocessors so when I managed to buy two 8" disk drives at an affordable ( just ) price I was able to design and build my own 6809 based system.

This has lain untouched for about 25 years until prompted by a post on this forum I dug it out and intend to get it going again.
Its not exactly small or light!, case is 3mm aluminum sheet, the card frame on the left takes up to 6 U4 cards on a custom bus, the 2 drives are on the right and the PSU ( linear ) at the rear, 5,12 -12 and 24V for the drives.

The cards are as follows
Left front, the CPU with 6809 , 64k Dram, monitor and boot EPROM, RS232 for console and parallel ip for keyboard. Dram was cheaper and smaller than static and its fairly easy to use with a 6809.

Middle card, more I/O, 4 x 8 bit ports and a handful of handshake / strobe curtesy of a 6821 PIA and a 6522 VIA, 2 more RS232 ports and a real time clock, the NiCad of which is responsible for the blue corrosion all over centre of the board Grrr.

Front right, display card, HD6845 based with added BBC mode 7 using a SAA5050.
This in fact was the mark 2 card, the mark 1 being a straight TTL,RAM, EPROM, 32 line x 64 char card.

Top Left, Disk interface using MB 8877, I think that's a Fujitsu part, I think Im running MFM but will have to check as it is a long time ago.

Top Right, not a wire wrap card, but also not a commercial but a 6800 card from my work, my design though, only used here for power up reset as I couldn't be bothered to put this on the CPU card !,
The OS was the TSC FLEX 09 OS

Progress so far, PSU run up and test with cards out, ok, cpu and display cards in , nothing !!. I have found my original design notes and drawings so there's some serious trouble shooting coming.
DaveH


Cards bottem.jpgCards Top.jpgSystem unit.jpg
 
Hi All;
DaveH, I am always amazed when someone build their own system !! Great Work !! And that it works is an added Bonus..
"" The OS was the TSC FLEX 09 OS "" Did You write the OS, or did it come from where You worked, or somewhere else ??

Battery Leakage can be a Bear to fix, from the pictures, it almost looks like it affected multiple Boards, But, it could also be the angle of the shot..
Also, check for continuity as these are old Wire-wraps..
Since You have Your old notes, start with what You have in Your notes, and trace out each section.. There could be multiple mis-connects..

THANK YOU Marty
 
Hi all,
Dave,
I was with a company called Osprey Electronics, based in Wick on the north coast of Scotland. We designed and manufactured underwater imaging systems for the offshore market , mainly CCTV but also one combined 35mm and video camera and by the mid 80's had a good part of the world market. Very interesting work, especially taking a project all the way from concept through design, testing (including offshore ), and finally production . The cameras were mostly tube based as CCD's were only just becoming available. On the micro side, one of my 6800 designs went literally to the bottom of the sea. :) The company was taken over by Kongsberg Maritime some time ago.

Marty,
Flex was a very popular OS for the 6800 and 6809 system in the 70's to 80's, I got my copy when one of our work systems was retired and I designed my disk drive hardware to be compatible with it.
Yes Nicads can be a pain, in this case its not to bad as the affected cards are not needed for the basic system , mainly extra I/O and they should cleanup OK, the corrosion is not in fact too bad.

Last night I got a little bit further, hooked up a terminal emulator to the console serial port and got one "FLEX BOOT LOADER" message, then nothing. So its out with the scope time , the one thing Im missing is the card extender, I put it down only 25 years ago and can't find it now :)

DaveH
 
Interesting Dave, I was wondering if it was the same folks who built the 6809 system I sued to have, but I have now remembered that was Fortronic. That had a range of cards including a PROM programmer. I used it for Ham Radio logging software. I now have a couple of SWTPC 6809 system I picked up recently from another guy on here. Not done much with them recently as I have been cleaning and testing my EAI-TR10 analogue computer (see the blogs section), and building a K8200 3-D printer.
 
Hi Dave,that's a coincident, it was Fortronic of Fife that founded Osprey, the intention was that they would do subcontract assembly but that did not work out so they went there own way, this was before I joined them. I know that Fortronic used the 6800 in there products but not that they sold actual systems, that's if its the same Fortronic of coarse .
Found the extender card and uncovered my MSI 6800 system at the same time
 
I assume its the same Fortronic, they were in Scotland somewhere. From what I remember , the system I had was built for the Post Office. Apparently the Post Office then put the project on hold and there was a pile of 6809 systems "surplus". I think at one time I had three. I was probably using one in my Avatar... Then when the Post Office said "we want it" all the stuff was obsolete and unobtainium, so there was a mad scramble to recover all the systems that were lying about, including the ones I had, while a new system was designed. So I lost it. I then got an Acorn 6809 rack system from some one else.... Yes I am a right scrounger....
 
Nothing wrong with a good scrounge, very valuable skill :)
Reseated all the chips on the CPU, did start to spray with Deoxit but it made a bit of a mess, most chip legs were fine but some had a black tarnish and they got the deoxit. Can now get the machine code monitor up most times so PROGRESS.
Found the listing for the Monitor/Boot Eprom, dated 15/3/1987 !
DaveH
 
Last edited:
Hi Dave,
It's great to see this thread you started about your 6809 homebrew rig. I only have internet access for about 4 days out of every two weeks, so just now seeing this thread today.

I had troubles getting my old rig going last summer. The major problems were that several of the 2102L1 SRAMs had failed, and the interface to the console terminal was intermittant. I used memory diagnostics to find the bad SRAMs, except one was a very elusive failure that I finally found by using SWI instructions inserted into my "Space Voyage" game. The intermittant interface problem finally went away went I reseated the MC6820 interface chip.

I will be keeping up with your progress, good luck!
-73
 
Your wire wrap cards look amazingly like mine, I suppose with Vero selling the prototype cards you were limited to what was available.

I got mine running with Flex, but I was amazed that the disc used a 16 bit link forwards and back between the consecutive sectors of a file. Meant you couldn't read more than one sector at a time, system ran like a drain. After being used to the Exorciser with zero interleave I just couldn't put up with it, got binned. Still have the boards but Flex got chucked.
 
Hi Random, So you got your system up and running , nice one, I remember 2102's, 1k by 1 I seem to recall, took 64 to make an 8k byte card, and lots of 5 volts. :)

Hi Woodchips, I don't remember my system being particularly slow, mind you I did not have much to compare it against, only the systems at work. I do recall that the disk did have an interleaved sector format, so that logical sectors 1,2,3,4 would be physically something like 1,16,2,17,3 etc on the disk, giving time between consecutive sequential sector reads. if you have a " missed revolution" you only get 1 sector read per rev and that's going to be SLOWWWW.

Slow is also my progress, mainly due to lack of time, getting a reliable monitor prompt now, changed a dodgy reset switch. Before I try the disk interface card back in I would like to get one of the two display cards running rather than jusat run with the serial console. The 6845 based card really did not show any sign of life so Im working on the TTL/RAM/EPROM , the Test Mode was with the ram removed and the data lines jumped to the lower 8 address lines, so the basic logic is ok as is the EPROM, using the monitor to write to the buffer ram did nothing so the next step is to do some proper write/ read testing on it

DaveH
Some signes of life.jpgTest mode .jpg
 
Hi All;
DaveH, I am impressed with the progress You are making..
I wish I knew enough to make any one of my systems work as well as You are getting Yours to do..

THANK YOU Marty
 
Thanks Marty, we are getting there slowly, Im not really a very fast worker, I have a tendency to get sidetracked. Working on something like this is MUCH easier than an 11, ive not yet got anywhere with my 11/50.

Got both display cards working , faulty bus buffer in the first and a dud LS04 in the clock circuity of the other, complicated by a 'red herring ' caused by an intermittent faulty connection on the extender card .
Going to use the 6845 based card as the monitor VDU routines are set for this.
Next to do is the Green Screen video monitor, been using the bench monochrome video monitor so far, the green screen is again a bit home brew, a monochrome pcb I got from somewhere, modified to drive a green screen tube I scrounged from a rep, all in one of my sheet alloy boxes.

VDU1.jpgVDU2.jpg
 
Hi All;
DaveH, I have a couple of 6" Monitors, which look alot like what you are Showing.. Great Job, getting it this Far..

"" ive not yet got anywhere with my 11/50 "" Tell me more about Your 11/50, Since I don't know what is the Difference between the 11/50 and the 11/55, or for that matter the 11/45..

THANK YOU Marty
 
Hi, sorry for the gap in postings, lots of "unimportant" stuff ( vat return, year end, client computers/systems/software ) getting in the way.

Well after getting the display card to work I looked at the green screen monitor, careful power up resulted in a faint but correct sized raster but no sign of the test pattern I was feeding it. This was in fact a good sign as it indicated that he vertical and more importantly the horizontal sections were ok, line output stages can be a bit of a pain to troubleshoot. Found the circuit diagrams , my goodness I must have been organized in those days, a file for everything and everything in its file, amazing !!.
Started signal tracing at he tube end, easiest to get to , worked my way back "up" the path and suddenly realized that I now had a display, I suspect that I had had a slightly leaky coupling cap , low value electrolytic, and this had been throwing the biasing of, after a bit it had reformed and was now ok.

SO , nearly time to connect the disk drives but first a basic check on them, clean the heads, check the head carriage free, one was a bit stiff , lightly oiled the rails, reconnect.

Power up. insert a disk labelled " sys #3", 3/11/86 " press reset, and WHOPEE , drive activity, an error about the RTC time and data and a FLEX prompt.!! SUCSESS.
Well mostly, next day nothing, oh rats, now what's wrong, looked closely at the drives why is that one not going round, surely there should be a belt on that bit, for some reason drive 0 , the boot drive had lost its belt, easy fix.

So started playing a bit but before long the display started to play up so I'm going to have to go back a step and look again at the display card. This sort of thing is par for the course on gear of this age, its just a matter of getting the MTBF up as you soak the faulty bits out.

Somewhere I still have the ex GPO 300 baud modem I used with the system, any bulletin boards out there ?
Cheers
DaveH

Monitor ready to be worked on.jpgIt works.jpgDrive ready to go back in.jpgFirst boot in over 25 years.jpgOne step back.jpg
 
Bulletin Boards

Bulletin Boards

Hi Dave,
It's good to see your update and the steady progress. I've been checking for updates since your last one. Regarding bulletin boards, I have been curious about that, too. I have a "FastMac v.34" modem for my Mac512K, but nothing either local or toll-free to call anymore. The modem will do 300 baud, I think, but goes up to 19200. I used to use it on America On-Line back in mid-90's, with my Power Mac 7500

-73
 
Hi all, and apologies Random fore not replying to your post, if I find any BBs I will let you know.
Only been able to work on the system sporadically over the last few weeks but we are now basically all up and running after the following work.

Got ready to debug the display card, card in extender, scope warmed up, drawings laid out, heat gun, cold spray and coffee at the ready, spent 4 hours trying and failing to make the card FAIL, it worked perfectly!!, typical.

Replaced the card that generated the power on reset, it was one of the 2 cards that had a NiCad on and it was easier just to put in another card, however I had forgotten that this card also has some address decoding for the I/O space on it and I also forgot that this needed to be disabled. This showed up later when the Flex copy and Format commands failed, due to the I/O space ( E000 EFFF ) also appearing at 7000 right in the middle of the Ram space, easy fix but caused some head scratching.

Fixed the real-time clock, this card also had a NiCad and was the worse one from a corrosion point of view, however most of this was on unused track, the sockets were ok and although some of the wire write wraps did look a bit black ( silver coated wire ? ) whey all checked out ok. I did have to clean the legs of the clock chip, 146818 and replace a 74138 and a CD4011 but then it came up, replaced the NiCad with a CR2032 . This card also contains 2 serial ports and 4 parallel ( PIA and a VIA ), not yet checked these out , I am not to bothered about the parallel but do want one serial as I would like to hook up the printer.

Flex was now booing quite reliably , using a disk marked "sys 3 3/11/86 " so time to generate a new boot disk, id not really tested drive 1 so in went a disk from the " re use " box, hmm, first CAT command ok the next , disk error, not good, looked at the disk, bad score marks. Now I was sure that I had cleaned the heads but I also remembered that double sided drives, this one was, drive 0 is just single, needed the heads cleaned on a very regular basis, for some reason the double sided is much harder on disks, sure enough the heads were dirty, cleaned BUT I think I may change this back to a single sided to be on the safe side. I dug out a spare SS drive, had a label " track 0 sensor faulty" ordered a replacement but so far not had time to fit it so Im still using the DS drive for now.

Format, Copy files, Link, new bootable disk , WHOPEE.

Now the big question, what to do with it now its running. In the day I mainly used it for going on line to bulletin boards and for developing embedded systems using an ICE cable and blowing the E Prom's, all that now replaced by the PC.
However I did find my copy of the Colossal caved adventure game, it was for the 6800 not the 6809 so it wont run however Iv got a partial dis-assembly so the PROJECT is to port it to the 6809, that should while away a few hours.

The System pic show both the system and some of its peripherals, the large box the top of the system unit is the 300 baud GPO modem, the box in front with the 3 green ZIF sockets is a home made E PROM programmer. The ribbon cable along the front is the ICE cable , it connect the system bus to a target system ( card on the right ) for hardware/ software development, its not actually plugged in to the system in this pic as I didn't want to move the cards around to make room for it.
Not shown is the printer, I still have it , its a Logabax 9 pin, built like a tank, weights a ton, and Im not gonna haul it out of storage right now. :)

Now to remember/ relearn 6809 assembler .( and the dreaded text editor )
Cheers
DaveH

The System.jpgReady to Format.jpgWhat modems used to look like.jpgTheres a project here.jpg
 
Super Congratualtions

Super Congratualtions

Hi Dave,
It's great to see that your 6809 system is working again! Thanks for the postings. Its too bad the 6809 cannot run 6800 object code. In the May 1976 issue of BYTE magazine was an article for "Shooting Stars" game, written for Intel 8080. Shortly afterwards I re-wrote it for my MEK6800D1-based computer.

Yesterday evening I posted a new thread in the Software subforum where I attached a text file containing the MIKBUG-format object code. Here it is again.
View attachment SHOOTINGSTARS.txt
I have been meaning to recapture my 6800 source code by copying from a hobbyist newsletter that published my original source listing.
-73
 
Back
Top