Unknown 1802-Based System

[NeXT]

I'm bad for this sometimes. 🤪

Occasionally someone in one of my circles of friends finds really weirdo stuff on craigslist and second hand shopping and passes around photos of "what is this"? This thing came up today somewhere in Vancouver.

Some sort of a homebrew thing made out of what I think used to be a word processor? Inside is a pretty bodgy keyboard cable, itty-bitty linear supply and a few boards. 6-slot backplane with a connector arrangement I can't identify, one board labeled "COLOR Machine MAIN" and "COLOR Machine Video". The boards don't look like one-off etches but from the photos I could tell the vias were open so it was hand soldered. Wasn't gonna pay $70 for what could be anything still so I asked if he could sneak board photos.

There wasn't a better photo of the smaller board next to the backplane but I'm making a wild guess from the size of the edge connector that it's some sort of a C2N cassette interface? If that's a triac I see on pin 3 and the markings of a hand cut alignment key between pins 2 and 3 that's MOTOR on Commodore's pinout.
The photos are terrible but I have enough to work with to identify one EPROM, two ram chips, what I can assume is the keyboard interface or an I/O expander, an MC6847 video controller and an RCA 1802. It's a COSMAC system.
Alright he won me over since I've never seen another 1802 based system besides the VIP, ELF and the COMX so it's waiting for me the next time I can make it down to the coast, at which point I can get better photos.
What is getting me is the backplane. It's too unique to be one-off but I cannot recognize it or find it in any of my books yet. Does any of this look familiar to a kit perhaps? Is it a clone of another system? Even a board identification will help me start to look into the proper documentation.

 

[NeXT]

It finally showed up. The system indeed has absolutely nothing I can consider a brand, but it is an 1802. Dating of the IC's places it no later than 1982. From this point onwards if anyone can recognize the kit this was built from (and curiously if the layout of the keyboard is familiar to anyone!) please chime in.

The enclosure is beautifully done. I wish I knew who the last owner is because they did a very good job forming the steel and cutting the sides. It's indeed an extremely custom job but there's no name or signature anywhere.

The boards are all photo etched. There is no markings as to who manufactured them however both the processor and video board seem to refer to themselves as the "Color Machine".

The backplane and power supply is equally unbranded.

The cassette interface however seems to at least denote the various input and transistor pins.

The keyboard is pulled form something else. Doesn't appear to have any smarts. Don't give a toss what type of switch it is because who the hell cares?
Unfortunately the cabling back to the processor board is pretty jank. The red button seen earlier was directly soldered here but I noted all the current wiring points before I pulled everything apart.

Edited: Further searching leads me to believe the keyboard is from a Vydec word processor. Both the 1400 and the 2000 have very similar but not exact key layouts but some of the special keys are in the exact same positions.

 

[NeXT]

The keyboard is from a Vydec 4000 word processor. I'm positive it's from that but this is the only photograph that seems to exist of the machine.

 

[twolazy]

Perhaps dumping the roms would give more information? I tried searching the BYTE archive, but didn't find much due to such generic naming...

 

[NeXT]

I can try an EPROM dump but I've had past issues with 2716's on my Willem. Perhaps this chip doesn't have as many weak bits as the last chips I tried.

 

[twolazy]

Do you have the usb version or parallel version of Willem? I know my parallel version was a pain, esp with the dip switches. The total usb version I have, havent had any issues, using a 12v power brick. Just dumped a 2732 out of a keyboard no problem.

 

[twolazy]

On another topic emptied the storage lockers and still havent found the cable for that Digielec. Thinking might just get some connectors and a breadboard and try to figure out wth it wants. I know what storage bin the orginal cable is in, just cant locate it. Driving me bonkers! I need it running to dump the keyboard controller on a few keyboards. Rather not buy yet another adapter for the Willems' I will only use once or twice.

 

[NeXT]

Nah I have the parallel port version. About a decade ago I bought one and was immediately fed up with how crude it was so I built a far more elegant solution around the board. It's still a Version 5 Willem. Just looks a bit different.

I have both the Digelec 825 and the Data I/O 29B and both have failed me more than this damn thing. I guess that's what happens when the Willem is your only programmer you didn't get for free.
Anyways I dumped the EPROM and immeidately I could see readable text, so we got a good dump and I've regained some confidence in my Willem and my previous work with a 2716 was more a bad chip. I can see cassette routines and prompts. I've put the bin file and a DASMx disassembly listing into a rar file and you can look at it here:

Mystery_1802_ROM.zip

Shared with Dropbox

www.dropbox.com

 

[NeXT]

Had time tonight to clear the rest of the table and inspect the power rails and the power supply. The giant blue capacitor is a 40v 10000uF and it tests okay.
After a check for shorts and grounds I hooked up a little color TV and flipped the switch.

There's something on channel 3. The audio is also giving a strange "beep beep beep beep beep beep" but I can see something in that garbage picture is changing in sync, so it's probably video noise as opposed to some voodoo magic deliberately generating audio in the video's analog circuit. The color killer isn't running either, so whatever the video signal is, that's color. I think the set is too new and the automatic tuning is messing it up. Let me find a black and white monitor.

 

[NeXT]

So after remembering that my portable Sony B/W has a broken fine tuning knob I had to dig deeper and pulled out the ol' Electrohome.
After some fiddling I was able to lock in a picture and it seems to be sitting at a READY prompt.

"MACBUG" does not ring a bell. Google is entirely useless. I'm burying myself in the old Ipso Facto and Questdata newsletter archives trying to find something because we have another mysterious name we can go by.. The easiest way to test what on earth it wants was if I had a keyboard attached but the wiring on mine is a mess with a number of broken wires and when plugged in makes the machine lock up with:

MACBUG
READY 
>:
READY
>Z

Edited: Oh man, the price of small 8-12 inch color TV's went nutty. Ignoring the Trinitrons because of course those are worth silly money people want anywhere between $150 to $300 for something with a manual tuning knob.

 

[DeltaDon]

I did a quick search on the 1802 COSMAC ELF group forum and didn't find any reference to MACBUG either. So it must have been an in house development by the maker of the boards.

 

[twolazy]

I looked through most all the Popular Electronics archive, all of Byte and some Radio Electronics and found zilch. Couldnt even get the rom you posted to boot in Emma 02, the elf/1802 emulator. BTW love the chassis you put the willem in, very classy.

 

[NeXT]

Couldnt even get the rom you posted to boot in Emma 02, the elf/1802 emulator.

EPROM checksum on the good chip is 3D2D. The checksum on the dump I generated matches and the checksum generated separately on the 29B matches three-way, so that should be a good bin file at least.

Edited:
Gee, at this point it would be cool if I could work around the keyboard issues and have some sort of a device to let me interpose the CPU and analyze what is going on but oh right, that's got a keyboard issue as well. (shakes fist at idiots)

 

[DeltaDon]

Is the keyboard putting out serial or parallel data? Or perhaps just decoding the matrix? That should be easy to figure out. If you can get a keyboard working I guess trying "L" or other commands might get some display results.

You are forcing my to dig out my decades long stored Quest Super Elf and see if I can get it running again. I had the S.E. plus the Quest expansion board running with a S100 video card and S100 RAM board. Ran Quest Tiny Basic from cassette before moving on to CP/M machines.

 

[Chuck(G)]

I note that there are read/write routines for audio cassette data recording in that code. There doesn't appear to be a lot in that monitor, judging from the very sparse ASCII content.

 

[NeXT]

You are right Chuck. I completely forgot that if there was a syntax error it wouldn't display it because it's not *there*.

I need to sit down and figure out where the pins on the keyboard socket are going, then try and piece together something from the remaining dozen or so pins that are still soldered to the keyboard. My guess is that if the keyboard socket lives next to a CDP1851 it has to be accepting parallel data from the keyboard. there's small bits of discrete on the keyboard.

 

[NeXT]

So tonight I noted down what is (or was) connected where on the keyboard via the ribbon cable to the processor board, then dismantled it.

Here's what I've got for connecting each pin on the keyboard's edge connector to points on the board. While I removed the solder from the edge connector I noted which fingers had bits of broken wire in the solder so I could get a better idea what *might* of gone where. Most of the wiring I could make end to end connections to were going to the 1851 I/O controller, which isn't a surprise.



The keyboards has no intelligent logic. There's two LS138's, one LS04 and an LS244N.

Trying to find more information the keyboard in the rare chance this isn't a custom layout (Micro Switch SD-16325, Cat. List. 85SD18-1) reveals our good-old enemies over at...Deskthority.....have nothing but praise for the switches, probably because all they'd do is strip out the keys and toss the board.
https://deskthority.net/wiki/Micro_Switch_SD_Series

If someone wants better photos of the board I can supply but I'm not about to strip the topside to find where traces go as everything is soldered in.

 

[NeXT]

And I've verified now that pin 12 on the keyboard socket ins normally a second ground pin but has been isolated by a trace cut and routes up to the 4049 to be conditioned before joining the rest of the bus as the CLEAR pin on the CPU. It's a reset button. (physically in a really awful spot no less)

 

[acgs]

In case any of this helps (and I'm really very UNfamiliar with any 1802 machines, so maybe no help), I've found an old Jameco Electronics ad from the Oct. 1982 issue of Popular Electronics that shows that Jameco sold the keyboard as a part. This was, I'm guessing, a surplus part left over from when the Vydec wordprocessors were discontinued. So, since the issue's date matches the IC's date, the keyboard might have never been in anything other than this kit/homebrew. The link to the ad is here:

https://ia601609.us.archive.org/vie...p2/PopularElectronics1982-10_0099.jp2&ext=jpg

Interesting to note that DeltaDon mentioned his Quest Super Elf. It and another Quest Electronics offering called "Venture" appear in an ad just 2 pages later:
https://ia601609.us.archive.org/vie...p2/PopularElectronics1982-10_0101.jp2&ext=jpg

Also, in case some of us don't know any of the background on Micro Switch, this website has a general overview:

https://telcontar.net/KBK/Micro_Switch/

Info about the SD series products and specifically the keyboards (no specifics on NeXT's keyboard, unfortunately) are here:
https://telcontar.net/KBK/Micro_Switch/SD & https://telcontar.net/KBK/Micro_Switch/SD_keyboards

 

[NeXT]

That's a nice find acgs. 1982 puts it at the same date as the "newest" IC's in the system and confirms the keyboard is from a Vydec.
It's funny how it was one of the cheaper keyboards. These days a particular group now sees that style of switch as the most valuable and most desireable switches available....