I need help finding information on this mysterious computer

#kitcomputer #70sTech #diy
On today's video, I try to figure out what's going on inside this big silver box that was recently given to me. Who made this and is there any hope in getting it working?
-- Video Links
High resolution photos of all the boards:
imgur.com/a/oW...
Adrian's Digital Basement Merch store:
my-store-c82bd...
Support the channel on Patreon:
/ adriansdigitalbasement
Adrian's Digital Basement (Main Channel)
/ @adriansdigitalbasement
-- Tools
Deoxit D5:
amzn.to/2VvOKy1
store.caig.com/...
O-Ring Pick Set: (I use these to lift chips off boards)
amzn.to/3a9x54J
Elenco Electronics LP-560 Logic Probe:
amzn.to/2VrT5lW
Hakko FR301 Desoldering Iron:
amzn.to/2ye6xC0
Rigol DS1054Z Four Channel Oscilloscope:
www.rigolna.co...
Head Worn Magnifying Goggles / Dual Lens Flip-In Head Magnifier:
amzn.to/3adRbuy
TL866II Plus Chip Tester and EPROM programmer: (The MiniPro)
amzn.to/2wG4tlP
www.aliexpress...
TS100 Soldering Iron:
amzn.to/2K36dJ5
www.ebay.com/i...
EEVBlog 121GW Multimeter:
www.eevblog.co...
DSLogic Basic Logic Analyzer:
amzn.to/2RDSDQw
www.ebay.com/i...
Magnetic Screw Holder:
amzn.to/3b8LOhG
www.harborfrei...
Universal ZIP sockets: (clones, used on my ZIF-64 test machine)
www.ebay.com/i...
RetroTink 2X Upconverter: (to hook up something like a C64 to HDMI)
www.retrotink.com/
Plato (Clone) Side Cutters: (order five)
www.ebay.com/i...
Heat Sinks:
www.aliexpress...
Little squeezy bottles: (available elsewhere too)
amzn.to/3b8LOOI
--- Links
My GitHub repository:
github.com/mis...
Commodore Computer Club / Vancouver, WA - Portland, OR - PDX Commodore Users Group
www.commodorec...
--- Instructional videos
My video on damage-free chip removal:
• How to remove chips wi...
--- Music
Intro music and other tracks by:
Nathan Divino
@itsnathandivino

[@ericjuve6885]

I worked at Tek in the mid 70s. to 79. I believe what you have is a “Bit Box.. it was a custom designed development system.that the company used to develop software for embedded systems within instruments. It is definitely a tek design as few manufacturers gold plated all the runs on the board. Also the bread boards are the way we made prototypes.

[@MatthiasWelwarsky]

I was thinking along the same line, the builtin EPROM (EEPROM?) programmer being the clue.

[@dannyhilarious]

That's seems to be more likely and would also explain the existence of the Z280 MPU (not a copro nut a full CPU) with the "modern" 128k RAM board next to the "classic" Z80 board together with the 64k RAM board and the prommer. In trade school, we also had similar machines based on the Zilog line of processors.

[@deansmith4752]

@@MatthiasWelwarsky an Eprom , EEPROM are electrically erasable and were not available for at least 15 years after these boards were made

[@analebanane508]

I completly agree with that conclusion. Some of the wires on boards seem to be interesting: does anyone followed them up to see what they do? I saw some resistores wildly added that was kinda interesting i guess? Some wires seem to be just there to open up more directly communications between modules so .. "for efficiency"? Idk i am a 90s kid XD

[@johndododoe1411]

@@analebanane508 Direct interboard connections are for any signals that aren't important enough to get assigned one of the main bus wires that go to all boards via the passive motherboard (in the original meaning of the word), those bus wires would be just the main power lines and the fundamental data / address / control lines, like on the PC ISA bus (minus the PC specific signals). Probably same signals and levels as the Z80 CPU chip itself.

[@EEVblog]

Classic look and feel of a custom controller for a specific scientific or industrial application. Video video because it's not a general purpose computer, it was likely desinged for a specific control application.

[@ThePoxun]

Its also possible a lot of the bodges are to repair it long after it should have been retired becuase the machine/device it was attached to would have been too expensive to replace but was still in use. I would guess that it was built somewhere around 79-80 and any newer chips were repairs.

[@Rob2]

Well, it depends... my home-made computer stuff from those days (1980-1984) looks exactly like that! I was amazed to recognize the techniques used on the different boards. "factory" boards with extra ICs stacked on top of existing ones, bodge-wired to the remainder of the board. And those prototype boards with enamel (transformer) wire to connect the ICs together, I have many examples of that here. E.g. a floppy controller for 5" and 8" drives, a harddisk controller, graphics card, all made using these techniques. It sometimes makes me sad seeing them (and this system) and realizing how much time and money I spent on it, and what it is now: just an old 8-bit machine. Adrian repeatedly wondered why there is no graphics card, well that was one way of working in those days. Either you had a graphics card, or you had a serial terminal. And this system has many serial ports so it likely was used with a serial terminal. And a parallel printer. It most likely was running CP/M 80 and applications like Wordstar.

[@kevinbroderick3779]

@@ThePoxun You'd be surprised the amount of 70's-80's tech that was used well into the 2000's and beyond. Adrian's critique of the design of the power supply, he didn't test it but I wouldn't be surprised if that power supply still work today. Some devices are not designed for 24/7/365 use without seriously shortening the lifespan of applications for industrial or scientific needs. A piece of monitoring equipment at a refinery for instance might be designed to run for 10 or more years without failure. The power supply probably still works perfectly with maybe the Caps needing to be replaced.

[@beauregardslim1914]

Yep. Looks like lab gear to me.

[@proxy1035]

yea i'd imagine they used some different CP/M Computer to write software for it, and then used the boot ROM and disk controller to load and test that software on that thing. maybe they used that EROM Programmer card to write the boot ROMs in the first place

[@mikeselectricstuff]

I doubt it was "home" made as it would have been pretty expensive in those days, especially the PCBs. I think a prototype or 1-off for some kind of industrial/scientific use, or maybe a college.

[@ionstorm66]

Yeah even the protoboard cards would of cost a pretty penny.

[@jeffreyphipps1507]

It would depend on the actual date. I made circuit boards in the 80's using photographic processes (PITA). However, how systems tended to be wire wrapped. Schools made systems like this (well, students did) in order to have computing resources. Seems like early 1980's

[@crashoverride328]

Actually the copper clad board / prototype boards were cheap, homebrew was done sometimes with etch resist marker or tape, then etched with ferric chloride at home. Double sided board not much more. There was much more emphasis and encouragement to homebrew. Companies like Heathkit, Velleman, Radio Shack, magazines like Popular Electronics etc would give ideas...

[@rdmclark]

I am offended by home made term used here, its definitely hand made. Its a work of art, too neatly wired to be a home made project

[@maltronik]

@@rdmclark I agreed.. the casing look rack mounted.. definitely not home use...

[@Zekrom_64]

Correction: The "Z80 math coprocessor" is actually a Z280 CPU; it is binary compatible with the Z80 but is *much* beefier, including instruction pipelining, a built-in cache and MMU, and more complex instructions like hardware multiply/divide, similar to an 80286 vs an 8088. Being Z80 compatible it could probably run CP/M but considering the rest of the hardware it looks pretty custom. My guess would be the Z280 is the main CPU and the other CPUs on different cards act as coprocessors for other subsystems.

[@drj113b]

That explains the 128K of RAM, as well as the fact that the original Z80 CPU card was in the 'storage' area of the machine.

[@ieatspacemonkeys]

Looks like a job for a z80 disassembler of the ROMs. Then start writing an emulator for the system that at least implements the bits that cause the code to block and get the serial output running. After the system is running in simulation, you can use co-simulation to recover each card using a FPGA to drive the bus in substitution of the processor board. But life is short, so just scan the ROMs for copyright comments and be done with the thing ;)

[@lucasrem1870]

he is crying boot disk, thinks it's IBM gaming system, mad....

[@adriansdigitalbasement2]

Indeed! So I'm thinking that the Z280 is the actual CPU for this as the Z80 CPU card was not installed any longer after an upgrade

[@adriansdigitalbasement2]

@@ieatspacemonkeys I'm going to dump the ROMs and upload them -- perhaps the Z80 assembly experts could see what they do. I'm certainly going to try to power this up too :-)

[@stephensanner1315]

It's an STD Bus machine ( en.wikipedia.org/wiki/STD_Bus ). Edit: The board with the four 40-pin packages and which is connected to the ZIF socket board appears to be a Mostek 3870 system ( www.cpu-world.com/CPUs/3870/index.html ), which is a direct derivative of the Fairchild F8 family ( en.wikipedia.org/wiki/Fairchild_F8 ). Wild! Edit 2: Z280 is a 16-bit extension of the Z80. TIL! ( en.wikipedia.org/wiki/Zilog_Z280 ) More edits: - The 6402 is one of the earliest single-chip UARTs I'm aware of. Both 6402s are associated with the expected level shifters for RS-232, in both cases in the form of the little 8-pin SN75150 driver on two outgoing lines and the SN75154 receiver for four incoming lines. One line in each direction will be for data (TX and RX) and the others will most likely be modem status/hardware flow control. - So there are... what, three separate CPUs in that thing? Four? And at least three different architectures too (F8/3870, Z80, Z280). Wow. It's a beautiful machine. Those handmade cards are works of art. I'm glad you were able to rescue it.

[@LegalizeAdulthood]

The connectors don't look right for STD bus, though. The fingers and spacer look wider than STD bus cards. It's unclear if the connections on each side are electrically connected in the backplane. If we assume they're not electrically connected in the connector, then it's got 56 connections total which does match STD bus. Maybe it's STD bus electrically but not mechanically?

[@lindoran]

I'm beginning to think this might be part of a collection of separate STD bus computers with cards just being stored in this chassis to keep them safe? It does seem odd that there are several cpus, unless the toggle was to select from the 3 cards?

[@zbradbell]

@@LegalizeAdulthood Pin 1 of these is ground, so it doesn't match electrically either

[@SomeMorganSomewhere]

Yeah, the many different processors is super weird... It's like there are three or four computers all crammed into the one box. I suppose given the somewhat ridiculous number of UARTs in there this is maybe not a stupid explanation? And yes, those handmade boards are GORGEOUS, in all likelihood a complete nightmare to troubleshoot but gorgeous none the less...

[@phirenz]

It's not a single configuration. Based on the date codes, I suspect the F8/3870 was the original CPU that shipped in this machine. The ROM board appears to be of the same vintage, the PROM programmer and maybe the 24k ram board, though it looks slightly newer. maybe the first SIO board. All chips on these boards date to 1979 or earlier. Then the Z80 board was added on later, maybe in 1981?, along with the 64k dram board and the CPM ROM chips, and the machine was converted into a CPM machine. I think the Floppy controller, and RTC/serial/FPU board dates to this config. Then the Z80 board was removed, that's why it was floating around the case, along with the CPM rom, and was replaced with the handwired Z280 sometime after 1986. The handwired 128kb of SRAM probably also comes from this era, and the serial/parallel/timer board. I suspect this is when many of the other boards got most of their bodge wires. I suspect this chassis actually carries two reasonably independent systems at once. The original F8/3870 machine, and the Z80 or Z280 machine. Not sure if both can run at once, or interleave execution. Or if one of the toggle switches switches between the two systems.

[@Mrshoujo]

This seems like CuriousMarc and his friends could decipher this box, but they're busy working on Apollo microwave circuits, the blackest of black arts in analog electronics.

[@BilisNegra]

Love the reference, and the quote, too! Greetings, earthling!

[@Tedd755]

@@BilisNegra *Greeting

[@Mrshoujo]

@@Tedd755 Don't incorrect his correctness.

[@Tedd755]

@@Mrshoujo You both need to listen better. Chosen at random: kzbin.info/www/bejne/jIPIloOHf7eFf5I

[@djmips]

@@Tedd755 True. 😄 It's 'Greeting Earthlings'

[@Gunbudder]

what a trip, this box looks like it came right out of my old lab. i worked a LOT with boxes very similar to this one. Its early 80's, late 70's so i'm not surprised it isn't covered in asset tags, but i am a bit surprised it doesn't have a single company mark on it. the boards were almost certainly rolled by a company with their own pcb manufacturing, like any aerospace from that period. Most aerospace firms plastered their logo on everything they rolled though, which makes me think this isn't aero. It's definitely not "homemade" though, and has all the hallmarks of a professional research lab. In general terms, i'd put my money on this box being a control system. I'd trace back the ribbon cables and see if one of them is actually DIO or not. Floppy drive cables got used a lot for DIO because why not; you already have a thousand of the cables. I would also be pretty shocked if it had any visual output. Even today, that is exceedingly rare for this kind of hardware because its a beyond a waste of time and effort to add video out of any kind. I'd take a close look and see if its RS-232 or RS-422 though, that would be a big clue. basically, different industries use different serial methods (more or less). If it is a control system, or a test bed for control systems, then it probably does have a boot loader or boot boot loader output on the serial port when its powered on. I found it pretty common to get at least something from serial ports at each level of firmware because it saves you a ton of work to quickly see what is going on (is it working, what version, etc). The fact that it seems to be completely Zilog based is also interesting. That makes me think this box was used as a test bed or for R&D for some consumer product, or at least not a government only product. I'm a lot more familiar with 90's junk, but i don't think Zilog stuff saw a lot of use in aerospace or similar areas. This video makes me miss my crusty dusty lab with its mix of 90's and modern hardware, all custom made with 80's design like the box you have! See if you can dump those roms, and then look for ascii text baked in (with all offsets). If there is something like version info for a boot, it will be in there stored hard coded

[@Drmcclung]

That's exactly what I was thinking too, that it's a control system and you're right, totally custom. My mind automatically went to plotter/printer control because I've worked with a lot of controllers full of bizarre cards and handmade/custom circuits just to get whatever machine/software the company was using to work the print/plotter/press etc

[@jagmarc]

I used to design this kind of thing for aerospace engineering, R&D and manufacturing, I had outside companies making the custom pin & wire boards, and sometimes I had internal resource to build them. I wouldn't have time to build any of it as I was fully occupied documenting everything. The mods like added 0.1 capacitors and cut tracks/added wires I would had done myself.

[@TheOleHermit]

Yes, but I would only point out that back then, about the only choices for 8 bit dev CPUs were either Z80's or Apple's 6502's. It was also common to boot from 5 1/4 floppy drives, instead of from RAM, w/ a bootloader. My initial guess was a laser controller, with a bit pad interface and a RS-232 dumb terminal. But, this looks more industrial, like an early CNC machining/robotics controller.

[@abergethirty]

It could be custom instrumentation for a test lab. I used build stuff like that for engineers in an Automotive Safety instrumentation lab. They would hand me drawings and I'd create it all from scratch, even the boards.

[@mikelunsford2587]

Allen, you're the bomb!! Let's see some of your stuff too

[@chriskukowski398]

I just worked on a PC like this where I work. I can second that this looks like some sort of testing unit. So a company produces an assembly then hooks it up to this and tests the product. Multiple ports means multiple tests running at the same time. Result of pass/fail could just be a series of beeps or blinks.

[@liam3284]

The large connectors remind me of a HP test equipment interface.

[@robertoricardoruben]

I think about the same, specially because the power transformer is marked "Tektronix". I would say it is a standard instrument customized into a test rig

[@samcross1833]

Greetings the device you have is part of ATE system bulit by a company I worked for in the early 80's in salem oregon. It was bulit for one of their customers on contract.. please contact me for more info.

[@adriansdigitalbasement2]

Very interesting! Could you email me at the address on my channel about page? Would certainly be neat to hear more about the history of this thing. kzbin.info/door/btwi4wK1YXd9AyV_4UcE6gabout

[@strangersound]

Those hand made cards are a work of art. :)

[@EsotericArctos]

This is definitely a really bad thing if you built this for a customer. Using parts from TekTronix equipment and dodging a power supply board? This is dodgy to say the least for something that is meant to be sent to a customer. I hope this one was more a prototype and that all those corrections done by bodge wires were reprinted into a proper PCB for the actual customer.

[@redryder6987]

@@EsotericArctos Why bother? If you are only going to make a dozen or so of these, having new boards printed out every time a revision is made get's ridiculously expensive, especially in the early 1980's, you couldn't get ten boards printed for $2, and it would also certainly explain the wire wrapping. As long as it works, who cares? It's not like this is an Apollo guidance computer or it's memory core, and I'm sure that there were still bodges on those to. Power supplies aren't a huge deal either, I've seen worse in crap ordered off of Amazon and in those Keurig coffee makers, power supplies that can literally kill you.

[@EsotericArctos]

@@redryder6987 This would not be a very cheap coffee maker from Amazon for one thing. Even in the 80's, if you were paying for something custom, you don't want something that could risk burning your office down, and I can guarantee the price of this, being custom made, would have well and truly covered a proper PCB design, given they were likely hand made in house for such small runs :). Surely you wouldn't expect a second hand tektronix transformer and dodgy power supply in a commercial product, even if it is custom and from the 80's. That's why I feel this one could have been a prototype.

[@wh1tewolf4]

It’s not a homemade computer. It’s in a rack mount 3U or 4U box. The boards are made in house to support the various test equipment/instrumentation. I worked in the oilfield industry and as various instrumentation was being converted to the digital age, it wasn’t uncommon to make in-house boards to bring the various instrumentation we acquired from all the acquisitions during the oilfield crash in the early 80’s. The funny soldering are field repairs we made in field because we couldn’t drive back to the base camp as we were on customer time. We pulled the rack mount, opened the top, slapped replacement boards or started soldering jumper wires. The braided soldering wick ground wire was just some guy panicking, and slapped that as a strong ground. We ran off generators that would surge or during lightening storms and not fully grounded, so I think they panicked and used that. We had humidity problems as our mobile office is a peterbilt unit with a box office and on hot days in West Texas, the AC and the heat from the rack mount systems didn’t play well. Good times then. Nothing is homemade, it’s propertary boards.

[@jxh02]

Card ejectors are what's missing on the corners. And yes, that is enameled magnet wire used for rework on the back. Mil-grade stuff had it all embedded afterward in epoxy so this is likely not that. They would also have littered everything with identification and serial numbers, also in epoxy, each numeral applied with a tweezer and a little rubber stamp. In yellow.

[@MarkEichin]

yeah, but you need holes in the board to attach the ejectors, and I didn't see any on these...

[@1944GPW]

Extremely similar to pieces of equipment I have that were obsoleted from radio telescope installations. They were rack mount, custom made with bespoke proper tools such as panel punching, nicely engraved fascia panels, custom PCBs but hand soldered/soddered, good nylon card guides, cable hold-downs, star washers under screws etc etc. They are well made because they had a government budget to do it, and it had to live a long service life.

[@bcostin]

Tech Time Traveler might be a good resource for deciphering a machine of this era. It feels like it would've been used for some sort of scientific or industrial control application to be. It looks pretty clean so maybe not a shop-floor system. Then again, it could've been tucked away in a clean environment and just operated from a nearby terminal. The newer bits obviously suggest that it was actively maintained and expanded for a long time.

[@adriansdigitalbasement2]

Thank you (and everyone else) for recommending his channel. I've starting binging his videos -- really good stuff! I sent an email off to Brad although he might have my problem where it's email overload :-)

[@Ampersand100]

@@adriansdigitalbasement2 Yeah, that guy's videos are really well made! It seems like he should have a lot more viewers.

[@kalilay]

"some sort of scientific or industrial applications"............ uhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh what else could it possibly be?

[@clifffton]

@@adriansdigitalbasement2 They actually covered it "The Digital Group"

[@Steven_VE9SY]

My first job was in a computer store starting in the latish 70's. All that wire wrapping was original From the factory. When IBM was selling for $10,000, these sold for around an easy $4,000ish. I couple of the boards are home brew. We sold the blank cards. Unfortunately I'm at that age my memory is failing and I can not remember muck about them. I sold 3 in my first year. All to the same guy. He was a Dr, and set up his office. One was a server and it was headless. The other 2 both had video cards. It used the old green or orange screens. I may have some documentation in my filing cabinet, but it is in storage, and right now in New Brunswick it's COLD!! Especially for this old fart! I plan on going down there sometime next week or the week after to bring a few things home and dropping some stuff off. I'll take a quick look through the cabinet, and if there is anything there I will male it out. I have cleaned out a lot of that stuff as I never thought I'd see the day I would wish for the old slow machines!

[@nuttyjawa]

So - at one point in my career I worked at a Company that produced alot of high end sensor equipment, and they had to support said equipment for a life of at least 25 years, they had alot of this sort of kit around - none of it was my responsibility so I barely got to interact with it much Long story short they had alot of sensor calibration gear in ancient rack mounted machines, in cases just like those - with machines built 20+ years ago either in house or typically from small engineer sector businesses that would sell some kind of bespoke equipment that would come with a machine as part of the rig. Mega obsolence in place, if there was ever a problem with these they'd be out of order for a long time whilst guys who had no idea what they were doing would try diagnose because the guys that used them in anger were long retired. They'd have serial coming out of the senors, into these machines and out again at times into other hardware. I'm talking 2010s with even Sun servers and VAX terminals still being actively used Whilst I couldn't/ wouldn't say that is the purpose of this machine, the similarity is pretty remarkable

[@absalomdraconis]

I once heard of an original IBM PC being used to control a piece of production equipment in the 2000s. Apparently, the equipment designer had looked at the extra ROM sockets on the PC, thought "I'll use that!", depended on the bus frequency as well if I recall, and the company never saw a point in replacing the system because it would be cheaper to track down another original PC.

[@timmooney7528]

@@absalomdraconis In 1997 I worked a temp job in IT at a printing company. First repair job was on an 8088 desktop in a cabinet under a binding machine. A pressure switch on the output tray would fire a relay on a custom circuit board, and the pc recorded copies coming out of the binder. It ran DOS and fed a monochrome display. In 2018 the car manufacturer I was working for had robot arms which still ran on Win98. It was what shipped with the arms when they were bought in 1999. The tire changing machines ran on Win2000.

[@fonkbadonk5370]

@@timmooney7528 I have seen a DOS machine (a 468 no less) no one was sure what it did (other than knowing it was something central and important) at a major paints manufacturer's production plant as late as 2015. The dude that made the custom software for it had long retired and moved across the country, and the disks that once contained the sources had been unreadable for a long time before I set foot in there to top it off. We were hired to assist the new guy in charge of the department in something unrelated to this machine, and he was clearly stressed out about the state of things - with good reason!

[@TheSimoc]

@@absalomdraconis If it ain't broken, don't fix it. At least never replace it with modern crappy hardware and bloatware.

[@5Hydroxytryptophan]

@@TheSimoc If the things are in this state, you better fix it before it hits the fan. You can let it run, but you really want to have some drop-in replacement as backup. A lot of this old stuff can be replaced by a cheap microcontroller these days. No need for bloated software. Unfortunately most programmers are not that good, so they use what they know and you end up with an Java-Application on Windows.

[@twol78s90]

For development work at Tektronix without use of an operating system on the board bucket, the Microprocessor Development group also had a skilled software team that developed a nice Macro Assembler, Re-locating Linker, and downloader toolset that ran on the Tektronix Control Data Cyber 73 timeshare system running KRONOS. Most of the toolset was written in PASCAL, and worked very well. The assemblers for the different processors were made to be as consistent with each other as possible. The team also wrote great documentation for the software tools they made, that made it quite easy to come up to speed on the use of the tools for developing code for the system. This was the environment most frequently used for doing software development for the board bucket systems within Tektronix. The earliest versions of the Board Bucket system were used as development systems for what became the famous and rather incredible (for the time) Tektronix 4051 all-in-one BASIC-language Graphics Computer. The 4051 was the first of its kind, providing a cable Tektronix-written dialect of BASIC in ROM, augmented by the ability to do high resolution graphics, drawn on the systems' storage tube display. (kzbin.info/www/bejne/rojVeoaVoZp7jc0). The success of using board bucket as a development system for the 4051 made them quite popular as a development platform for all kinds of things within product groups at Tektronix. The board bucket circuit boards were hand laid-out by a very nice lady named Dixie who worked in the Microprocessor Development Group. It was amazing to watch her working with tape and Mylar on a big light board to lay out the pads and traces for the boards. This was before there were CAD systems were commonly used for this kind of thing, although Tektronix did have an early REDAC printed circuit board CAD system based on a Digital Equipment PDP 8/I system down in the basement CAD laboratory at the main Tektronix campus, but this system was typically very busy doing production circuit board layouts. Dixie was a real artist. The art of her work is evident in routing of the circuit boards. Most of the boards were two layer boards, with a sandwiched ground plane, and etch on the top and bottom of the boards, with plated-through feedthroughs. Dixie was well-known throughout the company as one of the best hand-tapers around. The boards she laid out were always beautiful to look at, but as electronically sound as could be made. It was an extreme rarity for one of Dixie's circuit boards to not come up when a prototype run was done and a board stuffed with parts. Almost every time the problem proved to be in assembly, not having to do anything with the board layout. As much as I can remember, only once was a second-run of a board required due to a layout issue resulting in some kind of noise problem that took quite some time to track down. Tektronix' ECB (as they were called there, for "Etched Circuit Board") manufacturing plant was legendary for creating some of the best circuit boards in the industry. Gold was heavily used, even on internal engineering boards, and the quality was impeccable. Even the internal engineering boards were produced as if they were going to go into a Tektronix product, making them vastly superior to most all of the hobby computer S-100 boards on the market. And, in the end, they were a considerably less expensive than comparable boards in the hobbyist marketplace if you built them out yourself. If you opted to buy the boards fully built-out, they were still less expensive than an rough equivalent S-100 board in the hobby marketplace. I did not work in the Microprocessor Development Group, but because of my interest, I became quite close with the members of this group, and spent a lot of my free time over in their area hob-nobbing with the folks there. I learned a great deal from them in terms of digital design and microprocessor interfacing. I purchased a number of the unpopulated boards and built them up myself to make a small system to play with at home, which grew over time. Later, as a "spare-time G-job " project, I was "signed up" by the Microprocessor Development group to develop a custom system based on this hardware to serve as a wafer tracking system in Tektronix' IC fabrication facility(later sold to Maxim). I designed a video board based on a Texas Instruments video controller (can't remember the part number, but it was compatible with TI's 9900 CPU and required a little work to interface it to the bus). The board would hook up to a large (CRT) TV. There were a number of these systems planted around the fabrication areas, mounted so they were easily readable from just about anywhere in the area. The systems all talked to each other over a syingle RS-232 line where I used a time-slicing protocol so each system got a slot of time where it could talk to the others to relay status information. Each system had a CRT terminal attached via serial port where operators could enter information and query the system. The video board also had the National Semiconductor real-time clock chip on-board along with battery-backup. I also modified the floppy controller with a daughter board that plugged in place of the NEC floppy controller that provided the capability to used double-sided, double-density 5 1/4" floppy drives so I had enough disk space for the data the system needed to store. The system used the 6809 CPU (my favorite 8-bitter). It was packaged in a similar chassis, and used Technical Systems Consultants' Flex-09 operating system, which I wrote custom drivers for to talk to the floppy controller, serial ports, and video display subsystem. The system was prototyped and tested out with a couple of select folks from the IC fab, and they made a few recommendations for some changes, which I incorporated, and then it was installed in the Fab. It worked great, and really improved the efficiency of the fab's production workflow. The system that you have uses a different bus structure, but the concept is very similar to the board bucket. I think that the board with all of the 8?-square-pin connectors along the top is a multi-port serial board. The sockets might be for async interface chips.

[@jessiec4128]

I had a job regarding computers. My boss sent me to a company to troubleshoot a system. They had a guy who worked there, but retired. When I arrived i quickly found it was Pascal. I found out what they needed, and what was not working. I found the areas and had to add a few lines to help it run. You should have seen the Boss there. He gave me a hug and said, they have been wondering when someone could help. And it took me only about 20 minutes. It was not hard and I just had to fix a few lines.

[@thisolesignguy2733]

WOW! I haven't seen a Bus machine in years! Back in the late 80s, I worked at a brick factory where we had a packing machine that used an articulating robotic arm. It used a full array STD bus to run the software. As the bricks came down the line, the arm would swing around and grab the bricks than pallet them up. It only had a maximum of 12 complete movements and 3 pinch positions, the bus array was as tall as me and had 12 total boxes. Gawd, they were a pain in the arse when one went down. We kept extras in a storage room because they had to cleaned and maintained constantly, they would suck in dust and the circuits would get damaged or burnt. I even remember one catching fire when a capacitor popped...You'll never forget the smell of burning circuit board and brick dust, it smells like burning hair ugh. Great find! this really brought back memories...both good & bad lol

[@pride293]

lol bro

[@arjovenzia]

Haha, yeah, I did 6 months working in the packing plant of a Titanium Dioxide plant. its the stuff that makes your toothpaste white and abrasive. the SCADA stuff was pretty up to date, but for whatever reason, the printers were all apollo era. they printed the packing slips that went with each consignment. apparently there was a bigwig who liked paper dockets. so, the whole place ran on paper. unfortunately, hyper fine abrasive powder is Bad for printers. If one lasted more than a week, it was a miracle. but the whole line would grind to a halt if you couldnt fix it in about 20 minutes. Lots of hot spares. when I started, took me over an hour to clean n refresh a printer. my Personal Best by the time i was done there was 7 minutes. other parts of the plant used either digital systems, or ink blasters. but packing used the ol bangers. the buggers kept working tho. you can love, hate, and respect a thing in one sigh.

[@cryptnotic]

The EPROMs might contain clues as to the source of this machine (e.g., a message printed to a serial console).

[@oddvertex9429]

👍 good idea

[@gorak9000]

Yeah, I mentioned in another comment, read the eproms and run "strings" on it - likely some juicy tidbits will fall out. This is long before evil "firmware encryption" - all the clues should fall right out

[@8bitwiz_]

Yes, dumping them would likely reveal something about the company that produced this system. But maybe not all that much, because it ran CP/M, so everything important would be on floppy disks.

[@alexanderstohr4198]

Eproms might have lost their memory over the ages. But i wish you good luck that they are still data intact.

[@Ariannus]

Being a rack mount system I find it extremely likely that this would have just used a serial terminal and not had a display adapter

[@jpnc2319]

It operating system could be a dos or cpm specific or maybe a type of basic.

[@squirlmy]

@@jpnc2319 If this was scientific or industrial, it probably would have been programmed with a more efficient language. You seem to be thinking in terms of "microcomputers/PC" and this isn't one of those. Highly unlikely it used DOS, and considering CP/M is more of a monitor than OS, I'd look in entirely different direction, perhaps even a custom "OS". Looks a little too primitive for unix-type OS. But this was from a time before "PC standards" existed.

[@scottcall7207]

I had an S100 based Z80/CPM system loaned to me in Jr High (it was out of date then) by a teacher who was happy I was interested and even though it was a "desktop" it was connected to a serial terminal.

[@kaitlyn__L]

@@squirlmy yeah probably a basic RTOS, likely with less functionality than we’d expect from modern firmwares. Could well have been custom depending on the client. Edit: if indeed industrial. Another commenter has an uncommon, expensive, custom built Z80 home machine from the time and mentioned CP/M and WordStar on theirs.

[@gorak9000]

Being rack mounted and from the 70's, probably some custom built or custom modified machine for a particular scientific or industrial application. I see no need for a display if it runs some code from the ROM chips, and then talks to some other devices over rs232. It could also be an early interface between an old CNC machine that expected to get it's program from a paper tape reader, and instead this thing transforms it into a disk-baed system. Best bet would be to dump the roms and run "strings" on them to see if there are any clues in the firmware.

[@gordon9042]

The whole package looked like something out of Tektronix so I was pleased that it had the Tek label on the transformer. In a previous life I worked at Tek and took a leave of absence in 1973 to go to the Oregon Graduate Center ( later to become the Oregon Graduate Institute and eventually merged with Oregon Health and Sciences University) to construct a special piece of test equipment. When I completed that project I decided that it was more enjoyable than my job at Tek and stayed there until retirement in 2008 at 70. The prototyping boards looked like some we used at Tek and it seems possible that it was prototype for a piece of test gear. When I was in the component evaluation group at Tek, we would often have to build a one off piece of gear for testing some particular component or group of components. I later moved to the CRT instrumentation group where we designed and built test equipment for the CRT manufacturing department. In that group we usually made custom circuit boards since there were typically multiple pieces needed for the large number of test bays for the testing process. The wire used for those hand built cards looked like it was transformer wire which seemed logical since it was easily available from the transformer department and was not going to have insulation melting problems to dal with. I am afraid that I have no idea what was built for and could have easily been someone's home project. A lot of us built our own home projects using our experience from work and that build looked like something I might have tried. Hand manufacturing such equipment just seemed normal for a lot of us.

[@mikeenkelis8522]

This is classic Tektronix R&D type of development computer. Lots of gold plated circuit boards, and the crimped wires were the normal engineers method of design. I was at Tektronix from 1980 until late 2000, and have seen plenty of these during my time there. Could have been used for prototypes of there graphics terminals or Phaser printers.

[@frugalprepper]

The security was really good on the Star Trek Computer Too. "Captain, Someone is accessing the Transporter" "Lock them out Mr. Worf".... Guy in transporter room opens a unlocked panel and moves two control crystals... "Captain, he has over rode the lockout and transported off the ship."

[@jfbeam]

There's not a lot one can do to stop a hacker with physical access to the hardware.

[@timmooney7528]

@@jfbeam I never understood how security could be so lax on a starship. When I worked in nuclear power production, there was no way a random person was going to walk into a restricted area such as a control room. All critical doors were monitored and lock through access control, and attempting to enter a door you didn't have access to would result in angry security guy showing up. Even in a hospital if there is an alert about a missing or wandering patient, the staff can lock the place down pretty quick.

[@jfbeam]

@@timmooney7528 In the Star Trek future, everyone is trustworthy. Picard says something to this effect in the S1E25 / The Neutral Zone. (Data finds three people in cryo.) Everyone knows the rules, and almost always obeys them. Don't forget about the incident where they even shutdown power to kill the transporter, but the guy used a phasor to power it directly. (S3E11 / The Hunted)

[@firesurfer]

Even in 1967, I was wondering why there was no security for the transporter room.

[@timmooney7528]

@@jfbeam Time and time again has shown that every guest on the ship is not to be trusted.

[@BigDaddy_MRI]

My Z80 system from Heathkit used a terminal that emulated a monitor system (VT-100??). 2 8” DSDD floppy’s, and 2 5.25” DSDD floppy’s. I think it was the H89. Ran CPM and later MSDOS. It was a workhorse and I used to write Z80 Assembly code for Z80 projects I used to wire wrap. The PIO is parallel input/output (2 8bit ports) with printer handshake but also highly configurable for bit input and output. The Serial Input/output chip (SIO) is a dual port serial (both ports can be asynchronous or synchronous) with modem handshake. The Counter/Timer IC (CTC) is used to generate the clocks for the SIO baud-rate generation. The CTC has 4 independent counters or timers as needed by the hardware. Precision outputs for different clock frequencies. Love the Z80 family of chips. Very well engineered.

[@pauldunecat]

z80 mini computer running CP/M, I used to go to a video store for rentals, and they ran one. This was in the late 80's, I seem to recall it used a serial terminal as the console.

[@solitairepilot]

Well, that was quick 😂 (EST Timestamp 17;42 March 10th, 2021)

[@chouseification]

yeah I was going to suggest something similar as well... point of sale system or similar. One serial for modem, one for console terminal, one for serial bus replicator to talk to POS terminals?

[@pauldunecat]

@@solitairepilot Well, may not be this exact thing, there were alot of these out there back in the day. But I found it funny that it was still running in the early 1990's, I guess since re-doing the catalog would have been a hassle, even if you could technically do everything in DacEasy database. :-D

[@pauldunecat]

@@chouseification serial dot matrix printer! :-)

[@chouseification]

@@pauldunecat good point - each POS terminal would have needed one of those bad boys for receipt printing - double copy with "carbon" often.

[@bobpurdon69]

Z280 is a 16-bit version of a Z80, so I'd speculate the actual Z80 card was the original, and the machine was upgraded to a Z280 and the old board just left floating around in the case.

[@TomStorey96]

Hey Bob. Fancy seeing you here 😄

[@tassiebob]

@@TomStorey96 Hehehe, Adrian's channel(s) are on the must watch list :-)

[@Peter_S_]

That EPROM programmer's processor card is based on a Fairchild F8 CPU which is made up from those four 40 pin chips; a Mostek MK3850 ALU, an MK3852 Dynamic Memory Interface (DMI) to control additional RAM or ROM (or EPROM), an MK3854 DMA controller for the MK3852 connected memory, and finally a MK90071P a.k.a. MK3851 Program Storage Unit (PSU) which contained 1 KB of program ROM and handled instruction decoding. The MK90071 number on the MK3851 is the ROM mask number for the EPROM programmer code. The Z280 was to the Z80 like what the 65C816 was to the 6502, but much more so. It looks like two complete separate systems in there with the EPROM programmer as a smart peripheral. Everything is serially interfaced and the EPROMS on the card likely contain a very basic ROM Monitor program for developing firmware on the Z80 system. The floppy controller PCB also contains a 1K PROM, possibly for boot code. I could tell more, but you keep moving the boards around randomly and holding them at angles so it's terribly hard to see.

[@adriansdigitalbasement2]

Interesting stuff. Check the description, there is a link to high res pictures of all the boards, front and back.

[@rogeratygc7895]

Having worked in steel industry research and development from 1975 to 2013, I feel this has the look and feel of the sort of the sort of equipment we might have made (we didn't, I'm in the UK). A wide range of dates on the I.C.s and the modifications suggest it might have been made for one purpose/project, and rebuilt for another. In the 1970s and 1980s, if it needed to generate graphics there were "graphics generator" units, also 19" rack mounted, which took HP-GL commands over RS232 and created the video signals. If the method being developed succeeded a more finished design would have been made.

[@alantaylor3910]

I worked as an electrician in a steel rolling mill (retired now) and this looks like something that was used to replace some analog logic. Proms were used to help match different production modules as needed. This was all done inhouse as you could not buy anything that would work with the local setup. Serial ports may have gone to seven segment displays

[@JaimeWarlock]

Around 1980, I helped build industrial control systems and this computer is very similar to the stuff we did. We even had the extra solder wick on the riser boards to handle amp loads that increased over time as we got more power hungry chips. I actually kind of miss those days when I could have told you exactly what every chip and/or trace was for.

[@pirobot668beta]

Prototyping station/University research project. Get an idea, work out the logic, make the board, install it for testing. If it works, you have a new functionality for your computer! I saw some amazing machines when I worked at University. Example: A multi-touch stylus pad in the form a glass slab; CRT behind it. Solid slab of glass, about 22 mm thick, slightly larger than a typical monitor of the day. Dozens of piezo-crystals lined the edges. Sequential 'pings' through the glass would be attenuated by finger contact, more pressure, more signal loss. By pinging from many different locations, up to 4 finger contacts could be read with confidence. All hand-made using common shop tools, 1973 according to the label. No CPU: all 7400 series logic, analog or digital outputs. It got sold 'by-the-pound' at surplus...

[@ncdave4]

All those extra wires reminds me of when I built my Heath/Zenith 148 (ca 1984, my first computer). Although, H/Z was nice enough to show a diagram of where the trace jumper wires should be and what (if any) component should go with it. The video card never did work, and, at that time, there were H/Z stores you could go to and exchange non-working boards for good. Great vid

[@timmooney7528]

The Heath/ Zenith techs' motto was, "Never let the customer fail." For a short time I worked on the Z248 motherboards. The US government had loads of them. They were a planar system with the large DIN keyboard connector on the motherboard. It was intended to be a cost saving machine, allowing ti to be upgraded by swapping out the CPU card and other old cards. By the time the 486 came on the market, a complete "system on a card" came out containing the cpu, memory, local bus video, I/O, floppy and hard drive controllers. Everything fit on a full length ISA card, allowing nearly every card to be removed. I don't recall if ZDS sold many of the 486 cards. Most of what was being sold during that desktop refresh project was complete desktops.

[@ncdave4]

@@timmooney7528 Tim.. yes, the 148/151/158 was a planar system with individual boards for all the peripherals. In Oct of 1989 I added an Intel Inboard 386/PC card to upgrade my 8086 and I had a story on the install published in the Nov 1989 issue of "REMark", the official Zenith/Heath Computer Magazine. Heady days back then.

[@timmooney7528]

@@ncdave4 I worked at ZDS from 1990-96. I remember we had a couple of bookshelves full of reading materials, and REMark was one of them. We'd test laptop modems by calling the Heath/ Zenith user's group (ZUG) BBS. ZUG would have listings of clearance items as well as an area for private sales and correspondence.

[@bobblum5973]

Back in the day we called those riser cards "extender cards" because they would extend the signals outside the card cage so you could actually probe signals on the card you were troubleshooting.

[@geoffreykeane4072]

Yep - in rack mount usually extended out the front of the case because you wouldn’t be able to get in the top!

[@bobblum5973]

@@geoffreykeane4072 Exactly! I hated troubleshooting boards that worked on the extender card outside the enclosure or card cage, but failed once they were seated in place. You had to get creative, tack solder a test lead or use mini- or micro-grabber clips, then run the test lead outside the cage.

[@DavidRavenMoon]

I wouldn’t say “home made” necessary. I worked in a prototype lab at ITT Defense Communications in the late 70s and. We made lots of stuff like this. A lot of it was wire wrap.

[@iancooper2338]

We used that method too for backplanes using Kynar or Tefzal wire wrap and for Baudrate linking on DEC PDP11 boards (instead of red handbag links) but also for chip to chip wiring in proto too with fine enamelled wire and It’s trade name here was called Roadrunner. The varnish insulation on the wire spooled on a pencil burned off where you soldered it to the pin so you could could quickly join pin to pin around to make a bus between chips. Great dev method back in the day 👍 now PCB are very affordable for development and better for todays higher speeds etc. but this was great to see again, bit of nostalgia for me 🙌

[@rog2224]

I had the feeling of defence or academic.

[@karlmatthias2698]

Tektronix is in Portland. Almost certainly from Tektronix lab.

[@bobblum5973]

I worked at McDonnell Douglas in the late 70s onward, and saw a lot of similar hardware and boards, too. We had prototyping boards for things like these, plus wire-wrap, and even something called wire-weld that was similar to wire-wrap but instead would make traces using special wire spot-welded to small flat-headed posts. A nightmare to work on, especially when the welds were lousy, you'd pop them loose and make things worse.

[@dennisp.2147]

@@rog2224 nothing Defense. I've built defense systems. That wouldn't have passed muster.

[@adriansdigitalbasement2]

Since I can only have one pinned comment, I'm consolidating there. Side note, I plan to dump all the ROM chips with hope they can give us a clue about the origin or this machine. . I got an email from a viewer: "Many of the parts of the units were made by Tektronix. The case with the "special" openers are Tektronix stlye. The powersupply comes from a Tektronix 31 calculator. The numbers on the PCB (EXXXX = Tek prototype boards) and arrows indicating pin1, the style of the pin-row headers, the style of the extender boards,.... Thats tektronix all over the place. Since you're from the protland area you should contact the vintageTEK museum. I think they can help you to identify the unit." . I reached out to the TEKmuseum and heard back from Bob: "The consensus was that some boards were indeed made by Tek. But 'Ennnn' numbers imply an engineering project, as opposed to a product for sale. The boards are the same form factor as used in the first Tek terminal, the T4002. As to whether the wired boards were home-made or Tek-made, we couldn't decide. But it looks like a home computer made with partly Tek-made parts. A lot of employees were doing the same thing in the late '70's." . Also another KZbin comment seems to be revealing: Sam Cross: Greetings the device you have is part of ATE system bulit by a company I worked for in the early 80's in salem oregon. It was bulit for one of their customers on contract.. please contact me for more info.

[@jeromethiel4323]

I actually bought 4 gigantic dual 8" floppy drives made by Wang from a surplus auction in the early 80's. Absolutely massive units, with linear power supplies inside. The two 8" floppy drives. And then a card cage that held huge circuit boards with 30-50 74LS logic chips. Nothing bigger. There were 8-12 of these cards. All of that just for a dual 8" floppy storage system. Would love to have those today, simply as a source of parts.

[@mikelunsford2587]

I also worked with 8" floppies. Good times, and still had to strap ground straps to my wrist AND cabinets were little air-conditioners.

[@Azlehria]

Four massive Wangs? Most guys would be satisfied just having one . . ..

[@jeromethiel4323]

@@Azlehria my pants fit me like a glove! ^-^

[@phirenz]

Adrian, are the two backplanes even connected? Based on the fact that both sides had their own CPU installed, (Mostek 3870 in the right backplane and Z280 in the left backplane), and each size had both RAM and Serial IO boards, I have to wonder if this chassis might actually have two almost completely independent computers. The only thing the left side is missing is a ROM. Maybe this is a development platform, where you access the right computer via serial and then use it to upload software to the left computer and debug it. But that would require partial interconnection between the buses.

[@pragmax]

I agree that this is a workstation for building and maintaining other systems. The EEPROM programmer gives it away. The level of re-work and one-off boards in there would be risky to support for users, but totally fine for an engineer that's inside this case half the time anyway. Given the vintage of the parts, I'm guessing that some of the serial ports (DB25) are intended for some kind of serial terminal, like a VT100.

[@johndododoe1411]

Side to side interconnect could be with prototype wires between the bent pins at the top of some cards. Just 3 loose wires to hook up serial, or use that loose DB25 serial cable.

[@scsirob]

Wow, you got yourself a treasure trove! Z280 is a very fast enhanced Z80 with expanded memory (bank switching), and extra instructions. I built a CP/M computer in the mid-80's using similar technology (Eurocard experimenter board and wires, 64-pin Elektor backplane), using a 10MHz HD64180 CPU, which is virtually the same as the Z280. Also built a floppy controller (FD1793 I believe), a Z80-based VDU/keyboard controller, and a SASI (not a typo) controller to run an Adaptec ACB-4070 SASI-to-RLL bridge board. Hundreds of individual wires, can't even fathom how I found the time for this back then. Wrote my own Z-80 monitor program and CP/M BIOS. I really wish it had survived the cleaning frenzies, but unfortunately it's gone.

[@lucasrem1870]

ne needs boot disk...mad...

[@quetzalcoatl-pl]

I just hate "cleaning frenzies" even more now than before :((((

[@nigeljames6017]

Myself and three other guys started a company out in Silicon Valley. I wrote the software, they did the prototyping boards using your “hodge” technique. It was highly effective and quite quick to implement. The only drawback was if you found out you had made a mistake ten wires back you would have to carefully pull the unsoldered wire through pack of good ones hoping nothing would snap.

[@thomasives7560]

Those boards are definitely Tektronix made, but it is hard to tell where they originated from. Tek made a lot of automated test equipment in the 60s-90s, so this may have been a prototype system, or maybe just something that a Tek engineer cobbled together. If you could find some additional part numbers, some of the Tek-savvy community might be able to track down the data, maybe even an instruction manual. Great video, love the channel. Cheers!

[@cambridgemart2075]

The transformer part number would almost certainly narrow it down.

[@canman4058]

I have seen very similar one-off / modified computers used for tracking radars. The radars were used on a military test range in the 1980s. The computers were used for an upgrade to the signal processing side and/or the control of the radar. The computers were mostly hand made, the circuit boards were made in house. The jumpers were for correcting errors in the design or using slightly different components. There were probably only 2 made, 1 for spare.

[@OscarSommerbo]

My first though with any strange/mysterious computer was support equipment for testing, there are a lot of custom and strange boxes that were used to interface with sensors/testing equipment. But it looks more like a true home-built computer. Proto boards for actual components, home built floppy controller and rtc/serial board, and to my mind that Sony SRAM chip. The computer was built and modified with considerable skill over at least 5-6 years. I would assume there never was a display controller, as serial access was very common in the late 70's when this was being started. Further, I assume it was a hobby project by a real electronics/computer engineer, those proto boards are just beautiful, and routing 5v along the perimeter of the board just screams of experience. Many cite cost or quality of materials as a factor against being a home build, I disagree, just because there were cheaper alternatives doesn't mean some wouldn't buy it. And the level of detail and knowledge going into these components hints at a professional building it, so maybe some stuff from the office just fell into his bag. Or he was allowed and or encouraged to take a few odds and ends home.

[@LatitudeSky]

We had a similar box where I used to work. It was issued to us by a government agency and the only ports were power and network. It has a CD ROM drive. All black paint, no markings whatsoever, and every screw and panel was epoxied shut. We were shipped the box, told to install it and connect it and leave it alone. Once in a while, we got CDs in the mail and it was marked in the order to install them. Put in a disc, wait until it popped out, insert next disc. No other instructions. Yes I tried to read the disc. It was all encrypted. That black box ran on our network for years and was actively sending and receiving network traffic. Eventually "they" sent a shipping case and it got sent somewhere.

[@Jerrydt44]

Looks like aerospace ATE I started in 1972 at King Radio that made Avionics for aircraft. They had 2 different test equipment departments. Automated and Custom ATE and CTE. I worked in CTE. Back then it was very expensive to lay out PCB's A draftsman to lay it out, artwork fees and several hundred for the boards. They would hand wire them. This looks ATE, they would have a large box of boards and cables going out to a bed of nails, That's what they used for laying a PCB production board on for testing.

[@Rorschach1024]

I'm betting this was some prototype/engineering built test interface equipment for Boeing. When every plane is being qualified, a lot of instrumentation is mounted in it and the equipment is short run/custom built. I designed and built a combo vga computer display/analog video display/vt100 terminal for such a setup for Boeing about 20 years ago.

[@ErraticPT]

A college near me had around 6 of these (or very similar) sat in the corner of a storage room in the early 90s. Each had similar cards to what you see and one had a HDD (5MB I believe) and what looked like a memory mapped jumper board (confirmed by a old nerd and tutor who worked there) with over 500 wires. He would occasionally fire that HDD one up with a grinding a screeching noise and some of the loudest beeps you've ever heard! Apparently they came from a "database farm" at a local ship builders/heavy engineering company.

[@EngineeringVignettes]

Homebuilt CP/M machine, maybe for industrial CNC or some other numerical control? A bit pricey and complicated for a hobby machine.... I say industrial as it looks beefy enough to be running 24/7 in some plant... The PSU was stolen out of some Tektronix equipment and definitely over-kill for the purpose. It's not S100 though. I have done similar wirewrap jobs in my past but this guy didn't use wirewrap so it would have taken a looong time to wore though boards up. Nice neat job though.. well done. A quick tip one doing a neat wirewrap job is to put in "croquette hoops" of bare wire on the wiring side prior to starting the wiring then use the hoops to guide and hold the wiring. The little plastic thingys that were missing were called "card extractors". Cheers,

[@leon545b]

I remember a prototyping system in the late 70's early 80's that utilized a special type of "enamel" coated wire that was actually a plastic coating that would vaporize when you applied soldering iron heat; it eliminated the need to strip the wire. I don't recall the name of the system but it looks like it was utilized extensively in this box.

[@karlkiessling]

I remember using that a lot! And wire wrapping thousands of Vero interconnects....

[@nophead]

Yes I used it in the 80's and also used it as recently as last year! It was called Verowire and can still be bought. The only difference is I use little plastic "combs" that have pegs that fit into the verboard holes and have tiny cleats to wrap the wires round, rather than being tied into to tiny looms as this was. I made my own computers and floppy controllers, EPROM programmers, etc, with it in the 80s.

[@Otakunopodcast]

Someone ought to show this to the KZbinr "Tech Time Traveller." This looks like one of those home-built computers from back in the early days of computing, like something you would put together based on plans/schematics from a magazine back in the 70s (or maybe even the 60s?) Tech Time Traveller specializes in those type of homebrew systems, and might know a thing or two about it.

[@matthiassteinbrecher473]

Back in the late 90's I was working in elevator maintenance and I remember we had some elevators from the mid- to late-80's that were controlled by similar computers. We had maintenance units that provided graphics, keyboard and HDDs to perform software updates or changes. The riser cards were used to make the cards acsessible for onsite-troubleshooting. Empty spaces in the housing were usually used to store spare cards. The power supply units often were modified to provide some extra power and/or to bridge power failures (sometimes by big capacitors, sometimes by rechargeable batteries). Later versions also provided control-LEDs in the front panel that showed the most common errors to save time for diagnosing. Since this looks very common, the units we used back then were more evaluated and normally had standard PCBs without all the wild wiring on th e back. I don't think that one of these self-engineered elevator controls have made its way from Germany all over to your place in Portland. But since this is no rocket science, I could imagine that someone over in the U.S. developed something similar to the units we used back then.

[@BilisNegra]

I have a feeling that browsing early issues of Byte magazine (started in '75), one would find something made of many of the parts we can see here, back in the era of partly homebrew/built in kit form computers, and yes, this is a reference to Shelby's video on his complete collection of the magazine.

[@arjovenzia]

I opted to submit my assignment for my TTL digital course as a physical built unit. you could have submitted just the design, but I like building stuff. i wasnt great at digital logic, but managed to design and build a discrete logic digital clock. the combo wire-wrap/point to point style Is a really satisfying way to build a dense, pro looking board. until you look at the back... but I very much enjoyed the process I only got an OK grade for the project, because it was a pretty basic design. but we did use it as a demo unit for other classes and it proved a very robust build. If i was 20 years younger, I could really see my self rolling my own Z80 CP/M machine like this. I built my first PC with the proceeds of my 12th birthday money. I made it known that I didnt want any plastic rubbish, but needed a bankroll for this project. I got a VGA screen and a 800mb HDD n some cash. enough for a Mobo n ram n awful GPU. case, CD, sound, better graphics, were added over the years. I 100% Grok the origional builder of this rig. such things are not bought, but lovingly crafted. my pinnicale was a hodgepodge 100% DIY watercooling rig that got my Athlon 2100+ to benchmark as a 3500+. 3200+ was the fastest off the shelf. Alot of ram tweaking went into that. Dual PSU (one for the mobo only, the other ran all the disks), Full tower AT case that was modded for ATX, window, neons (this is pre CCFL's, neon baby), paint job. 7x HDD's, all PATA. only weighted around 30KG. everything that could be modded was. I'd love to have a pint with whoever built this rig and just let their eyes glaze over as they talk about their baby.

[@8bitwiz_]

I tried to make my own 8080 computer from Radio Shack parts when I was a kid back in 1979. Of course it didn't work, and it never could have worked because I didn't understand bus fan-out. But more importantly, it would be decades before I would have the discipline to test things as I was building them. I went with software because it was much less annoying to recover when you screwed something up.

[@erickvond6825]

Beyond the fact that the riser cards were also likely used to make the ROM programmer accessable in addition to their normal diagnostic function. If I had to wager a guess I'd say it looks similar to an early TRS-80. Like maybe a development model or something. Beyond that I'm stumped. You might try asking @CuriousMark if he knows anything about this unit. It's certainly fascinating...

[@MarkHopewell]

At first glance it initially looked like the rubidium reference clock used by the BBC in their HF broadcasting stations. This "homemade" equipment was commonplace with them. Their various teams, working as if isolated islands of development, were scattered about their network, of which was full of this shabby looking heath Robinson stuff - aluminium absolutely everywhere - even the toilet seats were made of aluminium! When you threw this box in the bench, it sent a shudder down my spine. When I went to work there 20 years ago, this bespoke stuff was antiquated then. It was quite a shock to the system seeing such primitive systems serving the network like that.

[@rauldasilva3861]

It's a CP/M STD-Bus machine of some homebrew vintage. I ran an IMSAI 8080 S100... brings back memories. Most all CP/M systems were terminal based via RS232, You'll need to find an old ADM-3a, VT-52, Hazeltine 1500 or VT-100 terminal.

[@bobpoortinga7352]

No need for a physical terminal, just a PC with a serial port (or USB/serial adapter) and a terminal emulator program like Tinyterm or Putty.

[@rauldasilva3861]

@@bobpoortinga7352 Absolutely true. A PC with terminal emulation will do the trick, but if Adrian wants to set it up as an 'authentic' system, he'll need a vintage terminal.

[@thehearth8773]

You can indeed still buy those odd sockets you mention around 5:40--I use them all the time for making sockets for parts that aren't DIP or SIP (and you can get them rated to quite high temperatures too, so I use them for that as well--my job involves testing electronics at up to 175 °C). I get them mostly from Mill-Max, but another company called Preci-Dip makes them too, and I'm sure there are others; Samtec might make them too, I wouldn't be surprised.

[@cowasakiElectronics]

Z280 is a 16bit Z80. The Z80 itself had 16 bit registers and in a lot of ways was like a 16 bit processor although only an 8 bit accumulator.

[@DavisMakesGames]

No idea what this exact machine was but definitely something custom. The EEPROM programmer reminded me of a story my dad told me of when he repaired Apple machines. Came across an Apple II with some wacky custom cards and 2 hacked-in EEPROM readers in holes cut in the case at the front. His best guess is it was used for copying cartridge ROMS or something as the custom card allowed the contents of one EEPROM to be dumped to RAM, then written to the other. Made sense also since it had lots of RAM upgrade cards. Wish he still had it.

[@garthhowe297]

Fascinating! Some kind of development chassis? It looks too expensive, in money and time, to be any sort of homebrew. I wonder if dumping the ROM's might provide a clue?

[@Francois_L_7933]

It reminds me a lot of the things my dad used to work with in the aircraft industry. From what I've seen during family day visits, you either have a "server" that connects to some dumb terminals through serial connections or a buffer to fit between equipment and a plotter.

[@johanlaurasia]

In a conversation with Don French who was one of the developers of the TRS-80 Model I he mentioned that he liked the S100 bus, and based the Model I on the S100 bus, and even wanted it in the computer, but Tandy wouldn't have it because they wanted their own proprietary system instead.

[@holger8802]

All the boards (even the hand-wired ones) inside of the unit have E0000XA or ...B numbers. This means those boards were produced by Tektronix for internal purposes. So this is either a unit which was created to do some factory internal development or one of the engineers took the boards and other parts home to build his own computer.

[@zaprodk]

The funky IC sockets, component choice, font on the boards and the circuit board layout says Tektronix to me. That would correspond nicely with the transformer. Probably a one-off test system for internal use.

[@djmips]

Not to mention Tektronix was nearby Mr. Digital Basement who lives in Oregon.

[@F0NIX]

I worked in the Norwegian navy on a submarine that was a test bed for different equipment for future use. There was a lot of one-off computers or other electronics in rack. The front on this machine you had here looked a lot like the some of the ones we had. The rack mounting seemed to be standardized and most of them had a bus system with slot in card like this inside. But much higher density and made much more rugged and grounding everywhere. But there was a lot of CMOS circuits. And there was a lot of bodge wires all over, since they was a one-off system. I remember we at one time got a huge rack about 1,5m high full of equipments. A few month after installation we got a replacement that was a revision and it only was a 2U unit that replaced all the units in the rack. I also remember we had often go into the program and fix bugs by instructions from the makers. That was the last time I was programming in Fortran :)

[@lindoran]

I had 24 hours to think about this. Start with the cards that have a known pin out cpu. You should be able to map the bus on from there. Presumably the rest of the cards will follow the same standard. Doing so will also help you troubleshoot the backplane and the voltages on it. It is very possible if there are other voltages besides 5 you can look at the cards with ram to further help map the bus. It's as good a place to start because once you know the bus you can figure out part by part if things are not working. But first map the bus then test the power rails.

[@phirenz]

It's hard to tell from photos, But I think D0 though D8 are on backplane 2 though 9. A0 though A11 are on backplane 29 though 41, A12 though A15 are on backplane 43 though 46

[@michealcollins8383]

I have a case like that. Most of those cases are meant for things like Print Controllers(Large endustrial printers), or CNC machines. Judging from the cards you have and the controllers / Roms i am seeing. it is mid level CNC controller / Print Controller from mid 80's. In many cases like that, there is a main board that supports mutiple blades of montherboards with specific roms.

[@Vermilicious]

Seems industrial or scientific, and surely functioned as some kind of server. Realistically, I think you've just got a bunch of spare parts. I suppose you could try to power it up and see if you can communicate with it serially somehow.

[@thedopplereffect00]

Doubt it was a "server" in the modern sense.

[@richfiles]

Would love to see the ROMs dumped. Maybe you'll find a clue, like a text string with a company name, or a testing prompt. This definitely looks like a custom industrial controller of some kind. I used to hand build test panels for testing electric motors, at my old job. We were just doing analog circuits to connect things like motor windings to off the shelf text equipment, switching the windings read based on hall effect commutation signals, so we didn't need a whole computer inside, but it's the same principal. All hand made equipment meant for a specific application. You might have a few identical panels, or a singular, one off machine, depending on the application requirements. For the stuff I built, we had maybe a dozen test stations around the plant. I was responsible for occasionally making a new one, as well as calibrating and repairing the existing ones. The nature of what you have makes me think it was operated remotely, by way of a terminal. I think that's the right assumption. With so many serial interfaces, I very possibly could have been connected to not only a terminal, but one or more bits of industrial hardware, with serial control I/O. About a decade ago, I snagged a really intriguing bit of custom industrial control kit. It has dozens of BNC connectors on different modules. Some are marked as monitors with In, out, and monitor BNCs, Others have BNC marked with things like Clocks, burst clocks, indexes, etc... No idea what it all does. The highlight of the device is a module that uses TI smart LED displays and thumbwheels to read or load data to or from various addresses. No idea what it was used for, but it's neat that it can power up. It's obviously some sort of custom digital industrial hardware, but it seems to be all discrete analog and digital I/O controls. Very interesting.

[@slabo8171]

This might be the most beautiful thing I've seen on your channels so far. I love those PCBs with wavy traces, presumably hand-drawn. I'm having a hard time finding anything from the history about how exactly such circuits were made. I would love to see a video documentary with someone plotting those. And also the later techniques like using a tape and things like Bishop Graphics Inc. templates and how scaling down was done. There seems to be very little on those topics, maybe due to the rivalry in electronics between countries. It's a pity. Finding someone who worked on this machine, whether it was an early expensive Arduino-like home project of a hobbyist or commercial prototype, to tell the story of it, would be great.

[@gorak9000]

I'm pretty sure these boards with rounded curvy traces were laid out with tape - that's how you get those smooth yet consistent lines. This is almost definitely not a "arduino-like home project". This looks custom for a very niche application. Something that only 15 or 20 of them were ever built. I'm sure the original price tag would knock your socks off.

[@DeVerm123]

I made lots of those PCB’s. You have transfer sheets for all the pins and little rolls of tape for the traces. That tape rol was about 2” diameter and available in at least 3 widths. We used a sheet of 0.1” matrix, put an overhead projector sheet on top and did the work on that. Then we put positive mask on blank copper circuit board, exposed, developed and etched it.

[@slabo8171]

I've rewatched a video I've found some time ago while I was on my personal hunt for this topic. It's from Tech Time Traveler channel, titled "Vintage PCB Artwork From the 1970s". It features some of the originals made with tape and decals. Somehow I forgot that there were indeed curved traces as well. @gorak9000, I guess I have played too little with tape art, as it was something hard for me to imagine, that it would be easy to bend the tape like that. The adhesive tapes I've seen weren't too prone to twisting. Now I feel like I need to find some tape elastic enough to try it myself. ​@Nick Vermeulen, that's great! Now someone with good video production skills should kidnap you and make you show the original process on camera. I would love to watch it.

[@slabo8171]

The video I've mentioned: kzbin.info/www/bejne/j4rUhXaCgrOSgas

[@DeVerm123]

@@slabo8171 I did some searching but couldn’t find any of the supplies we used. The use of tape and transfer labels was frowned upon as we were supposed to use transparent paper, with Rote-ring ink pens and rulers with patterns for the solder islands etc. Those drawings we did 2:1 to achieve the precision needed. One of the reasons we liked the transfer labels and tape is that you had transfer labels with either one or two traces going between IC pins. You could just connect to those with the tape. This was all 1:1 on transparent film instead of transparent paper. The tape curved so easy because it was narrow and made much like painters tape: it could stretch at the edges. I may even have some left in a box at my parents place… I’ll look for it next time I’m there :-)

[@LegalizeAdulthood]

385x parts appear to be the components of the Fairchild F8 microprocessor line: en.wikipedia.org/wiki/Fairchild_F8 3850 CPU, 3851 1KB ROM, 3852 Dynamic Memory Interface, 3853 Static Memory Interface, 3854 DMA controller According to that article Mostek released an improved version of that CPU, so no surprise there are also Mostek parts on that board.

[@ThorstenDrews]

I just read an article in on a german computer news site maybe two weeks ago about the story of the CP3-F CPU developed in Germany that led to the development of the Fairchild F8. I'm in IT for a really long Time now but I never heard of them.

[@seanbarron2890]

That enamelled wiring is beautiful work! Whatever that machine is, it was a real labour of love. Maybe you should consider framing and wall-mounting those boards!

[@ronniepirtlejr2606]

Those were the days when you built a computer "You literally built the computer!" Some amazing work went into that masterpiece!

[@thiesenf]

Nowadays we program FPGA's the same way... :-)

[@larryh8072]

I could tell those boards were laid out by hand. I worked as a test engineer and built several test stands that interfaced with an Hp85 and several HP analyzes and volt meters. We created our own back plane which had a half dozen cards for signal processing and I/O. We had a pc lab so we were able to layout all of the boards. All painstakingly done either by cutting rubylith or using Brady tape and stencils. We did everything 2:1 and then made negatives for exposing the boards. It really was a dream job!

[@ChisakoYume]

I'm no expert, obviously, but running my imagination I came up with this wild guess: It could come from a small company or department of bigger one, that designed some devices / chips / software for such. I highly doubt this is a private machine. What threw me off the most at first was having programming logic inside a rackmounted chassis. wouldnt it be much more convenient to have it more accessible? as a private solution i would never make it so inconvenient to access. i had built a 90° angle raiser card and a cut out on the chassis to get there probably. but in my scenario it would make some sense. If a team of maybe 3-4 people would write code for a project on their client machines to programm something and "upload" it via serial/parallel the machine would get instructions like what type or chip is used and what addresses to write and right away burn the rom onto it and just sent a "completed" message back to the client machine when its done. at least to me that would explain why there is no video neccessary, if all it did was giving status messages and writing roms on to an IC That computer would be mounted in a rack maybe along some sort network storage for archival purposes or something. like saving revisions of what they did (and didnt work) leads me to a different version too. what if it was used in clean room reverse engineering? i think i heard of it once that creating own boards and stuff was necessary for that or am i wrong? finally, in a heavily customized machine like this, i'd be surprised if there is anything unpopulated for any reason. i think it is straight up missing and maybe that's why it is not making sense to people who deal with computers a lot. if they went through all the trouble building this, i just cant imagine an empty socket for no reason.

[@gorak9000]

The programming board is there so you can update the firmware (change the functionality in some way) and then burn new ROMs right on the same device. That ROM board is obviously generic and provides sockets for many more EPROMS than were required for the custom code that was running on this thing.

[@ChisakoYume]

​@@gorak9000 Since no one really knows for sure yet, you may be entirely right. I also might be an idiot and be entirely wrong. But I cant make sense of the purpose then if it works like you say. It leaves me with questions like: Why would an individual make it so complicated? As an individual, i'd try to get it as usable for me as possible. With all that modifcation done, I cant imagine that person made it so inconvenient to access for no reason at all. So, what if it wasnt a private project and how would it be beneficial for a company then? Someone still has to open it up, put a chip in the ZIF socket then run back to a terminal of sorts to give instructions then go back and fetch the chip from that awkwardly positioned programmer and install it in a socket on another board. Seems too complicated for just slamming updates on it. Or did you meant that it can update a chip on any other board inside of it right away? I highly doubt that, as the voltage is too unsafe to send it through multiple connectors, wire types and resistors. You cant possibly write something reliably like that, can you? Also the ZIF socket wouldnt be required anymore for that but it wasnt bypassed as far as i could see in the pictures. Only some wires have been added regarding the voltage from what I saw in the pictures. So why is that still there / still enabled? It must have been used, right? My guess on the ROM board is, that it could hold different "programs" or instructions on how to handle different kinds of eproms. The terminal would f.e. send a command to run code in address space 2000 to run the programm on that ROM or address space 3000 to run code from another ROM. If Adrian would try to read the ROM chips, that could give a clue on what they do and possibly solve the mystery. So I stick with my guess for now that it is a kind of glorified network-ish EPROM programmer that is used by multiple people. Either for developement or reverse engineering.

[@kaitlyn__L]

That point about Star Trek never mishearing the cast made me imagine something along the lines of: “computer, I need a 3 second tachyon burst”; “okay, 3 second tetryon burst, coming right up”; “no no that’ll be dangerou-“ and then the ship blows up. My personal pet peeve is when I ask for a two minute timer and I get a ten minute one. Or a quarter of an hour timer and I get a one hour timer. Different types of mishearing for each, but both happen about 10% of the time for me.

[@ethanspaziani1070]

Those homemade boards are beautiful that's basically something that's a lost art in today's computer world I hope the corrosion on the clock can be fixed and I hope that it works would love to see a part too

[@tookitogo]

Not a lost art, it’s still done (and taught) for rapid prototyping.

[@ARTofTY-TV]

I reeeeeeally hope he tries to get this working again!!

[@johnmorris3776]

I have a lot of respect for who ever wired these cards up.

[@lindoran]

24k is basically the bare minimum of ram needed for cpm with a relatively standard bios the programmer would have access to approximately 4k. This would depend on specific customization... Very plausible roms were used for program storage with very little ram, it's also very unlikely the "display" is just a serial terminal, as display routines would take much of the remaining 4k.

[@absalomdraconis]

As at least one other commenter said somewhere, this will be a "home" made machine, built for commercial or industrial purposes, and upgraded over the years with the old cards floating around inside the case so that they'd be easily available if the replacement boards went bad at some point. There's a decent chance that it doesn't match any spec, and probably was hooked to one terminal or another through it's entire service life. Could have even been used for something like internal inventory tracking at some repair shop.

[@gorak9000]

It definitely looks custom built for a particular application - testing, automation, something that runs some custom code (on the eproms), has the ability to change that code (the eprom writier), and talks to a bunch of serial devices of some kind - maybe terminals, maybe something completely different. This definitely is not "home" made - it's some small run custom controller for something. Probably one of those things where if you had to ask how much it cost, you couldn't afford it.

[@rickmartin6817]

If someone has never had-wired circuit boards like that, they have no idea how hard it is to do it right.

[@allangibson2408]

As part of my electronics course I had to build an 8086 computer using wire wrap and discrete components. Interesting… (particularly as the final assessment was getting it to talk to one of the other students projects).

[@rickvaneijck3016]

We used to do it at ibm

[@paulmichaud7565]

Yes, a sign of deep engineering wizardry. Perhaps to custom engineer support for factory sensors or manufacturing, but the box allows it to interface with an off-the-shelf mini. So, a development box for said mini-system. I bet a ROM dump would provide some interesting insight. Fun video, Adrian.

[@melkiorwiseman5234]

Stopping at 13:30 to mention that I'm pretty sure the ROM board was designed to take a certain size ROM chip but someone wanted to use a different size ROM chip and manually re-wired the board so that it could use the size they wanted, at least in one and probably two positions. A lot of the other "bodges" are probably for a similar purpose; to adapt the old machine to use newer chips. I feel certain that the switch is to enable the selection of alternate boot configurations/ROMs. Stopping again at 20:38 to say that it looks like the computer was designed to connect to a dumb terminal via a RS232-C port and so it never had a keyboard or video built-in. From the fact that there's one male port and two female ports, that indicates that it was intended to have just one dumb terminal as the interface with two ports free for printers, or perhaps a printer and a modem. At 24:05 I think that 64K of RAM might be DMA RAM for the floppy interface, so that data transfers can take place without the CPU having to be involved and it can be off doing something else while the data transfer is happening. The 64K RAM is almost certainly bank-switched so that a "slice" of it can be placed into a particular slot of the CPU's memory space. The EPROM programmer appears to have been designed to program the 2708 series of EPROMs since they required a 30V programming pulse. It seems probable that the ROM board was originally designed to take 2708s but was modified later to allow larger size chips.

[@mikeselectricstuff]

5:19 could be 7-segment displays perhaps?

[@arf20]

too many pins

[@physnoct]

The traces going to the headers are consistant with 74373 latches. My guess is a GPIO board.

[@TheExileFox]

14/16 segment maybe?

[@Obladgolated]

Looking at all those lovely old chips and hand wired boards makes me remember Friday evenings watching _The_ _Love_ _Boat_ and _Battlestar_ _Galactica._ Long time ago. Mostly sad memories, a few happy ones.

[@MarkEichin]

The NEC 765 floppy controller is pretty late relative to the rest of the system... definitely fits with the idea that this was adapted over the decade or more that it was used...

[@knockywigglesworth1909]

Companies used a lot of expensive custom built stuff back in the 70's. Pulled a huge rack mounted server & office system out of a large national insurance company a few years back. They'd actually been using it up somewhere in the 2000's. Custom software programs I believe! They just wanted it gone! Separate rack mounted power supplies were huge & heavy! Did you ever see the stack of manuals included with those early IBM office machines?? Ridiculous... Staff would call almost weekly

[@mikipank]

To me, this looks like a device (I can't call it a computer because it's more than one computer in one case) that was part of the McDonnell Douglas DC-10 control module, which served in this class of aircraft from mid-1978 until 1985 when it was replaced by a more modern, smaller and more powerful device.

[@jpdemer5]

Not without a hundred serial numbers, part numbers, and manufacturer names on every last piece of the thing.

[@richpayton7162]

@@jpdemer5 So true. They had to be fully traceable to enable the lawyers in plying their trade, if something went awry and there was a plane crash. While at GE I worked on an early system to control their nuclear reactors. It used cmos logic ICs, prior to this it was mainly relay logic doing the work. Back to your subject, there were several "isolation boards", each with over 500 diodes on it, all identical types, but each with its own serial number and probably a paper file folder full of testing data and whatnot. Being new to this equipment I asked why all the unique serial numbers. Answer: Legal Issues.

[@jpdemer5]

@@richpayton7162 My brother worked for Pratt & Whitney. He had similar tales to tell. Not so much for the lawyers, but so they could track down any weak links and make the appropriate fixes. If they identified what they thought was a "bad batch" of capacitors, they could track down and replace every last one of them, in every device that contained one, thanks to insanely detailed record-keeping.

[@alexkalish8288]

I was an Intel apps engineer back in that time. From the case I was thinking S-100 bus system but it looks like a STD bus MB used for instrumentation. The give away is the notched connector - my guess made 1982-87 roughly. We used a successor, the VME BUS in the space shuttle power control system as I recall but with a different CPU. This was a very common industrial control system in it's day.

[@scharkalvin]

DZUS fasteners, I've seen those screw locks on old military equipment.

[@ExtrvertEngineer]

Came here to say exactly the same thing … I designed military equipment back in the late 1980’s. The sockets on the Z80 cpu card were called Harwin Pins and were stupidly expensive!

[@JamesHalfHorse]

Makes me think of some of the early automation gear used in broadcast. If something failed or it didn't quite do what was needed then out came the soldering iron and there would be braided copper and caps bodged all over it to quiet down the RF blasting from a nearby transmitter a problem I still deal with and about the only soldering I do these days besides making cables. It's just board swapping if not entire unit swapping. It is difficult at times because the station owners are still in the mindset that the chief engineer could fix anything with a trip to the radio shack down the road. Such things just do not exist or are not possible any more. Newer stuff doesn't even come with board schematics or service manuals any more kinda back to the no user serviceable parts thing again. I doubt this was the exact purpose of the unit but given how many of us can go yeah I have seen something like that means it's overall style/design was used for a bunch of custom equipment. My closest guess that would fit with everyone else's is some form of early industrial control/lab/telemetry/automation system before there were a lot of standards and off the shelf components in that area. It of course would have evolved over the years as technology and needs changed until being retired completely. I hope you will find a way to dump the eproms and get some clue as to who was behind it but I am betting give or take dodgy caps that this thing might fire up enough to give you a serial console or two. Also Mr. Carlson might have an idea if not know exactly what it does by looking at it.

[@darkstatehk]

See if you can read out some of those roms on your retrochip tester and see if any of the dump is legible - maybe a copyright comment or sys reference or maker clue?

[@dr.decker3623]

It is a "server box" for an automotive assembly line. These are the modules that hook up to the different arm "robots" on the lines. May have been a diagnostic module, not a main one.

[@joeloliver7599]

I can't imagine how many days/weeks it took to painstakingly build that system!

[@paulstubbs7678]

I built a 6800 system, with 32k ram using 2k chips, also with 3 serial ports & a floppy controller, all with wire wrap - it took a while....

[@wishusknight3009]

Accomplished techs in the day could design and build a card like that in a few days.

[@timmooney7528]

It depends on if the designer/ engineer is working on such a low humber he makes them all himself, or whether manufacturing engineering develops procedures for production "assemblers" to make the boards, and stop and test at various test points along the way. In the latter case, there is possibly a build rate or quota they had to adhere to.

[@mikespangler98]

I'm in awe of the soldering skills involved.

[@dr_jaymz]

That's a lot of noxious fumes sucked up by the builder. Luckily back then, those fumes were good for you and the designer probably smoked in the office anyway.

[@keithrushforth4019]

I had something similar to this back in the late 70s early 80s. It was a development system based around an S100 bus back plain. Over the years it was modified and hacked about as it was used for different projects until it ended up in my "useful junk" pile, and got parted out. I wish I'd kept it now as it would be great to be able to play with it again.

[@horusfalcon]

Those "screws" on the top are Dzus fasteners. Start looking for 7805 regulators! This does not look like S-100, but does look S-100-ish (maybe SS-50? I'm not sure.)

[@joelavcoco]

Not SS-50. That was Motorola 6800 and 6809 bus, and was unusual (and hazardous) in that instead of having female slots in the backplane that expansion boards plugged into via card edge connectors, it had a backplane that was a forest of male pins sticking up in the air, with female plugs on the cards. A screwdriver or other tool probing carelessly in the computer case can easily contact some of those exposed backplane pins and pop a fuse. Ask me how I know.

[@horusfalcon]

@@joelavcoco Thanks for the lesson! It's been too long since I messed with SS-50 gear, apparently. (I've been an electronics technician since about 1974... a lot of beer over the dam since then.)

[@joelavcoco]

@@horusfalcon My exposure to SS-50 equipment (SWTPc 6809), (and beer) may be a bit more recent. My interest in the 6809 started with the Tandy CoCo series. The OS-9 operating system that runs on the CoCos was also an option for the SWTPc, so I took an interest when a friend inherited several chassis and boxes of boards a few years ago.

[@horusfalcon]

@@joelavcoco Wow... that sounds like a neat system. I've heard some about SWTPc, and I did have a CoCo 3 back in the day. Never opened it except to upgrade the RAM to 128k (bought it on clearance with 64k...)

[@john_in_phoenix]

Looks like sensor/instrumentation (data acquisition) machines we used to use for missle tests. (early 1980s) My memory after 40 years isn't so good, but STD bus was quite popular. That is what we used for our own EPROM programming machine. It isn't homemade, it is a prototype and or limited production machine. Likely to be totally terminal based and never had a video card. Quite common for CPM machines.

[@WarrenPostma]

Given the rack case, I'm pretty sure this was industrial in use. I could easily see this being used in a process control, environmental monitoring, any late 1970s/early 1980s small company build. RS-232 ports, a serial terminal and maybe one serial port talking to a device or monitoring sensors. That linear power supply looks like it could blow up if you plugged it in without testing it.

[@timothyp8947]

Would be interesting to know if the two sets of backplane were connected to each other (beyond power)… is it possible it is actually 2 machines housed in the same case? A Z80 machine and the more prototype looking Z280 machine? On retrobrewcomputers.org there are a few people with Z280 projects - a relatively rare CPU due to rather a lack of commercial success. Presumably if this was constructed at a time contemporary with the date codes (especially with 8” floppies too), it would’ve been a very expensive bit of kit.

[@qdaniele97]

Yes, that thing is hand-made but far from home-made. More like some industrial prototyping machine or scientific university project.

[@8bitwiz_]

These are clearly three eras of prototypes for making many more boards. They were all together because they were kept together at the original company.

[@johnbandhauer9224]

I have no problem seeing this as a serious hobbyist's home built personal machine. It almost certainly *evolved* over many years. For instance, it might have had many different cpu cards over time etc. The handbuilt cards are amazing and very consistent in build style. Nothing wrong with terminal interfaces. People wrote software and books and did serious science with machines like this. Graphics are not always necesary. Also, just because this stuff wasn't cheap doesn't mean it had to be a business or institutional machine. The lack of labelling suggests this was one person's machine. Some people pour their money into hot rods or boats or vacations or big families. Some people build their dream computers (and maybe some of that other stuff too). This is definately outside of Adrian's wheelhouse. I hope he finds someone to take it on if he's not gonna try to make it work again. It looks pretty complete. Not obvious that it won't boot CP/M off of those roms, btw.

[@ronniepirtlejr2606]

It's a time machine! If you can figure out how to get everything going & apply 1.21 gigawatts of electricity. You will have access to the secrets of all mankind! 👍😎👍

[@EngineeringVignettes]

Either that or a very large fire in your basement... In any case it will be very exciting.

[@GPFlljk]

I actually had one like that before. It ran CP/M off 8" floppies and had two destroyed 10 meg hard drives. It required a dumb terminal and once I stacked the hard drives, the computer, the floppy drives and the terminal, I could just pull up a chair in front and it was at the right height. It ran dBase II, WordStar, and some kind of version of VisiCalc. I miss that system.

[@jensschroder8214]

When this part was built, many pages of data sheets had to be bought on real paper for each IC.

[@EngineeringVignettes]

I still have a shelf loaded with my old databooks...