Love on the buses: The S-100 and me

Although the IBM PC had been released in August of 1981, the system of choice for true geeks in those days was a home-brewed computer based on the S-100 bus architecture. I may have not been a bull geek in those days, but I was working on it - so I built one in 1982. It served as our family's computing workhorse for a few …

[Anonymous Coward]

OMFG

Get Mr. Peabody and the way back machine. This takes me back to some of my early dealings with PC's. Oh the hours I spent in dBASE, the even longer hours I spend hacking together bits of code, days debating the relative merits of keeping things on the standard 8" vs the new fangled 5" floppies. Those were the days :-).

[Anonymous Coward]

Those Were the Days

I had a Southwest Technical Products 6800 (http://en.wikipedia.org/wiki/File:SWTPC6800_Computer.jpg) PC back in 1978. It was a bit frustrating: The tape storage unit was a portable audio cassette player/recorder that was very flaky and if the room went above 78 degrees it would crash.

I did manage to program a pretty nifty space travel game in 6800 assembler.

I loved the IBM PC clone I bought to replace it in 1984 with PC-DOS, 1 whole megabyte of RAM and a 20 megabyte hard drive (it cost several thousand and was state of the art at the time).

[Martin Gregorie]

..and the SS-50

I built my first microcomputer about the same time, but using the SS-50 bus and the vastly superior (to assembler programmers) MC6809 8/16 bit MPU. Superior because of its nicely orthogonal address structure - if it seemed reasonable to use an address mode for the instruction you needed it was usually permitted: suck on that, Intel: no wall charts of instruction vs address mode was ever needed.

I still have the system. It would boot into Flex9 right now, if only I had a reliable 8-bit parallel keyboard or unlimited supplies of switch cleaner. Its successor, an MC 68020 running OS/9 as I write, is accessible from this Linux laptop and is run daily. OS/9 predates MSDOS and, as a multi-user, multi-tasking system, relieves itself all over Bill G's bought-in OS.

Its about time 'This Old Box' covered the SS-50 bus, Motorola chips and their operating systems.

[K. Adams]

KayPros and Micro-C

Brings back memories of my Kaypro II with a customized BIOS provided by the folks at Micro Cornucopia Magazine. Those were heady days, when one could actually call up the magazine and speak personally with the likes of Dave Thompson, Larry Fogg, Bruce Eckel, and Scott Ladd. I wonder what they're up to now-a-days? Bruce is still writing programming textbooks, but I've lost track of the others... P.S.: I miss you guys!

My jacket's the one with the beat-up copy of Micro C. in the inside pocket...

[Christopher Ahrens]

Anyone know where to get the parts?

I am working on my degree in electronic engineering, I was wondering if anyone could point my to a place where I can get all the schematics for this thing and possilby some good sources on where to get the components (although the lab at scholl has a huge assortment, i would guess that they don't have the parts...)

[Murray Pearson]

Reminds me of Dad

In 1981 or so, I went with my father to his office at a satellite campus of UC Berkeley. He showed me an 8" floppy and said, "This thing can hold almost a megabyte!" It was the very first time I heard the word "megabyte". I was about 10 years old, and SOME kinda impressed.

[Anonymous Coward]

and in the days of yore...

things happened that i dont give a toss about

is it a slow news day???

[Herby]

First liar doesn't get a chance...

My early entry into computers of the age (1976 or so) was with 6800 (Motorola) processors. I ended up designing a floppy disk interface (using a 5 1/4 "minifloppy") with then a wonderful single chip controller (WD 1771). I ended up writing an "operating system" that used the floppy disk as secondary storage, and it worked quite well, even adapting to any memory size it was given (8k bytes up to 60k bytes). I even adapted the Basic of the time (not Bill's) and a text editor. It was a going machine. I ended up making test fixtures for Qume Daisy wheel printers, and all was well with the world. The system was based on the prototype boards I had at the time (44 pin cards). Pretty simple for the day. A few years later I had boxes working as answering service computers based on this design. That lasted for a few years (1983) until the company went Chapter 7. Oh well, life in the big city, in a galaxy far far away, a long time ago.

It was FUN though!

[Geoff Campbell]

@AC

Run along, little boy, the adults are talking.

I must have a dig around in the attic, I'm sure there's an S-100 bus system up there somewhere, amongst many, many other bits of obsolete old junk.

Pirate flag, because that was what it was like in the '80s.

[jake]

@Murray Pearson

"In 1981 or so, I went with my father to his office at a satellite campus of UC Berkeley."

Civil engineer? Big, solid guy? Liked to get his hands dirty in the field? If your dad was who I think he was, he was a GreatMan[tm], and more than just in boot size!

[jake]

@AC 05:17; @Herby

AC 05:17 wrote: "things happened that i dont give a toss about"

It's called "nostalgia", youngster. Or better "history", as in "those who forget ..."

Herby: A lot of people reading ElReg were there in the 60s & 70s. What's with the assumption of anybody being full of it? My first "from home" computer access was Stanford's Tymeshare in (probably) 1974 or thereabouts. That was accessed via paper terminal & accoustic coupled modem ... Cost a lot of money, wasn't useful for much of anything but learning Fortran, COBOL and BASIC ... mostly I played StarTrek and Wumpus.

My first "home personal computer" was in late 1977. It was an LSI-11 based Heath H-11 (yes, I had to lay down the traces, boil the boards, hand populate & solder (and wirewrap, in places)) ... It was bloody useless for the most part, and bloody noisy, but at least I learned PDP-11/40 assembly language on the thing. (And the Heath-bus was a hell of a lot less trouble-prone than the s-100 ... which should bring about a few arguments ::grins::).

Eventually, I bought a 300 baud Hayes Smartmodem for the Heath/DEC gear. Usenet and the DECUS archives at Stanford beckoned. A little later, TheWELL & BIX arrived.

Eventually, an AT&T 3b1 (7300, AKA "UNIX PC") fell into my lap ... I'll stop there :-)

[Adam Williamson]

I'm with AC

It's harmless enough, I s'pose, but what exactly was the value of that article? I mean, it didn't really make much of an effort at historical context, or relate things to the modern age in some interesting and heretofore little-known way. That might've been interesting. No, it just seems to be "I once owned a computer and now I will bore you to death with the details" hour. This guy must be a hoot at parties.

[davcefai]

Memories of a Morrow.

My first "real" as opposed to "home" computer was a Morrow II bought in 1983. Also was the first computer at work. It cost £1999.40 with a Dot Matrix printer.

Within 2 months, finalising the departmental monthly accounts was taking 2 hours instead of the 2 weeks we were allowed. Suddenly everybody wanted one!

Incidentally the machine had 64K of RAM and 2 180K floppy drives. A 5MB Hard Drive would have cost another £1000.

[Camilla Smythe]

That will be Intel as in Not Intel

http://regmedia.co.uk/2009/04/29/s100_cpus.jpg

"Intel 8088 (left) and 8085 - notice that each chip has just 40 pins"

Oooooooo, gosh yes 20 each side. Ermmmmm I also notice that your Intel 8088 on the left is really an AMD 8088. The logo sort of gives the game away as do the letters AMD.

Who'd a thunk, the rivalry goes back such a long way......

[Jason Bloomberg]

SWTPC

I too cut my teeth on a SWTPC 6800 back in the late 70's. Had the luxury of floppy drives and a 40 x 16 terminal. The disks were on a metre long ribbon cable and would only work in the office in a particular orientation ! We found that out when we redecorated and had to reinstate the old two-tier shelving system and put it all back exactly where the feet marks indicated or 'no go'.

I also remember getting our stock control system in dBase II off CP/M and onto our first MS-DOS system. Ported flawlessly once we hacked some COM Port file transfer together. The whole thing fitted on a 720KB floppy with plenty of room to spare.

Happy days.

[Peter Gant]

Memory Lane

Back when these systems were current technology there was a store close to the Royal Exchange in the centre of Manchester. They had all sorts of S-100 systems and since the world wasn't so market driven in those days they didn't mind a kid playing around with them. There was also the Tandy Computer place around the corner and a shop that sold all sorts of Commodore kit on Deansgate.

Happy memories.

Ganty

[JohnG]

I still have a book on S100 somewhere

I miss the pioneering nature of those times and the BBSes where people swapped code without a lawyer in sight.

[Jeffrey Nonken]

Qume drives

Are you sure those aren't Tandons? I remember the Tandons having those brackets for mounting them in pairs (in place of a single full-height drive, and providing mounting holes for the same) and the turn-lever to open and close. The Qumes that I remember had a more traditional centered tab to close and push-button to release arrangement.

Now I have to go into storage and dig one out. (All I have left are the Qume drives, I believe.)

I also remember that the Tandons -- probably due to space restrictions -- didn't lift the heads off the medium, but rather stopped the drive. This caused a lot of problems because the systems didn't expect it, we sometimes wore our disks out in a few hours. The Qumes came later -- at least in our shop -- and were welcome relief. (I also recall the Tandons being slightly larger than half-height.)

Ah, here we go:

http://computer-refuge.org/bitsavers/pdf/tandon/TM848_service_1982.pdf

http://computer-refuge.org/bitsavers/pdf/qume/36027_QumeTrak842_Maint_Dec82.pdf

As for the rest -- thanks for the walk down memory lane. Me, I've got an old beat-up IMSAI-8080 that I need to restore someday. Might sell it afterwards. Also have some Godbout and other miscellaneous hardware that I could turn into a second system, though I'll be darned if I know what I'd do with it once I got it running. :)

[Jeffrey Nonken]

@Camilla Smythe

Ah yes, AMD. I recall that they made 8086 and 8088 clones that a) had the full op code set of a '186 and b) had a special 8080 mode. With the proper wrapper you could actually run CP/M-80 programs natively.

(Of course with modern computers you can emulate them at large multiples of the performance, but at the time it was pretty cool.)

I think I still have a couple of those. Seems to me I put one on my 8086-based Godbout board, in fact.

[Anonymous Coward]

Love on the buses: The S-100 and me

And I thought the Packard Bell Legend I still have was old!

[Anonymous Coward]

Everything old is new again...

So basically the floppy-emulator RAM card was my generation's version of the SSD. Except without that persistence without power thing...

This article made me nostalgic for Motorola's 6502 and then 680x0 family -- as a teenager earning spending money writing small business software on 6502 systems, interning writing assembler on 8086 embedded systems, and discovering the revolutionary 680x0 family at college, I was absolutely convinced that the elegant, orthogonal Motorola architecture would win out over the baroque Intel instruction set, or even its Zilog Z80 cousin... ah, the joys of youth...

As for the "what's the point?" whiners above... what exactly is your problem? Is there a shortage of space for articles? Was some other more interesting article pushed aside for this? Did you pay too much for an article you don't care about? Oh wait... this is free and online! So you're just making a fool of yourself complaining about it!

[William Boyle]

A trip thru memory lane

Boy, does this article bring back memories! I worked at ComputerLand in Los Altos in the early to mid-80's where we sold Cromemco's, CCS, and other S-100 bus systems, along with IBM PC's, Apple II, III (there was a dog!), and later the Lisa and finally the Macintosh.. I learned WordStar on an S-100 system belonging to my good friend Bruce Ravenel (one of the architects of the 8086/8087 processor family) in Mountain View. All that great stuff helped me segue into software engineering where I have developed a very satisfying career over the years. In those days, the Silly Valley was the place to be if you were into computing.

[jake]

@AC 01:07

"So basically the floppy-emulator RAM card was my generation's version of the SSD. Except without that persistence without power thing..."

Exactly. Later, as 2+megs of memory were standard on PCs, but DOS couldn't handle it[1], some people used the "extra" memory as a "RAMdrive", which is the same basic idea.

I still have a 386sx16 (+math-co) with 8 megs of RAM ... It runs DOS 5.0 & DesqView[2]. Back in the late '80s, one of my standard boot configurations[3] used one meg for the basic OS, and 7 megs as a RAMdrive to run the Mark Williams C compiler, mostly to develop small utilities for DOS and UN*X systems. In that configuration, it was an order of magnitude faster than the 64 meg 68030-based Sun 3/470 I used at work, at least when compiling the same files.

Before you pooh-pooh the above paragraph, consider that MWC was hand coded in assembler. As a result, the entire system (sans "demo" junk) fit on a single 1.44 floppy. That's the compiler, the assembler, the linker, the standard C libraries, an excellent debugger, some pretty good in-line help (not quite man pages, but close) AND a pretty good screen editor! I had a 7meg partition on a hard drive that was copied over to the RAMdrive when booting in that configuration, leaving me 5.5 megs for work space. Various environment variables were set accordingly.

Obviously, I had batch files to sync drives, and to halt or reboot the system when using the RAMdrive, so I didn't lose data. Also obviously, a crash would lose current work, so the batch file that called the compiler saved the "current state" to a second 7 meg partition on a hard drive before running the rest of the job. Not so obviously, I could use the hard-drive version if I wanted when using the 7megs of RAM for something else, by changing the environment variables set on boot, including saving the "working" 7 meg image to yet another 7 meg partition before starting to muck about with the "in progress" source ... Most folks never really figured out how to configure DOS to suit their needs ...

[1] Bill Gates never made the famous "640K should be enough" quote. The 640K "limit" under DOS was actually because IBM decided to put hardware RAM above 640K. In fact, most folks with a clue from the era had 720K or more of lower DOS memory available.

[2] I had a configuration that allowed me to run Win3.0 & 3.1 reliably under DesqView, but I never really saw the point.

[3] Am I the only one who customized config.sys & autoexec.bat to make booting into any number of different configurations possible? DOS was a down & dirty OS, but if you understood it's ins & outs, it was actually quite usable for little stuff[4].

[4] Note that I learned UNIX[tm][5] long before [PC|MS]-DOS existed. I'm not saying DOS was a good OS, just that it was usable IF you knew what you were doing. Kinda like most OSes.

[5] Yes. UNIX[tm]. Pre-BSD. I have a signed original Lions Commentary in the same UV resistant glass case as my nano second, my coffee & Guinness stained signed first edition K&R, and my wet-ink EWD. I'm not a computer guru groupie, just been around the block a few times, I guess :-)

[Anonymous Coward]

16-bit 8088?

"8-bit processor: 6MHz Intel 8085 ¶ 16-bit processor: 8MHz Intel 8088"

But the 8085 and 8088 are both 8-bit processors.... Maybe you meant 8086?

[Les Matthew]

£ Prices quoted

Haha, I seem to remember that anyting computer related back in the bad old days was priced at £2 to $1 minimum.

[IGnatius T Foobar]

Nostalgia!

Thanks for the trip down memory lane. My early days were also spent on an S-100 based machine. The photos and writeup were very nostalgic.

[Albert Waltien]

Ahh, the memories...

of tying down the data bus.Radio Shack made an "Expansion Interface" for the old TRS-80 Model I but they neglected to do that - caused no end of grief for their users. Micromint made an expansion box based on a Steve Ciarcia design featured in Byte magazine.Among other improvements over the RS version was terminating resistors on the data lines. As an aside to the (relative) newbies, this box allowed you to up the user RAM to 48KB and attach floppy drives - made it into a real computer <G>.

[Chris Gray]

Very similar to mine!

Wow! Rik, your system was pretty similar to the one I had. Similar Godbout cards. I wish I still had my CPU board to see if my 8088 was an AMD as well! My box was the full 20 slots - I had a separate box for dual floppies (only single density though). Did the "Ram disk" work under CP/M-80 or CP/M-86? I mostly used CP/M-80, and I programmed my own RAM disk for it - I used the 24 bit addressing mode of the Disk Controller to write instructions to the high part of each 64K bank of my 256K DRAM card, and I think it was 5 bytes per cycle that I could transfer from the "main" 64K bank to a "cache" bank, inside the CPU registers. Wonderful fun!

Before that system, I built my own diskless one using the original Intel 8080 chipset. I also went through a Radio Shack Colour Computer and an Exidy Sorcerer, before I got my first Amiga - the A1000 - that one I still have.

[Mage]

Not really 16 bit8088 8086

Most S100 used 8080 originally and then later Z80 or 8085. Few used 8088/8086 as it wasn't really 16bit. 68000 was really the first S100 16Bit. I remember Cromemco UNIX like Cromix 5 years later on dual Z80/68000 "minicomputer".

the 8088 uses an 8085 8 bit data / 16 bit address multiplexed and 4 extra address bits on the 8088 for 1024K MAXIMUM. But the various peripherals and screen RAM was mapped between the 640K and 1024K. The 8086 had 16 bit data on external bus otherwise identical to 8088.

Additionally the 8086 and 8088 are not true 16bit CPUs but basically the 8085 with a few 16 bit instructions and only 64K memory blocks like the 8080/Z80/8085. Later a Z80 clone had segment page memory manager for 1024k and the 8bit Z80 in Amstrad PCW could access 512K.

So the 640K was really a limitation of the small segmented address space of the pseudo 16bit 8088 and 8086. The amount of contiguous RAM depended on if text, hercules, CGA, EGA or VGA to and extent. It was much later that TSRs (AKA dirvers) could be loaded into unused holes in the 1024k as early DOS didn't support it and early IBMs had 128K, 125k, 384k, 512K or 640K physical ram. Later 1M physical DRAM was used.

The first real 16bit IBM PC was the 286, but DOS used it as an 8086. Only UNIX flavours of OS used it as a flat 16M address space.

The PC and DOS held back PC computing nearly 10 years as it was really not much different to the CP/M and 8080/8085/Z80 8bit machines, just more standardised.

[Anonymous Coward]

Windows 7 capable?

If Microsoft can trim down Windows 7 to run on this little beast, I'll buy that for a dollar.

[Gordon Crawford]

that's the name

wumpus , could not remember that name

But I do remember trying to decide on the s100 or a heath kit .

And the winner from Japan , boot leg Apple 2 mother BD...

thanks for memory lane down the Byte way

[jake]

@Gordon Crawford

"wumpus , could not remember that name"

If you're running a Linux or BSD variant, try typing "wump" at a command prompt :-)