Tales from the S-100 dark side.

[jcdammeyer]

It's been too wet to cast aluminum. Waiting for needle bearings from China for the harmonic drive project. Haven't decided what to do with the power drawbar project. So no machining for a while. What to do......
Well I've been meaning to clean up some of the old S-100 computers that have been collecting dust for the last 30+ years so I thought I'd pull them out and inventory what I have. Turns out to be a lot.
Found an S-100 facebook group and have had some nice conversations and help.
At this point I've now been able to get a Hydra 80186 CP/M-86 up and running. Has 5.25" floppy and hard drive.
Next was the Ithica Intersystem Z80 system with two 8" floppy drives. What is amazing is all the floppies are readable so far.
Current reforming the power supply capacitors on the third system which has a M68000 with CP/M-68K. The capacitors in it are pretty poor. Way more than two hours per cap to reform them. Going to be a while yet before I can see if it boots.
Another system that was taken apart, cabinet scrapped for the metal used to have two 8" floppy drives that were also scrapped. It's a Tecmar 8086 with 256K memory and it ran MP/M-86 for a multi-user system. I think I trashed the hard drives on that one. So all I have left are the boards.
And finally my own personal Z80 system that had a dual Persci drive. The heads were moved by a voice coil system instead of stepper motors. I may plug all those boards into a motherboard (I have extras) and see if I can make it work. Just for fun.
What's interesting too is the tools for these systems. I have COBOL, FORTRAN, FORTH, PASCAL, DBASE-II, Plotting programs and of course the Adventure Game.

 

[Dabbler]

what, no FOCAL? Back in '74 FOCAL was all the rage - I even helped port a copy from a PDP-8 to an Intel 4004 system.

-- I have very little hardware left that is older than a few years. (I'm a little envious)

the only program I got to write (in FOCAL) was a version of Mastermind (the colour matching peg game)

Before I hit U of Waterloo, most of my programming was in FORTRAN or machine code...

 

[jcdammeyer]

what, no FOCAL? Back in '74 FOCAL was all the rage - I even helped port a copy from a PDP-8 to an Intel 4004 system.

-- I have very little hardware left that is older than a few years. (I'm a little envious)

the only program I got to write was a version of Mastermind (the colour matching peg game)

The big language, other than TRS-80 and Microsoft Basic was Pascal. Ithica had PascalZ and Turbo Pascal was $49 which included the book that I still have. Borland's Turbo eventually became Delphi with Object Oriented Programming. It's still around in RAD studio although they've drifted towards C and Python.
Delphi OOP Pascal Programs

 

[Dabbler]

I didn't see PASCAL until around 1977. By then I had used BCPL and C enough to stay away.

 

[jcdammeyer]

I didn't see PASCAL until around 1977. By then I had used BCPL and C enough to stay away.

Wirth's book on Pascal is dated 1975, Kernigan and Richie C is 1978. But it was the dominant language in Unix systems so I used it in a 500 series Comp Sci graphics course on a PDP-10. But the year before we'd used UBC PASCAL for a telecom course along with a number of other languages.
After Uni I discovered ICON in a book store in San Fransisco. Now that was a cool language and Python now has a lot of what was in it.

What I've found now is that the worst languages have the least amount of documentation with the programmers thinking the languages are self documenting. Really just an excuse for not doing the hard part which is explaining why rather than how. Python is the worst for that. But let's not go there.

Here's the Hydra 80186. All on one board with a big cable going to a DTC-520A hard and floppy controller. Runs CP/M-86. Nice little machine. Of course a Pi or Beagle runs circles around it.



Next we have the Z80 system. Thse are the two 8" floppy drives.


And inside the processor box are a lot of RAM cards. Why if the Z80 only has 64K memory address-ability? Well there's a paging system on the CPU board that lets it use at least 512K in 4K pages. Shades of IBM 360 architecture. Anyway each track on the floppy is buffered so once it's read into ram it stays there so reading files is slow the first time and after that pretty quick.


Finally we get the M68000 16 bit computer that can linearly address 16MB. A 256K ram card and a 64K are in there along with the 8 serial port card, and CPU, and disk controller card. and big heavy power supply and box for a 5.25" floppy and hard disk. Overkill. The front panel for the Integrand disk box has gone missing...

 

[jcdammeyer]

Aaaargghhh. I have three different power supplies that used to work with the older 8" drives. One, a Power-One CP272 is supposed to supply 24V, 8V, -5V and +5V. The 24V is fine but the 5V can't even provide 1A and it's rated fro 2.2A. So the 8" drive won't work with that.

Another supply I built years ago is 2.5A at 24V and 1A at 5V. No problem with 5V but as the regulator gets warm the 23.8V drops to 17V.

The other supply which should be able to run the hard drives is friggen huge and meant for a 20 slot motherboard in addition to two 8" floppies and a hard drive.

So I buckled and ordered this for delivery on Sunday. https://www.amazon.ca/gp/product/B07RT54H9V/

BTW, if anyone is cleaning out a closet and stumbles onto a 5.25" floppy drive or/and a 5.25" older style hard drive with the two cables I'm interested. If it's really cheap. Like shipping costs...

Thanks
John

 

[Canadium]

Aaaargghhh. I have three different power supplies that used to work with the older 8" drives. One, a Power-One CP272 is supposed to supply 24V, 8V, -5V and +5V. The 24V is fine but the 5V can't even provide 1A and it's rated fro 2.2A. So the 8" drive won't work with that.

Another supply I built years ago is 2.5A at 24V and 1A at 5V. No problem with 5V but as the regulator gets warm the 23.8V drops to 17V.

The other supply which should be able to run the hard drives is friggen huge and meant for a 20 slot motherboard in addition to two 8" floppies and a hard drive.

So I buckled and ordered this for delivery on Sunday. https://www.amazon.ca/gp/product/B07RT54H9V/

BTW, if anyone is cleaning out a closet and stumbles onto a 5.25" floppy drive or/and a 5.25" older style hard drive with the two cables I'm interested. If it's really cheap. Like shipping costs...

Thanks
John

I'm pretty sure I have a floppy drive in an old computer I have kicking around. If I can manage to remove it you are welcome to have it for shipping cost. I'm guessing your power supply problems are nothing more than a blown electrolytic capacitor. Not too hard to replace. When they go the tops usually bulge out so they are quite easily identifiable and replacements are usually pretty cheap. Just a bit fussy to replace. You need a good soldering iron for electronics.

 

[Dabbler]

I think I recycled both my 5 1/4 and 8 inch drives a few years ago... shame...

 

[jcdammeyer]

Well the MPC 8" half height drive has some issues which explains why the Ready signal never shows up. A scope on the optical sensor that detects the index hole shows no activity. A meter set to diode mode sees a 1V drop across the double sided sensor but open circuit on the single side sensor. So I'm guessing the diode is toast or over time vibration has mucked up the diode which of course is glued into the frame. Or the wire is broken inside.

Second: the home slotted sensor has a broken wire where it snapped off the top. I might be able to fix that by carving away some of the plastic to get access to more of the wire. Don't want to unscrew it if I don't have to because then I have to go through the alignment process.

Guess I should have done those tests before I splurged $32 on the power supply.

 

[jcdammeyer]

I think I recycled both my 5 1/4 and 8 inch drives a few years ago... shame...

Yeah, I did the same about 5 or 10 years ago. Not only that, some of the CPM system disks, the manuals and the OS9-68K system disks and manuals. Silly me.

 

[jcdammeyer]

I'm pretty sure I have a floppy drive in an old computer I have kicking around. If I can manage to remove it you are welcome to have it for shipping cost. I'm guessing your power supply problems are nothing more than a blown electrolytic capacitor. Not too hard to replace. When they go the tops usually bulge out so they are quite easily identifiable and replacements are usually pretty cheap. Just a bit fussy to replace. You need a good soldering iron for electronics.

Thanks. Hold off until I get further along into this.

 

[Dabbler]

AI did the same ... Not only that, some of the CPM system disks, the manuals and the OS9-68K

OS9 68K rocked. I only got to use it for a year or so, but I like it a lot. Pity.

 

[jcdammeyer]

OS9 68K rocked. I only got to use it for a year or so, but I like it a lot. Pity.

I used it for two years in The Netherlands at a company that made Trim & Form equipment for bending and trimming the leads on Integrated Circuits. One of the reasons I got that job was because I'd ported it to a system here in Canada just before we moved. This was before Linux. Hard to say if it would be any better.

 

[Susquatch]

BTW, if anyone is cleaning out a closet and stumbles onto a 5.25" floppy drive or/and a 5.25" older style hard drive with the two cables I'm interested. If it's really cheap. Like shipping costs...

I am soooooo envious.

I will see what I have. Most things got put out with the trash in one of my moves. I had a giant S100 rack system. I cried as I took it to the curb to wait for the Hearst.

I am almost certain I have a few drives of various sizes laying around that you can have. I mostly kept stuff like that for parts (screws, motors, steppers, magnets, etc) I'll look later today or tomorrow.

I have very much enjoyed reading all the posts on this thread. Waaaaayyyyy too much of my young life represented here. I'm too old to start crying over hardware and software.

But while everyone is sharing their history, my own specialty was at the hardware/software interface. After the early days of designing and hand building small dedicated systems, etching circuit boards, boot loaders, tape readers, building my own computers, writing my own OS, etc etc along came CPM........ O M G...... Soooooo beautiful! Sooooo Structured! Sooooo Adaptable! So many others out there doing the same things. Assemblers, compilers, etc etc all became possible!

I started to focus on designing and building boards for specialty purposes - mostly early controllers (not microcontrollers) that were controlled by the OS. That meant writing drivers with time critical functioning. I didn't follow the coder rules so things actually worked......LOL! I recall getting blasted by a larger company for software I wrote that mixed data and code - in particular I modified the code itself according to inputs received. They told me to rewrite it properly, I told them to FO, they wanted their money back, I wanted my code back, they tried to fix it themselves, then folded and let me do it my way. In hindsight I was more stubborn then than now!

I continued to like that side of things through the advent of the first microcontrollers. I still have a tray of 8748 controllers that I used to do all kinds of things. Some probably still have code in them. As time passed, I also learned to like the Parralax Basic Stamp - talk about a fast development system!

That's prolly why I can't wait to get cracking on the Arduino.

But first I need to finish a few major hardware projects......

Ya, life is a big circle. Wtf is going on with that fountain of youth thing they are working on? They better hurry up or ill miss out and never get to my Arduino!

 

[jcdammeyer]

My first computer was a National Instruments SC/MP with teletype at work for I/O. After that S-100 Vector Graphics Z80, Vector Graphics Serial, a kit 8K memory board, Vector graphics Floppy controller board and Persci dual 8" disk drive with voice coil positioning. I used the monitor that came with the Z80 board to key in hex bytes, into ram to bring in the bios from the CP/M distribution disks. Modified that to access my console and disk routiines and then inserted a new 8" floppy and performed the write track code for tracks 0 and 1.

After a few attempts which memory has lost as to how many is a few I booted CP/M. From there the second drive was used to pull in the rest of the .com programs off the Digital Research disk and it was put safely away. First step after that was to use ed.com to modify the bios in the same way I'd done it manually. Assembled it with asm.com and was able to create a proper bios hex file that became the system bios.

In looking through my 8" floppies I've found the wordstar document files for several of the courses I took in University. I was the only one in the non-computing science classes who was using a computer and printer to submit essays. So long ago.

The Arduino is pretty simple and has amazing support for all sorts of hardware. The downside is it doesn't easily let you get under the covers. So no break points. No looking at variables or registers. So if a program that runs great for 17.5 hours suddenly reboots you are 'screwed' essentially because one of the many library routines has decided to lose memory or run a loop until it runs out of memory.

But it's really easy to work with too. And amazing things can be built with it.

 

[Dabbler]

I came along just a little after: My first micro was helping wire wrap a 4004 board, but my first design job was a 6800 'fox' generator for teletypes. I was paid with an ASR33 teletype. I designed the I/O part of the board and wrote all the software. This was for a Waterloo company that was still maintaining ASR and KSR teletypes. (in 1975)

I didn't actually own my own computer until a few years later - a used Apple II with one floppy.

 

[jcdammeyer]

So the latest in this saga is why the MPI 41 series 8" disk drive didn't create an index pulse. There are two sensors and two IR drivers. One of them had a wiggly lead and it was the one for the standard 8" floppy in the normal position. And measuring at the connector showed one had a diode drop of 1.3V and the other open circuit. Had to take a fair amount apart in order to get at it. Then heat the body with the soldering iron so it was hot enough to melt the glue holding it in. And the insulation for the black wire just melted and the wire fell off. Clearly snapped at the body which is why it didn't work.

Now I can read the part number. It's an extinct TL 31B with 100mA max current and emissions at 980 nm. I found a replacement for about $9 at Digikey that is now on the way. Spectral output is at 945 nm. Hopefully that will be close enough. The ony 980 nm ones l could find are surface mount devices.

PART: 1125-1026-ND
MFG : MARKTECH OPTOELECTRONICS / MTE9440N
DESC: EMITTER IR 945NM 100MA TO-18

There are more expensive ones at 950nm but we'll see.

 

[Susquatch]

In looking through my 8" floppies I've found the wordstar document files for several of the courses I took in University.

Someplace somewhere I have the source code for an early version of Wordstar when it was last written in assembler before it was rewritten in C. As I recall they were looking for a more efficient way to manage formatting control codes and more efficient data caching on disk. As I recall, they never paid me for the full job because faster processors and bigger faster disks solved their problems for them.

That was back before bloatware was invented. It had to be compact and efficient then because it had to fit in the available memory.

 

[Susquatch]

So the latest in this saga is why the MPI 41 series 8" disk drive didn't create an index pulse. There are two sensors and two IR drivers. One of them had a wiggly lead and it was the one for the standard 8" floppy in the normal position. And measuring at the connector showed one had a diode drop of 1.3V and the other open circuit. Had to take a fair amount apart in order to get at it. Then heat the body with the soldering iron so it was hot enough to melt the glue holding it in. And the insulation for the black wire just melted and the wire fell off. Clearly snapped at the body which is why it didn't work.

Now I can read the part number. It's an extinct TL 31B with 100mA max current and emissions at 980 nm. I found a replacement for about $9 at Digikey that is now on the way. Spectral output is at 945 nm. Hopefully that will be close enough. The ony 980 nm ones l could find are surface mount devices.

PART: 1125-1026-ND
MFG : MARKTECH OPTOELECTRONICS / MTE9440N
DESC: EMITTER IR 945NM 100MA TO-18

There are more expensive ones at 950nm but we'll see.

View attachment 23362

If it doesn't work, I might have something. But I will have to dig for it and might even have to test it. I won't do that unless you need it.

Also, have you considered buying the pair instead?

 

[jcdammeyer]

If it doesn't work, I might have something. But I will have to dig for it and might even have to test it. I won't do that unless you need it.

Also, have you considered buying the pair instead?

Looking at the curve of the new device it seems like it should be fine. It's on the way from Digikey so I'll find out next week. I likely also have some IR diodes in my parts bin but they are fully plastic and I wanted something with roughly the same beam angle too. We'll see. I don't want to remove the receiver assembly. Even if I scribe lines in the casting, which I'd do before I remove it I'd rather not. And a broken wire on the sender is a pretty big clue to the missing index pulse.