• Please review our updated Terms and Rules here

A Z80 based SBC for beginners for the S100 bus.

monahan_z

Experienced Member
Joined
Oct 19, 2008
Messages
263
Location
San Ramon, CA
Guys, I thinking of trying to solve the chicken/egg problem by doing a single Z80 based S100 board that would contain 64K RAM, EEPROM, PS/2 keyboard input as well as a host of indicator LED's and jumpers to deactivate ROM/RAM keyboard etc. one at a time as other boards are added to get new people started on the S100 bus.

The only part I'm not sure of is what to do about console output/video. Was thinking of using the basic Propeller Console IO circuit
http://s100computers.com/My System Pages/Console IO Board/Console IO Board.htm

but this requires programming its two wire serial EEPROM with Parallaxes software and USB adaptor/software. Do you think that is outside the range capability of most newcomers. Alternatively we could go with a straight serial output to some dumb terminal IBM PC terminal emulator. Open also to other suggestions.

John
 
Hi All;

John, For my two cents worth, I would go with the Serial Port, as more people would find that easier to get going..
I for myself would not know how, nor have the means to program the Parallex..
And the same would probably be for the USB adaptor/software..

THANK YOU Marty
 
I would use serial as well. Much easier for someone to implement. I have 2 terminals that would love to be hooked to something.
 
John,
I support your idea of using the Propeller. I have built two "Replica" Apple 1 clones:

http://www.brielcomputers.com/wordpress/?p=1152

Similar to what you want to do, the Replica 1 uses the Propeller to replace about 30 chips in the original Apple 1 design. Using the Propeller is very flexible and easy to program, the tools are free, and anyone willing to take on something like this could also handle using the USB programming gizmo (power plug). Go for it.
Rick
 
I used to use FTDI chips, but found that I can get USB-to-serial TTL adapters cheaper than the FTDI chips: For example. That leaves you with a TTL header to hook to either USB, or to a conventional level translator for RS232C.
 
John,

My $.02 worth... I think a standalone S-100 board like you describe is a great idea. For two 'target audiences':

1) people with an S-100 chassis can get their old box up and running with just a single card, before they start to flesh out all the issues with their vintage card collection. That allows a step-by-step approach to getting original cards to work, and will fill in the gaps in their old card collection.

2) this is a nice, low-threshold entry into the S-100 world for those new to it. Similar to the original N8VEM with its ECB connector, people can build a stand-alone single-board machine that - if they like what they see - can afterwards grow into an S-100 box over the course of time.

My suggestion would be: if (2) is indeed part of the target audience, then the board needs built-in CF or SD mass storage. More than Propeller VGA and PS/2 connector. IMHO, video/keyboar are not that necessary: S-100 goes together with a serial terminal. If you don't have a terminal, $5 buys you a USB-TTL serial converter cable... but mass storage is essential, and also the first thing that dies in S-100 machines.

Either way, I'd love a board!

Kind regards,

Oscar
 
The starter "S(bc)-100" sounds like a marvelous idea!

My vote was leaning towards a simple serial port, but it has been pointed out, not too many computers these days have those without buying an add-in card. And, they are notorious for causing builder confusion as seen with the N8VEM SBCs.

Chuck's example is amazingly cheap and versatile. If the Propeller I/O is just emulating a simple terminal, why not set the board up for that, but have a header where the USB-to-serial-TTL adapter can plug in?

I really like the idea of being able to switch out sections as boards are added to the system.

Jim

PS - John, your gorgeous custom case makes me drool every time I see it. Very nice.
 
The only part I'm not sure of is what to do about console output/video. Was thinking of using the basic Propeller Console IO circuit
[[url..]] but this requires programming its two wire serial EEPROM with Parallaxes software and USB adaptor/software. Do you think that is outside the range capability of most newcomers. Alternatively we could go with a straight serial output to some dumb terminal IBM PC terminal emulator. Open also to other suggestions.
FWIW, getting a Propeller-based board connected and programmed is possibly some of the easiest setup I've ever done with a board. It's so low-threshold that for a long time I used to walk around with a Parallax Quickstart board in a pocket, a USB cable stub, and a small notebook PC in another pocket. Because it was so simple to program that board anywhere I could sit down, in a cafeteria for example.
 
FWIW, getting a Propeller-based board connected and programmed is possibly some of the easiest setup I've ever done with a board. It's so low-threshold that for a long time I used to walk around with a Parallax Quickstart board in a pocket, a USB cable stub, and a small notebook PC in another pocket. Because it was so simple to program that board anywhere I could sit down, in a cafeteria for example.

This is the kind of thing I'd personally define as super cool.

Out of random curiosity, what notebook PC were you able to fit into a pocket?
 
FWIW, getting a Propeller-based board connected and programmed is possibly some of the easiest setup I've ever done with a board. It's so low-threshold that for a long time I used to walk around with a Parallax Quickstart board in a pocket, a USB cable stub, and a small notebook PC in another pocket. Because it was so simple to program that board anywhere I could sit down, in a cafeteria for example.
I very much agree. Programming the propeller is very simple and easy for anyone (who would also be interested in learning about S100) to do.
Rick
 
Thanks everybody for the suggestions.
Here is what I’m currently going with and why…

First and foremost I wanted a Z80 circuit that is proven and works with almost every vintage s100 board out there. I used the core circuit on our Z80 board (http://s100computers.com/My System Pages/Z80 Board/Z80 CPU Board.htm ). It’s based on the old Intersystems II CPU board. I have not come across a vintage S100 board it does not work with. It goes to 10MHz in my own system of mainly S100Computers boards. Even works with DRAM boards. I had to remove the >64K windowing circuit for board space (see below). The Z80 to S100 bus signal conversions take up quite a few 74LSxx chips. Granted that could be reduced with a GAL or two but newcomers will probably not have experience programming GAL’s initially.

Second, I brought in the circuit from the Propeller Console IO board (http://s100computers.com/My System Pages/Console IO Board/Console IO Board.htm ). Because of space I removed the HEX display and a few jumpers. This provides direct support for PS2 keyboard and VGA video. Besides the internal ribbon cable connectors to the back of a box, I also added IBM-PC P2 and VGA sockets at the top of the board for direct connections. Again my thinking was KISS. It’s a well proven board totally debugged and is completely programmable (currently has VT100 capability as well). I wanted to retain the flexible status port and bit configuration jumpers section so when sections are disabled to use an external board (see below) the monitor need not be. Granted there is a simpler solution such as the PIC32. I may have to go to that for space reasons but with the Propeller circuit 99% of the work is done it’s just a splice. Also later, the user can get and use the Propeller Console IO easier when they build a larger system. The interrupt circuit is still there BTW.

There was an outcry for a serial port. These unfortunately take up board real estate badly. I opted to add the (again well proven) USB port chip that we used on our Serial I/O board (http://s100computers.com/My System Pages/Serial IO Board/Serial IO Board.htm ). This 18 pin “chip” takes up little circuit room and actually is easier to interface to a PC than a serial port. USB the serial cables are also very common.

Added a few LED’s that’s about it.

The board has 128K of RAM (there are no common 64Kx8 static RAM DIP chips) that can be jumpered into two separate 64K sections. It has a 27C64 (or equivalent) ROM, keyboard buzzer, 8 interrupt jumpers, a Power on clear and a ROM power on jump to any 1K boundary circuit. I used two, a 5V and 3.3V Pololu regulators. There are a lot of chips on this board. A TO-3 regulator capable of > 1.5 Amps takes up too much space. Those Pololu units are really nice and take little real estate.

There is a socket to program the Propeller EEROM but the user will have to buy the “Propeller Programming Plug”. A serial interface circuit is too board space expensive. They would also need to buy a PROM/EEPROM programmer (e.g. Wellon VP-290). After that they should have a board that would light up an empty S100 bus card cage. Clearly the next thing to do would be to build a FDC board or IDE/CF Card board and then go from there…

The circuits are configured so the onboard RAM, ROM, Console I/O and serial sections can be inactivated in stages as discrete more functional boards are added. While the board will act as a master/slave S100 board it’s not really intended to function is a complex multiprocessor configuration. For example for debugging you cannot “see” beyond 64K of RAM in the bus.

A few of you wanted more options including a CF card etc. Unless we went with a 4 layer board (more expensive, hard to debug), there is unfortunately not enough room. GAL’s (or a CPLD) would help but I wanted to simplify the board for first time users. The current 74LSxx board is amenable to simple logic probe debugging in all the circuits.

Now the bad news. This board is very dense. The initial via count is ~1200 vias! (More in fact than my current 80486 CPU prototype boards). It’s taxing Freerouter to find a solution. I playing around with chip layouts. May need some volunteers to shuffle things around an try for a solution. The multiple jumper options really soak up vias.

Anyway stay tuned.

John
 
Out of random curiosity, what notebook PC were you able to fit into a pocket?
Well, that statement of mine should have been qualified.. it's no ordinary pocket. I usually walk around with a lightweight 'Scott e-vest' (there's a website for that, remove space and -), which has an 'iPad' pocket. And tons of other pockets, but that iPad pocket works great for notebooks up to 11.6", although those aren't not all exactly the same size. I specifically got one which was no more than 29.5cm long, 19cm wide, so that it would fit the pocket.

-Tor
 
Thanks everybody for the suggestions.
Here is what I’m currently going with and why…

First and foremost I wanted a Z80 circuit that is proven and works with almost every vintage s100 board out there. I used the core circuit on our Z80 board (http://s100computers.com/My System Pages/Z80 Board/Z80 CPU Board.htm ). It’s based on the old Intersystems II CPU board. I have not come across a vintage S100 board it does not work with. It goes to 10MHz in my own system of mainly S100Computers boards. Even works with DRAM boards. I had to remove the >64K windowing circuit for board space (see below). The Z80 to S100 bus signal conversions take up quite a few 74LSxx chips. Granted that could be reduced with a GAL or two but newcomers will probably not have experience programming GAL’s initially.

Second, I brought in the circuit from the Propeller Console IO board (http://s100computers.com/My System Pages/Console IO Board/Console IO Board.htm ). Because of space I removed the HEX display and a few jumpers. This provides direct support for PS2 keyboard and VGA video. Besides the internal ribbon cable connectors to the back of a box, I also added IBM-PC P2 and VGA sockets at the top of the board for direct connections. Again my thinking was KISS. It’s a well proven board totally debugged and is completely programmable (currently has VT100 capability as well). I wanted to retain the flexible status port and bit configuration jumpers section so when sections are disabled to use an external board (see below) the monitor need not be. Granted there is a simpler solution such as the PIC32. I may have to go to that for space reasons but with the Propeller circuit 99% of the work is done it’s just a splice. Also later, the user can get and use the Propeller Console IO easier when they build a larger system. The interrupt circuit is still there BTW.

There was an outcry for a serial port. These unfortunately take up board real estate badly. I opted to add the (again well proven) USB port chip that we used on our Serial I/O board (http://s100computers.com/My System Pages/Serial IO Board/Serial IO Board.htm ). This 18 pin “chip” takes up little circuit room and actually is easier to interface to a PC than a serial port. USB the serial cables are also very common.

Added a few LED’s that’s about it.

The board has 128K of RAM (there are no common 64Kx8 static RAM DIP chips) that can be jumpered into two separate 64K sections. It has a 27C64 (or equivalent) ROM, keyboard buzzer, 8 interrupt jumpers, a Power on clear and a ROM power on jump to any 1K boundary circuit. I used two, a 5V and 3.3V Pololu regulators. There are a lot of chips on this board. A TO-3 regulator capable of > 1.5 Amps takes up too much space. Those Pololu units are really nice and take little real estate.

There is a socket to program the Propeller EEROM but the user will have to buy the “Propeller Programming Plug”. A serial interface circuit is too board space expensive. They would also need to buy a PROM/EEPROM programmer (e.g. Wellon VP-290). After that they should have a board that would light up an empty S100 bus card cage. Clearly the next thing to do would be to build a FDC board or IDE/CF Card board and then go from there…

The circuits are configured so the onboard RAM, ROM, Console I/O and serial sections can be inactivated in stages as discrete more functional boards are added. While the board will act as a master/slave S100 board it’s not really intended to function is a complex multiprocessor configuration. For example for debugging you cannot “see” beyond 64K of RAM in the bus.

A few of you wanted more options including a CF card etc. Unless we went with a 4 layer board (more expensive, hard to debug), there is unfortunately not enough room. GAL’s (or a CPLD) would help but I wanted to simplify the board for first time users. The current 74LSxx board is amenable to simple logic probe debugging in all the circuits.

Now the bad news. This board is very dense. The initial via count is ~1200 vias! (More in fact than my current 80486 CPU prototype boards). It’s taxing Freerouter to find a solution. I playing around with chip layouts. May need some volunteers to shuffle things around an try for a solution. The multiple jumper options really soak up vias.

Anyway stay tuned.

John

John - sounds like you've got a lot going on on your proposed board, most of the Z80 CPU board, most of the Propeller board, and the 8255 PPI / USB adapter section from the Serial IO board.

Maybe simplify things a little?

I Googled a little and found this neat PIC ASCII terminal - http://geoffg.net/terminal.html - it attaches to a serial port and provides a monochrome VGA (or composite video) VT-100 terminal, PS/2 keyboard *and* a serial->USB conversion all with one PIC chip. This would allow us to use the Z85C30 circuit from the serial board and just hang this one small circuit off of it vs including the entire Propeller Console board circuit. Using one of these with the Z85C30 from the Serial board might free up some space on the board and be a good compromise? The builder could choose (or not) to stuff this section of the board depending on if they wanted to use an MAX232 to talk to a PC serial port, or adapt to VGA/PS/2, or USB.

Also since the Z85C30 is a two-channel adapter, maybe the 2nd channel could feed an SD card if there is enough space free after taking out the Propeller circuit?

I know it would be a bit of a stretch to use a circuit we haven't used before in the board, but I think it might yield a more useable board with less vias to have most of the Z80 board, the Z85C30 section of the Serial board, an MAX232, this small circuit, and an SD card slot?

One thing I would point out - lots of EEPROM programmers can program the serial EEPROMs that the Propeller uses. You don't need the special programmer from Parallax, you just have to be able to write the image to the serial EEPROM. I use the same Chinese TOP-853 programmer that I use to program the parallel EEPROMs with monitor ROMs, etc to program the serial ROMS for Propeller consoles. (I do not recommend buying this programmer now - it only supports Windows XP 32-bit - I keep an old XP laptop around just to be able to use it).

Just an idea...maybe I'll take the Gerber files and make some of these ASCII Terminal boards to test.
 
I think the 85C30 would be a good choice, but why the Propeller chip? To me, that's a whole new programming exercise--probably more involved the the Z80 itself. Most S100 systems had little more than serial I/O to an external terminal. A simple USB-to-serial TTL satisfies that. Little SD card adapters are available (mostly intended for the RPi) for less than a SD card holder would cost--and they include the 3.3V regulator and 5V-to-3.3V interface. Last week, I bought 5 of them for $1 each shipped. If you just bring out the serial terminal I/O as TTL, you can attach either a EIA level shifter for traditional RS232C or a USB serial-to-USB for those who'd like to use a modern PC for terminal I/O.

I don't know how many folks actually want to program this thing--but IMOHO, the 85C30 is a bit more complicated to program than a legacy Z80 SIO chip would be, but you'd have to include a Z80 CTC for a BRG also.

My Z80 days were years ago, so I'm not particularly interested in building one of these myself (I probably have a few Z80/Z280/Rabbit) boards kicking around. If you go SMT with one of the eZ80 chips, you could easily fit the whole thing in a shirt pocket with room left over for a pen protector.

FWIW.
 
I think the 85C30 would be a good choice, but why the Propeller chip? To me, that's a whole new programming exercise--probably more involved the the Z80 itself. Most S100 systems had little more than serial I/O to an external terminal. A simple USB-to-serial TTL satisfies that. Little SD card adapters are available (mostly intended for the RPi) for less than a SD card holder would cost--and they include the 3.3V regulator and 5V-to-3.3V interface. Last week, I bought 5 of them for $1 each shipped. If you just bring out the serial terminal I/O as TTL, you can attach either a EIA level shifter for traditional RS232C or a USB serial-to-USB for those who'd like to use a modern PC for terminal I/O.

I don't know how many folks actually want to program this thing--but IMOHO, the 85C30 is a bit more complicated to program than a legacy Z80 SIO chip would be, but you'd have to include a Z80 CTC for a BRG also.

My Z80 days were years ago, so I'm not particularly interested in building one of these myself (I probably have a few Z80/Z280/Rabbit) boards kicking around. If you go SMT with one of the eZ80 chips, you could easily fit the whole thing in a shirt pocket with room left over for a pen protector.

FWIW.

On the N8VEM group we've already got 3 boards that use the Propeller for VGA/Keyboard (The S-100 Propeller Console board, the ECB PropIO V2 board, and the ParPortProp add-on board for the Zeta). All a builder has to do is load a pre-built ROM file into the serial EEPROM, and it's ready to run.

The Propeller can generate color video up to 1600x1200 using its onboard video DACs without the need for additional external hardware beyond some resistors and a VGA port.

This was selected as an alternative to trying to build boards around vintage character generator/video controller chips like the Motorola 6845, upd7220, etc, because it uses components that are readily available and it can generate timings that work withe modern monitors. Also the extra pins on the Propeller can be used as a keyboard controller as well.

I have a small ECB-bus system with the 80C188 SBC and a 4MB RAM card. I decided to build the ECB Color VDU2 card which uses a vintage MOS8563 (Commodore 128 high-res text display chip, which I had to order from the UK) and a VIA 82C42 (PC PS/2 keyboard/mouse chip). But the 8563 can only generate timings somewhere between EGA and CGA, which don't work with most modern monitors, and also fall in a vertical/horizontal sync range (49Hz and 18Khz) where they don't work with readily available video scaler boards either. I've spent days trying to come up with a solution to make that thing work; if I had gone with the PropIO V2.1 board with a Propeller chip I'd have had a working console in no time.

Some people want real VGA/PS/2 setup instead of just a serial terminal and the Propeller is a good way to do that in color and SVGA resolutions. The PIC terminal version I linked is only monochrome and has a much lower resolution (but should sync on a modern VGA monitor) - it fits in a smaller board area and also provides USB->Serial.
 
Feedback on S100 Z80 SBC board

Feedback on S100 Z80 SBC board

A board described by some of you here would without doubt be a far better S100 board. Having an SD card on board in particular would be a great addition.

Because I’m mostly occupied with getting high end CPU’s on to the S100 bus these days 80386, 80486 and ARM/Raspberry etc. + adding MB and GB RAM boards I had in mind the Z80/Propeller/USB board as a ‘quick & dirty’ solution just to get new people up and running on the bus. The components are well tested having gone through a few years/versions both in terms of hardware and the ROM monitor. With the USB port it would be possible to download a program into RAM from a PC, indeed download a CPM itself (this would require some more software). I am doing this board more or less in between fabrication of the above prototype boards – which take 2 weeks to turn around these days with the Advanced Circuits PCB manufacture. I hoping one or at the most two prototypes will do it.

If somebody else can put together a “proper S100 SBC” that will be great but honestly for me I cannot do the splice right now. Any volunteers?

BTW I just updated the S100Computers.com site with the proposed schematic for the above board. Freerouter did eventually find a solution (currently down to 654 vias). See:-

http://s100computers.com/My System Pages/SBC Z80 Board/SBC Z80 CPU Board.htm

I will let it stew over the weekend and send off to Advance Circuits on Monday. I will get the usual/minimum 4 boards made, so two available if anybody is serious about testing/changes of this prototype.

John
 
I appreciate your comments and am familiar with the Propeller (still waiting for the second version of the chip). It's an interesting chip; but then so are scads of other chips. You miss my point.

I've participated in S100 stuff (not so much in the last 20 years or so) in S100 stuff since 1975. S100 graphics are pretty much akin to putting a fuel-injected V8 on your push mower. I'm sure that someone at some time has done it, but it missed the point. Given that this is a 'vintage" list, VGA color graphics are hardly vintage for the S100 crowd--and, I think miss the spirit. Of all of the S100 systems that I encountered during the 70s and half of the 80s, not a single one featured graphics, regardless of the price. Even at Sorcim, where SuperCalc for the PC was developed, part of the work was done on a VAX 11/730, a few of Compupro systems (with 85/88 CPU cards) and a couple of IBM PCs. Only the PCs had graphics. Did CP/M or MP/M ever acknowledge the existence of graphics?. By and large visual input and output was almost always handled through text-mode terminals. Consider, for example, the popular Morrow MD2--no color graphics there.

Sure, you can strap a VGA interface to a Z80 just as well as you can strap a gigaflop vector math box to one. But what's the point? Certainly not "vintage" authenticity. A simple serial I/O interface to a terminal is really all that's needed or wanted.

As we look backward in time to see the real application of this technology, we're in danger of misapprehension of the real world. In most of the 1970s, computers not used in defense applications were used for simple financial record keeping--extending to such exciting stuff as inventory and payroll and banking. IBM held the bulk of that market--grab a random issue of Computerworld from mid-decade, say 1975 and scan the stories. Most news concerned itself with IBM or IBM-compatible mainframe issues. Unless you were in some special project or worked for a computer manufacturer, it was white-shirt-and-dark-tie boring work.

In short, I think it's best not to over-romanticize this stuff. Most serious S100 systems were not doing very exciting things--unless you consider Accounts Receivable exciting.
 
Last edited:
Chuck - I'm no S100 expert, just an 80's kid who, growing up in the 90's - fell in love with green phosphor screens and command prompts on hardware that everyone else threw out as 'obsolete'. I STILL prefer working on the command line to working in a gui - I hate mice. Anyway, while I agree that there is a tendency to over-romanticize the 'early days' of computing, whether you define that as the very first machines built, or the early minis, the first micros, or the 80's when the general public began to accept the notion of having a computer in their home - any and all of those periods are fascinating, and easy to romanticize, I also think that, in this context it is less an attempt to maintain museum-level accuracy, and instead to continue exploring what is possible in an entire area of computing that eventually got left behind by the continued 'appliance-ization' of computing. I rant and rave about the heath/zenith h/z-100's all the time, they were very MUCH graphics machines. They had three planes of memory, each dedicated to one of R, G, or B values, and thus provided a REALLY impressive set of capabilities that, in my opinion, was never fully explored. Compared to the graphics in a PC, well, no contest. Certainly no VGA, but it took a WHILE for the PC world to catch up to what it could do. Were those machines USED for that? Probably not. But, they were an entire computer engineered around the S100 bus as an expansion system. Far superior technology to the IBM PC (which I also love) and even having MS-DOS and an incredibly high amount of compatibility with PC software didn't give it the ability to knock the PC off the market. It might have been interesting to see where we would be if THAT machine, instead of the PC, became the 'standard' that took over.

My comments are certainly not trying to argue history - you lived through a period of time that I have only read about. I'm merely offering my view on your comments, and I think that a lot of the work that John has done really IS an exploration of where things would have gone and wound up had the S100 bus continued as the standard, instead of the PC taking over. It would be easy enough to duplicate (legality/morality aside - lets assume for this comment it were legal to do so) any S100 board of the time, and make them available new. The majority of interest that *I* have seen in the S100 realm is of two types - restoration of classic systems, usually with original boards - and the exploration of what *might have been* and now IS. My understanding is that DOS was not nearly as popular as CP/M MP/M on the original systems. But it's lack of graphics also probably helped the DOS/PC market to gain share, rapidly. Now look at it this way: If you have a computer capable of running MS-DOS on a 386/486, you have pretty much everything you need to boot at least an *early* version of the linux kernel. If you are truly savy, you can probably compile a custom kernel with 486 support and much of the 'modern' functionality that was added after older processors got dropped.

This SBC concept is meant to be the 'fast, dirty, one - weekend and I have a working computer with I/O and know that my power supply and bus are working, and I can start testing my other cards' system. This is exactly the kind of thing I want a couple of, that I can slap one into my compupro chasis, bring it up, and - god forbid - if I'm going to fry something, it WON"T be my much rarer and harder to replace original Compupro cards. SVGA is a convenience, and is handy because most modern monitors still have a vga port - and most people that are going to be working with this stuff certainly have what they need to hook it up. Serial isn't much harder to work out, either via an adapter or whatever, but - really, it's not about giving you a perfect replica of how things WERE, it's about getting you going - so you can restore your vintage system to historical accuracy - or explore the future that never was.

Think of this SBC as a tool to get started, and NOT the 'end goal'.

Sorry, I have rambled greatly, and taken great liberties in interpreting other peoples comments, motives, and history in of itself. So, if I am wrong - please let me know! Everything else is meant as conversation on a topic I am far too enthusiastic about by an ignorant fanboy ;) :D
 
Side note, ronwoch, we're of the same vintage. I was born mid-80s, and growing up in the 90s I fell in love with phosphor-screen IBMs, and got to experience going from my father and I building 286 PCs in my house up to 486, Pentiums, the rise of the GUI, etc. In fact, this is probably nuts, but there was a game on a 286 IBM PC I played that cemented my love for the Model M keyboard as well as gave me a weird connection to the IBM character set, for it used this strange clover-leaf character (Unicode code page 437, 0x06: ♠) that I hadn't seen on a computer before. Simply seeing an usual character on a green, phosphor screen created an emotional connection, and helped hook me on computers.

I have a collection of computers from the time I was growing up, but my interests have expanded now to see what computers were like before I was born, and I find it fascinating.
 
Back
Top