I designed an 8 bit computer and it seemed surprisingly simple after I'd taken a digital electronics class and understood how to build the bus. I never built it, but assumed writing an OS would actually be far more time consuming than building the hardware.

Looking at the circuit involved I'm not so sure it's as hard as it sounds... Sure, if you were designing the IC's and system architecture as well, (or designing for a system based around a modern high-speed CPU such as an X86 system, where many of the main IC's have over 1000 connections, and as a result require multilayer circuit boards just to route it all) then it'd be a different matter...

Yeah, the data and address bus look complex, but only because they each have 16 connections, but those connections are all doing pretty much the same thing.

On the Timex (same as Sinclair Z80?) we could load hex into BASIC "comments" & hit an interrupt to execute. Then it was a simple matter of writing assy. on napkins & paper, then translating to hex by hand ;-) Ah how I *don't* miss those days!

Mao ting assembly can take quite a long time, I've done a bit of asm myself, it's pretty tedious. (a suprisingly large number of people know Assembly for the z80. Texas Instruments used a version of that cpu in their TI-83 calculators, and the more clever kids figured out you could load and run assembly programs onto them, once you'd gotten low level access. imagine having to code in an image, if it's supposed ot be animated, like in a game, get to code in every single frame, and oh yeah, the 83 ran at like 1 mhz)

Mao ting Elders... ha! I learned in assembler and didn't write in high-level languages for the 1st 15 years of my career. Writing assy isn't really very slow, many (most?) arcade games of the 70s-80s were written that way by very small teams, in a fraction of the schedule modern videogames require. What is slow & hard about assy is *maintaining* that code as requirements change!

Mao ting When I first started with the ZX spectrum, I did not have an assembler. I used to write programs on paper, then I translate that to machine code. Next is entering them to the computer using poke command. Most of the time, the computer resets because I entered the wrong number, or mistranslated the assembly command. Later on I got "McGrowHill Assembler" which is a software that comes with a book called "Learn assembly Language for the ZX Spectrum" or something like that. You write Rem statements in basic, the the comments should be assembly commands such as: REM ld a,0 REM ld b,10 ... then you call the assembler which will translate your program into machine code. I remember making a simple side scrolling shooting game using that once which took me about one month to finish. Later on I gave it to my friend for him to have a look. He lost the cassette tape, and all was gone since I did not have any backup tapes. I almost cried. :(

Awesome!

My first computer: www.pdp8.net/

My uncle gave me a ZX-80 back in 1983-4, he used it for programming BASIC for uni, it was a cheap and cheerful solution if you were willing to assemble it yourself, after his uni days he upgraded to a more modern machine and handed it down to me. I was too young to appreciate it, played with it for a day or two, boxed it up and put it in the loft, I was far more interested with my Atari 2600, only after I got a taste for computing, programming and micros in general, the machine was long gone, I searched high and low for it, just couldn't find it. Been kicking myself ever since.

I had the TImex Sinclair 1000, those membrane keys were terrible! It was a fun thing to fiddle with though, I still remember having to set the volume just right on the cassette recorder to get it to load a program. Remembering what a program sounded like makes me sentimental for the old modem handshakes.

John Hamson I remember loading a Space Invader game for the Spectrum, and at one point it made a sound just like the intro of the Six Million Dollar Man... the noise when he says "Gentlemen we have the technology... we can rebuild him". Ha - America looked very exciting through my seven year old eyes... I thought you really had those bionic people.

Go on. It's not that hard nor expensive ;) .

I think you mean, back in the old days the demands to games was much lower than nowadays. Today everybody can make a game, so the market is overflowing with crap and there are only a few pearls, that are shining bright. And I think it was the same 15 years ago, only that the quantity of produced games was not that high. But I share the nostalgia feeling as well. So old games feel much better, than new games, even when they are not.

I am computer science student and I can tell you they teach embedded systems, assembly etc. on Z80 and you have to make some projects on it (either just coding or physicall ones with wiring and all kind of theese).

oh please do!

yes my Stepdad had one or 3

If you still have to breadboard and program it yourself, I'd hardly call it cheating. Besides, using a microcontroller to handle I/O is not an uncommon approach, both the 8048 and 8051 have been used for that purpose from what I remember. If I were to attempt something like this (and I'd really love to!) I would probably use an ATmega1284 as a "system chip" to do keyboard/video generation. But only because I'm too stupid to learn VHDL.

Fernrat I was honestly kidding about the Arduino part :) I use Arduino all the time in my prototypes, without using the board themselves except to program the chips or FTDI for the wearables. I'm sure you could learn VHDL, you made a replica Z80. It's obvious you have patience.

Most retros projects that replicate an old computer or present a new one based on old technology, cheat a little, especially with complicated stuff like video. In many old computers, the video hardware made up a big portion of the electronics, especially if tricks had to be used to share the video memory between the CPU and the video generator. Nowadays, it would be a little silly to put 20 chips on a replica board to make it generate video, if the rest of the computer only needs three chips and you can emulate the video with one Arduino or Propeller. I don't have a problem with that, or with the fact that replicas usually have a lot more RAM than the original, simply because it's cheap nowadays anyway. Also, although a lot of chips that were used in old computers are still available (and the 6502 is even still in production), custom chips and chips to generate video are usually much harder to find.

Jac Goudsmit Graphics hardware definitely seems more challenging. I looked into building something along the lines of an Snes at some point... (Comparatively much more recent than something like a Z80 I know), but while the CPU is a custom variant of a common CPU from that era, the graphics chip contains a huge amount of custom logic. Although at that point it starts to look so complex the only halfway reasonable way to replicate a chip like that would seem to be using FPGA chips...

he probably just used the arduino to reduce his chip count a bit as some interfaces can take quite a few chips to setup if you go the glue logic route in which case its better to use a microcontroller of sorts to handle it rather than half a dozen or more chips though it can be fun going the glue logic route for some things if your not worried about board space as you can learn quite a few things from it as to how it works

100101101st!

colouredmirrorball overflow?

Nope