Johnny's homebuilt 8080 CPU corner (and other stuff, maybe)
8080 CPU, built entirely by DTL-technology
Once in the beginning of time (late 1970's slipping over to early 1980's) I came into contact with CDP1802 (Telmac 1800) which I played a little with until I made the leap over to Z80 (Nascom 1 and Nascom 2). That was a fun period of my life. I never looked back to the CDP1802 again.
At that time, before the Z80 introduction, I had a romantic idea of building a small 4-bit computer. It seemed to be not too complicated and a not so far fetched thought, so I started sketching down a few scribbles on paper, invented a small instruction set (15 of them) and even wrote a tiny tiny assembler program.
As it usually happens, interest, time, money, job, career and so on, pointed me away from my small 4-bit computer too be.
The Nascom took over, and not so long after that, around 1983 - 1984, I got stuck in the software industry as a programmer. I'm still stuck there, knee-deep in the swamp.
My old Nascom was replaced with an IBM-PC and after that, I more or less left the hardware part of my life.
Fast forward to the year 2006: I can't remember what happened back then but my interrest in hardware came back and has been a companion since then.
Another small leap to winter of 2013-2014: Somehow, I got an urge to take up the thread I lost 30 years earlier, to actually build my small 4-bit computer.
A frantic time was to come when I tossed ideas back and forth, all rejected in favor for even more ambitious ideas - and then back to the newly rejected ones. I assume you know how it is...
Then, I switched over to an even more ambitious thought: Why not build a replica, or something similar, of an old 1970's CPU? How hard could it be?
My first inspiration was to go for a CPD1802. It seemed simple enough, despite what I always had believed to be its biggest drawback: The instruction set, and in my opinion, an awkward way of writing assembler code.
Next, naturally, the Z80 came up. A little too complicated, so I had to drop that. I peeked into the 6800 and 6809 (and even the 6502) but those was never any of my favorites.
Then, unexpected, in one of my "deep sessions" with Google, I stumbled over an Intel 8080 book from 1975, and that was it!
I spent the rest of the winter/spring (2014) up until late summer drawing up schematics after schematics, trying to figure out the best way to make this happen. I spent hours drawing up circuitboards after circuitboards, always finding out that I had missed something and had to start over. Again, I assume you know how it is (for some of us).
In all this, I got an interest in finding out how old mainframes was constructed. IBM 1401 is a machine I like, partly because we are of the same age (to the month), but more to it, I sort of got in love with the technology of using small circuit board (SMS card), where each card represented a logic gate or flip-flop, constructed with transistors, diodes, resistors, capacitors and sometimes with inductors.
And then, everything changed. I had earlier studied (and was a little envy of) Dieter Mueller and his MT15 16-bit transistorized CPU, and the craziest of all crazy ideas came over me: Why not build this 8080 "thing" with no more than simple DTL technology? Crazy, huh? The 8080 project and the IBM 1401 somehow merged.
Said and done. I spent the fall of 2014 reading about DTL technology, while trying to find the simplest of simple solutions, using the minimum number of components, but still get some performance. Sure, it would be a compromise, but it might work well enough anyway.
Soon I had a concept to build on. I created the necessary basic gates (AND, OR, NOT, EXCLUSIVE OR), and began to build up some larger parts, such as a D flip-flop (7474), a JK flip-flop (7473), and a transparent latch (7475). Then I took on me to build a 7483 (74283), 4-bit full adder with fast carry. And it worked!
The stage was set.
To join the Homebuilt CPUs ring, drop me a line, mentioning your page's URL. I'll then add it to the list.
You will need to copy this code fragment into your page.
Note: The ring is chartered for projects that include a home-built CPU. It can emulate a commercial part, that′s OK.
But actually using that commercial CPU doesn′t rate. Likewise, the project must have been at least partially built: pure paper designs don′t rate either.
It can be built using any technology you like, from relays to FPGAs.