The chip that changed my world – and yours

It lasted 50 years, but history finally claimed it. Zilog has called time on the Z80 CPU. Readers may have owned one in an 8-bit microcomputer or showered coins on one in an early arcade video game. Z80 CPU Zilog to end standalone sales of the legendary Z80 CPU READ MORE It was ubiquitous in the early to mid-'80s and popular …

[Doctor Syntax]

The Z8000 had a market in Unix boxes long before Intel got there. However it was being squeezed from the other direction by the 68000 and derivatives which I think were a little ahead in the market and the trendy RISC stuff like MIPS coming along a little later. What might have killed it there was Zilog making its own systems due to its new owner Exxon's ambitions. The vendor eating its potential customers' lunch doesn't go down well. Then the Intel/SCO partnership ate everyone's lunch.

[toejam++]

I've always wondered how the Commodore 900, with its Z8001 running Mark-Williams Coherent, might have fared had Commodore not purchased Amiga.

[Dan 55]

I don't see why Commodore couldn't have sold both machines at the same time, they were different markets. The A1000 was a graphics and sound desktop computer, the 900 would have had text output or hi-res black & white bitmap graphics which would have competed aginst UNIX workstations.

That would have required a marketing department which knew how to market though.

[ecofeco]

It's not hyperbole that Commodore's marketing is what killed them.

[Empire of the Pussycat]

I'll raise a glass to the Z80

I bought a Nascom 1 with one of my first pay checks, what I learned kicked off decades of, mostly, great fun doing what I wanted and getting paid for it.

Beer icon only because there isn't a champagne one.

[Stuart Castle]

Re: I'll raise a glass to the Z80

In fairness, based on my own experience with blokes who got interested in programming during the late 70s and early 80s, the beer icon is probably more appropriate than the champagne one. Particularly if it's a beer or Ale only brewed by a small brewery, where both the brewery and the beer have a wierd name. It should preferably served in a pub rated highly by CAMRA, but if not, any decent non chain pub should do.

[MyffyW]

Re: I'll raise a glass to the Z80

Ah yes, a pint of Old Inscrutable up The Old Priest Hole (a public house, not a medieval indiscretion) where the pump label is printed in Braille and there's part of the original brewers foreskin at the bottom of your pint pot.

(with apologies to a certain Mr B Elton from the 1980s)

[Ferry Michael]

Re: I'll raise a glass to the Z80

My first programming was on a Nascom1 my brother had borrowed. Writing assembly on paper and tehn assembling by hand I learnt the instruction set well. "Progressing" to just typing in hex without any source code. I came unstuck when refactoring my code.

[PRR]

Re: I'll raise a glass to the Z80

> just typing in hex

My Dad, after a stint with a 7 Dwarf mainframe company, threw some early Intel chips together but programmed them in Octal, which was for Big Machines. He was very big on desk-checking but I know he could improvise Oct if he wanted to.

[dunbankin]

Nascom 2

This report about the death of the Z80 brought back happy memories of my self-assembly Nascom 2, and I followed a career as "the techie guy" in business as a result of what I learned about machine code and programming.

[Invisible Code]

Re: Nascom 2

Ah the Nascom II, I took assembled mine from a kit. It came "unpopulated" as those static RAM were in short supply and I used to code in hex in the monitor space. Worked a whole summer to afford a 64K memory board. All in the museum (attic) now

[Neil Barnes]

The Z80, however, was compatible with a rival, Intel 8080.

It probably didn't hurt that the same bloke designed both.

Had Nascom accepted my credit card, I would probably be Z80 fan (though oddly I preferred the 8080 code and when I've worked professionally with the Z80 it's always felt a bit odd); instead, Microtan led me down the 6502 road. So to avoid any fights, a beer for all!

[martinusher]

Re: The Z80, however, was compatible with a rival, Intel 8080.

Strictly speaking, it was compatible with the 8085, not the 8080.

The 8080 was probably too primitive for most users since it was built using an early pMos process that needed four supplies and the chips were so fragile that they'd die as soon as you'd even look at them. The 8080 also lacked some instructions that the 8085/Z80 had.

I had to do 'professional' work with the 8085 because although the Z80 was a better part it wasn't good enough for large organizations like BT, even back then there was institutional inertia that tended to favor Intel (better marketing?).

[the spectacularly refined chap]

Re: The Z80, however, was compatible with a rival, Intel 8080.

The 8085 came after the Z80, although it wasn't influenced by it - development was well underway by the latter's release. There were only two new instructions in 8085 relating to the built in "serial port" (bit banged) and interrupt masking. I suspect you are confusing the 8080 and 8008. Z80 had lots in contrasts, mostly additional addressing modes.

None of 8080, Z80 or 8085 were pin compatible with each other, although the latter two both needed only a +5V supply. Z80 separated the address and data buses but 8085 kept them multiplexed. Horses for courses, since that means extra logic elsewhere for timing and status signals as they are all constrained by a 40 pin DIP.

Strange, I still think of multiplexed address and data buses as "Intel style" despite the last such chips being the 80188 and 80186.

[Number6]

Re: The Z80, however, was compatible with a rival, Intel 8080.

I think officially, to be a CP/M program, you were supposed to stick to the 8080 instruction set.

I used to know the Z80 assembly stuff off by heart - I could write the assembly language and then go down and write all the op codes, effectively hand assembling it. Forgotten most of it by now though, although somewhere I do still have my home-made Z80 computer with 7-segment LEDs and a hex keypad. I think it may even still work if the EPROMs haven't given up.

[Bitsminer]

Re: The Z80, however, was compatible with a rival, Intel 8080.

...the chips were so fragile...

I was debugging an 8080 made with the white ceramic and gold-plated cover. Damn thing popped on me, and the chip cover just missed my eye.

The electronics engineer (ex-Atari, ex-MIC, ex-California) who had designed the circuit board looked it over and said "lucky you."

Never understood why or how it managed enough oomph to blow.

Anybody remember the HP 1641 in-circuit debugger with the 40-pin clip-on cable? Now that was a marvel. It recorded and displayed 8080 instructions in assembler upon hitting a trigger address.

[Andy 73]

All is not lost.. yet.

It's true that the Pi keeps you at arms length from the hardware, safe in Linux and unaware of the complexities of bringing up a modern SOC - but that doesn't mean there aren't still kits and computers that you can build and learn from that don't wrap you in a slightly smothering blanket.

If you are interested in learning, or just returning to past experiences of CP/M and 8-bit bare 'operating systems', there are brand new, fun kits and single board computers out there - and for now at least, the end of manufacture of the Z80 won't stop them being available. The retro and hobbyist computer scene is thriving, not least because people do want to use machines where they can have complete control and complete understanding of the entire device, not just a shallow 'user layer' and command prompt.

[Jason Bloomberg]

Re: All is not lost.. yet.

There's also the Raspberry Pi RP2040 available as Pico, Pico W (Wi-Fi, Bluetooth) and as third party offerings - "as cheap as chips".

More an SBC than a home computer but it can become one. And it's pretty capable, running a BBC Basic, MicroPython and even Fuzix if you want something Linux-ish.

I would accept the ZX80 and ZX81 likely were the cheapest home computers but I wouldn't say that was because of the Z80, more through other cost-cutting. They were great for what they were, but it appears to me 6502 ultimately won in the home computer field; Atari, BBC, Commodore. I wanted a 6809 Dragon but could never afford one.

[Dan 55]

Re: All is not lost.. yet.

I would accept the ZX80 and ZX81 likely were the cheapest home computers but I wouldn't say that was because of the Z80, more through other cost-cutting.

The Z80 could refresh cheaper DRAM chips without any support chips, I think the 6502 needed a support chip to do that.

They were great for what they were, but it appears to me 6502 ultimately won in the home computer field

This is where I play the MSX, Commodore 128, Amstrad CPC and PCW cards.

[Number6]

Re: All is not lost.. yet.

IIRC, the way the BBC Micro did it was to use the graphics hardware to effectively provide refresh, so it didn't really need anything extra. The problem with the Z80 refresh was that it was only 7-bit, and some of the 64K DRAM needed 8-bit refresh.

[Matt Dainty]

Re: All is not lost.. yet.

While the Commodore 128 does indeed contain a Z80, it's only really used when running CP/M software. When running C64 software, (as a C128 tends to end up mostly doing), it's running on the 6502, albeit an 8502 variant.

[fromxyzzy]

Re: All is not lost.. yet.

The 6502 won in the US consumer market, the Z80 won in most of the rest of the world. MSX, CPC, Spectrum and the dozens of clones. The 6502 was chosen as the basis for designs prior to the point where the Z80 became cheaper, hence it was retained for the major American brands for years as they kept updating their late 70s designs. The rest of the world came a bit later and chose the Z80 because it was more cost-effective and they didn't have the baggage (also the eastern bloc countries figured out how to clone it, which is why they took to the Speccy so much).

Tandy/Radio Shack could have helped make the Z80 more ubiquitous in the US, but instead they threw compatibility to the wind and just picked whatever chip struck their fancy for every new model. Hence, the Dragon.

Z80 remained the processor of choice for business computing, but that's a whole other market really and CP/M was all that mattered there.

[Dan 55]

Re: All is not lost.. yet.

You can have a bare metal Raspberry Pi if you want, but I think the learning curve could be a little steep.

[Andy The Hat]

Designed by a software engineer

The 6502 was a hardware tinkerer's device, the Z80 was a programmers device! I remember mapping the binary instruction set and being surprised how many of the "missing" instructions worked as expected ... Unsupported code obviously but for a geeky comp sci programmer it was fun at the time.

[Number6]

Re: Designed by a software engineer

I think some of the unsupported CB codes, SLI (Shift left and increment, basically shoving a 1 in instead of a 0) affected the flags differently to the related instructions, which is presumably why it wasn't on the official opcode map.

[DS999]

Re: Designed by a software engineer

The 6502 had similar undocumented instructions.

[Stephen Wilkinson]

I started with a ZX-81 in my early teens, not for very long, probably less than a year before I jumped to a Commodore VIC-20 but without the ZX-81, it's unlikely I would have got into the IT world so thank you.

[Howard Sway]

It lasted 50 years, but history finally claimed it

Considering how much things have advanced in that time, it's almost staggering how it survived that long, but some requirements are so simple that there should always be a need for simple chips that do simple things, without any superfluous functionality that is overkill for the job. This makes me wonder exactly why they have chosen to end its story now, and 2 possible explanations spring to mind :

1) The machines that make them have more or less conked out, and would be too hard to replace

2) The people who program them have more or less conked out, and would be too hard to replace

[Electronics'R'Us]

Re: It lasted 50 years, but history finally claimed it

The announcement stated that the supplier was unable to continue fabricating the part.

That is hardly surprising; it was on a 4 micron node (IIRC) and the fabrication equipment needs maintenance and spare parts which are likely to be totally unavailable now.

It is interesting how some parts live on where you least expect it.

The venerable 8051 is often found within a wide array of controllers (running at many times the clock speed of the original); I have seen them used widely to implement state machines in USB hosts. There are numerous 8051 devices in a weapon system designed in the late 70s / early 80s, incidentally. I suspect there are many such examples.

Many problems can be solved with an 8 bit device running at a few MHz, so it is hardly surprising (in that sense) that such parts are still widely available.

[Ferry Michael]

Re: It lasted 50 years, but history finally claimed it

8051 was still being used in some mobile phone SoCs 5 years ago as a PMIC for bigger ARM processors. I think 8048 is still present in modern PCs occupying a tiny fraction of some bigger device

[G.Y.]

8051 in PCs Re: It lasted 50 years, but history finally claimed it

I believe the canonical PC keyboard has an 8051 in it

[Number6]

Re: 8051 in PCs It lasted 50 years, but history finally claimed it

The original keyboard controller was an 8042, I think.

[TReko]

Re: It lasted 50 years, but history finally claimed it

The 8051 lives on, but as a simple way of adding programmability to other devices. Some software defined radio IC's have it embedded.

[Admiral Grace Hopper]

Re: It lasted 50 years, but history finally claimed it

I remember the feeling of surprise I felt while dissecting a washing machine when I found a 6502 processor sitting there. I spent many happy hours of my youth writing assembler for that chip, but only after (back on topic) cutting my teeth learning the base instructions for the Z80 chip on a ZX81.

[Toastan Buttar]

Re: It lasted 50 years, but history finally claimed it

As part of my job, I did some prototyping work on a Silicon Labs Busy Bee development board. It has an 8051-based MCU at the heart of it. It was great fun to program a device somewhat similar to a turbo-charged ZX81. The board had its own black-and-white LCD display, 64K Flash and 4.5 K RAM, with a pipelined processor core running at up to 50MHz! It could even be powered by a button cell for portable use. Specs you could only dream about in the early '80s!

[ITMA]

Re: It lasted 50 years, but history finally claimed it

There are a few integrated circuits which "bit the dust" not because they were "obsolete" or such like. But because the manufacturers had "lost the masks".

LOL

[David Hicklin]

Re: It lasted 50 years, but history finally claimed it

>> Many problems can be solved with an 8 bit device running at a few MHz, so it is hardly surprising (in that sense) that such parts are still widely available

These days a small PIC chip can do most far more than one of these, I love playing with these babies

[heyrick]

Re: It lasted 50 years, but history finally claimed it

"Many problems can be solved with an 8 bit device running at a few MHz"

Yup. Bread makers, microwaves with push buttons, smarter toasters... When I look up the part numbers, they're either based on a SAM8 or - much more frequently - on an 8051 clone.

What surprises me is that there's very little Cortex-M around (maybe they're too powerful?) and I've yet to see anything from the PIC family.

I'm guessing the 8051, due to its age, has a hell of a lot of support. I just looked up "8051 PWM" and... https://www.8051projects.net/wiki/Pulse_Width_Modulation

[Mage]

Re: It lasted 50 years, but history finally claimed it

I think PIC originated as 1684 in 1976. Maybe originally prototypes used EPROM inside?

Most remote controllers use a PIC.

The main advantage of later 18Fxxxx PIC is the Flash memory, loads of I/O, and packages from 16 to maybe more than 64. Some have done USB for years just with addition of a capacitor.

[Pascal Monett]

Re: there should always be a need for simple chips that do simple things

I totally agree. Nobody is going to put a Core3 Intel CPU in a "smart" doorbell. So there has to be a simple chip in today's tech that could do a job that does not require umpteen billion transistors if you know how to program it.

What might that spiritual successor be ? What chip is being used in all those damned IoT thingys that are being sold by the thousands every day (for the greatest joy of malware hackers everywhere) ?

[Missing Semicolon]

Re: there should always be a need for simple chips that do simple things

Atmel ATMWEGA328?

Or a little more esoterically TI MSP430 series?

[Rich 2]

Re: there should always be a need for simple chips that do simple things

I was working in MSP430 right up until today. My resignation notice period just ended - not because of the 430 though :-)

[Number6]

Re: there should always be a need for simple chips that do simple things

MSP430 series was nice, had a chance to play with it once. A nice compact 16-bit architecture, made it better than the 8-bit stuff because we needed 16-bit math and the 8-bit chips weren't quite fast enough.

[Mage]

Re: there should always be a need for simple chips that do simple things

Microchip (PIC) bought Atmel?

[martinusher]

Re: there should always be a need for simple chips that do simple things

>What might that spiritual successor be ?

ARM claimed that space before they went all upscale due to making parts for phones. Now its the world of PICs and ESP32. Most of the time the problem is thinking *down* to that level, its really difficult to make a part that big with that (lack of) performance. If you think about it then a modern Z80 would be practically invisible and in order to justify making it you'd probably want to throw in the rest of the computer -- memory (RAM and non-volatile), peripherals, everything. But then pins are always a problem so you'd probably want to add SPI, USB and maybe a network port or two, maybe even wireless, rather than a full memory bus.....you can see where this is going...

If you're really deseperate to run a CP/M in a Z80 these days then there's excellent emulator programs around for the PC. They work just like the real thing and use so few system resources that even a mediocre PC could run a couple of these without noticing it. Puts those old systems in their place!

[Mage]

Re: ARM, before they went all upscale due to making parts for phones.

Um, the Cortex such as M0 was going downmarket to controllers. The smartphone have used ARM instead of custom 486 since about 2001 or 2002.

ARM started with the Newton PDA and Archimedes workstation (RiscOS and maybe UNIX by 1987?).

[Anonymous Coward]

Re: there should always be a need for simple chips that do simple things

"What might that spiritual successor be ?"

Microchip's PIC line fits the bill* pretty well for a lot of use cases. What I like about Microchip is that the PIC family scales from "dirt cheap, tiny, program in assembly if you want to do anything meaningful" to "run a RTOS, lots of IO available, don't try assembly, there's way too much to keep track of".

*or it used to, I've been away from microcontrollers for a few years now.

[David Hicklin]

Re: there should always be a need for simple chips that do simple things

They still fit the bit and are quite easy to program, the hard part is selecting which chip to use!

[Mishak]

Don't forget devices from the STM32 family

I'm using an STM32G031 device at the moment for a small actuator in the "large" 8 pin SOIC package (I can't hand solder the smaller package with more pins) that costs virtually zero, is 32-bit, has on-board RAM and FLASH, loads of peripherals.

At the other end, for not much more, I've been using other members of the STM32 family for touch screen industrial control systems.

Though it's not really fair to compare a CPU with an MCU.

[Linker3000]

Gonna be taped out!?

There's an intention to reproduce a version for the die-hards (heh!)

https://studio8502.ca/@mos_8502/112349800192558629

[Toastan Buttar]

I owe my career to the 1K ZX81

"A 1K tape-based monochrome machine with no lower case, sound or graphics doesn't look like much. But compared to no computer at all, it was magnificent."

And that's it in a nutshell. The 1K ZX81 actually gave you the ability to program a genuine computer and watch it carry out your instructions. Unless you lived through that period, it's difficult to imagine the thrill of that experience. A few brave souls (myself included) decided to roll up their sleeves and get into machine-code programming in order to overcome the speed and space limitations imposed by Sinclair BASIC.

To this day, I've been more drawn towards low-level coding. I'm fortunate to work for a company that specialises in embedded designs, mostly featuring low-power processors or microcontrollers. But without that early advantage of ACTUALLY owning a REAL COMPUTER in my early teens, who knows what path my career might have taken?

[J.G.Harston]

Re: I owe my career to the 1K ZX81

I'm fortunate to work for a company that specialises in embedded designs, mostly featuring low-power processors or microcontrollers.

Ooo, who? Any vacancies?

[Phones Sheridan]

"But compared to no computer at all, it was magnificent."

I suspect you're just a tad older than me. For me I owe my career to the Spectrum 48. I didn't know this at the time, but we were quite poor in the 80s. The government of the day were killing the large UK manufacturing companies one at a time, and my dad was unemployed like most other skilled engineering workers of the time. My parents literally saved up pennies per week for 2 years, by which time the 48K Plus had been released and they got me one of those instead of the rubber keyboarded original, along with a black and white TV to go with it in my bedroom (I didn't see colour until I saved up for and bought a Sam Coupe' on it's release with my paper round money and a colour 14"" portable).

I'd say the price of the Z80 / Spectrum is probably what got a lot of people including myself into an IT career. The relatively cheap-as-chips price of the machine got it into many more bedrooms than would have been possible had it been more expensive. I remember drooling at the C64 colour graphics in games when round at the house of the 1 kid at school that had one, but there were dozens of us with the Speccy, and I guess it came down to your comment, "...compared to no computer at all, it was magnificent.".

[Plest]

Re: "But compared to no computer at all, it was magnificent."

Ditto. My parents barely had enough to pay the bills but my Dad used a credit card to buy a Dragon32 home computer back in the day, he told me a spotty little sod of 9 years old that working in computers would be the future, he didn't have a clue how or why, he just had a "gut feeling" it was my future. A few years later my Dad got a better job, he then bought two Amstrad CPCs ( used the venerable Z80 and which I learn assembler ) , one for me and one for my mum so she could work from home. Then the Amstrad 1640 PC in 1987, a genius move to let a nerdy teen like me loose on a full blown PC, then I headed off to work in the IT game around 1990 I had all the skills ready to go and got snapped up for some very well paid gigs.

35 years working in IT, house paid off in my 40s and fingers crossed heading towards early retirement, all thanks to a "gut feeling" my dad had. My dad is 83 and still tinkers with PCs and his arduino collection, he was there when I started in computers and likely to still be around to see my working career come to end.

[H in The Hague]

Re: "But compared to no computer at all, it was magnificent."

"... all thanks to a "gut feeling" my dad had."

Your dad deserves one of these --->

Cheers.

[Phones Sheridan]

Re: "But compared to no computer at all, it was magnificent."

A quick google tells me the release price of the Dragon 32 was £200, the same price as the Speccy at the time so looks like your dad made the same choice as mine to get the one that was less than half the price of the C64. Was it by any chance the only type of home computer they had in the shop at the time because they were rare. I had 2 friends with the D32, and I recall the games were either white on black, or 4 colours of radioactive green, yellow, red and blue. I remember there was no lower case in the character set, so to get a lower case A, it was a normal capital A, with the colour inversed. The one thing I always loved about it in spite of this, was that the keyboard was like a "real computer". Much better than either the speccy rubber, or the square keys of the 48K+, but also you could type properly one character at a time. On the speccy it was 1 keypress per keyword, meaning you had to learn the locations of the keywords you were wanting, and the key combinations needed to get them.

[J.G.Harston]

Re: "But compared to no computer at all, it was magnificent."

Yeah, I did a paper round for two years to save up for my Spectrum, and a second-hand 12" black&white TV. I'd been "into" electronics for more than five years prior, and this "beast" was so fascinating, especially the way I could build hardware expansions and get it to control them. I built a centronics printer interface, a 8-bit LED output thingy (ooo, knight rider!) ASCII keyboard interface, joystick interface, built a expansion backplane system out of case-off parts to minimise having to buy loads of edge connectors. But if you tell kids today..... :D

[Mage]

Re: I owe my career to the 1K ZX81

The later Jupiter Ace was nice.

Forth was much nicer for test gear than the Z80 controller I designed in 1981 as it only had assembler. Last PCB I ever taped. Prototype used that enamel solderable wire on a sort of pen with spool at the top. Showing that PCB to some corporate guys got me my first corporate job and then I bought the Jupiter Ace. We did have a Spectrum in a previous small company just to provide test cards for monitor customisation.

[laughthisoff]

"Readers may have owned one in an 8-bit microcomputer or showered coins on one in an early arcade video game."

'Owned'... past tense?!? 'own'... present tense!!!

And 'one' ... ? Um... more than one. *cough*

#retrogaming

[Plest]

I wanted to but as my wife put it, "You're not buying a load of old crap off eBay and setting fire to the bloody curtains! You make do with emulators.".

Things we do for love!

[dharmOS]

eZ80?

I am checking that what has been lost to new production/ purchase is the 40-pin DIL Z80 (I used one in my A-level Electronics course work)?

I bought an Agon Light 2 recently and the eZ80 seems to be all round a better chip and binary compatible.

[deanb01]

Re: eZ80?

The eZ80 is an interesting chip - it's binary compatible but not a drop-in replacement, even if a suitable adaptor board was available. For example, it has a three-stage pipeline which means that the same code running on an eZ80 clocked at the same speed as a Z80 will run much quicker. As a lot of code on retro computers is timing critical, then this would cause issues.

And it operates at 3.3V rather than 5.

So in the long term replacements for retro computers (ZX81, Spectrum, MSX, Amstrad, etc) will no longer be available as new stock. And there are many new boards that use the Z80 (RC2014, Microbeast, etc).

[Boris the Cockroach]

Real programmers fiddle with

bits.

Yeah I'm another who cut his programming teef on a Z-80 (homebuilt ZX81 to be exact, before moving onto a spectrum and a memotech 512)

Always remember my old instructor when I was less than a PFY "What you learning all this computery stuff for? you need to learn howto hit bits of metal with rocks" but I was always gazing at the big CNC mill and thinking "thats the future"

So the Z-80 programming came in handy when finding out just how basic CNC is... until you get to more complex jobs involving robots, and on-machine measuring, and multi-spindle/axis programming

Still think I could knock up a 16 bit PLC controller in less than a week using a Z80, an EPROM, some battery backed RAM , a couple of PIOs, a CTC and a DART. (the knowledge is old... but still there)

[J.G.Harston]

Re: Real programmers fiddle with

As I mentioned in the parent article, that was the attitude of the people I was working with when I was doing work experience in 1985 building a super-fast ADC interface and writing the software to read from it. Fascinatingly interesting stuff that I found I had a natuiral aptitude and skill for. But: STOP DOING THAT!!!! YOU'RE WASTING YOUR TIME, IT'S USELESS!!!

[readman]

I purchased the Z80 card for my Apple II+. I was into assembly programming the 6502 and was looking to learn the Z80. I missed the 6502 memory mapped IO.

[Number6]

I did the RAM-less Rugby MSF clock on a Z80 once - saw someone had done it and so figured out my own version. Still have that around here somewhere, minus the 60kHz radio.

[RobThBay]

ZED

Good idea using ZED80 instead of simply using Z80. That'll stop the Yanks from calling it a ZEE80.

[deanb01]

Re: ZED

I guess this is why the eZ80 is so called - it's not E-ZED-EIGHTY, it's E-ZEE-EIGHTY (Easy 80).

[CowHorseFrog]

Zilog and variants like Z8000 etc died because they never tried anything REALLY different. Intel brought protected mode in the 286, and then virtual X86 and more, all zilog ever did was add a few extra instructions, the ISA never really changed.

[_andrew]

That's not true at all. The Z8000 was quite completely different to the Z80, fully 16-bit with instructions and addressing modes to match, and it was paired with a memory management unit that did protected segmented memory. The first Unix system that I touched was some sort of Z800[01] box running Xenix, I think. The problem, as I've since read, was that they used a hand-designed (and laid-out) network of discrete logic for instruction decoding, instead of a more structured microcode ROM system, and they never got all of the bugs out of it. I'm not sure how that manifested (probably instructions that ought to work but didn't?), but it probably contributed to the demise. The 286 did its segmentation and memory management in the single chip (only the FPU was an add-on) and what's even better, you could bypass it if you wanted to run an unprotected operating system like MSDOS.

DEC's LSI-11 chip(set) was contemporary with the Z8001, and it didn't get very far in PC-land either, although it wound up in quite a bit of industrial control systems.

[TBi]

And the Z8000 had sort of two operating modes, one segmented and then the other sort of completely flat. At around the time it was available I was designing and building an OS for a now long gone Italian computer company and there were many debates as to which mode we should use. It was also touch and go as to whether we would use the Z8000 or the M68000. Eventually for various financial reasons the Z8000 was chosen and we were dissapointed although we did get a free Unix based system (based on Z8000) from Exxon to play with (and use).

As regards bugs, they were legendary. We were coding in C but had often to resort to assembly to work around the bugs. And when new devices became available, there were other bugs.

It kept us all amused....

[CowHorseFrog]

Yes the Z8000 is different in SOME ways... but its very much the same.

*registers*

Its still doing the bank of registers and grouping pairs into a large register, so that hardly counts as a BIG change.

*addressing modes*

https://sourceware.org/binutils/docs/as/Z8000_002dAddressing.html

They all seem basically the same as the old Z80 modes, just with bigger offsets and bigger registers.

Its basically the same cpu with everything just stretched to 16 bits and the old Z80 16 bit register pairs streetched to 32 bits Z8000.

THeres no FPU, no SIMD, the later weree invented later by others.

It doesnt even have a MMU, the MMU was optional.

[Dan 55]

It wasn't the "brain dead" (billg) 286 protected mode which meant Zilog lost 16-bit and 32-bit, there were more things happening than that.

So Zilog decided to retreat from the Z8000(0) to the embedded space where they still are today with the Z8, eZ80 and other chips.

[CowHorseFrog]

Im not saying 286 protected mode was genius im just saying that intel tried to move things forward, while zilog did nothing.

[Charles E]

8 bit processors

I remember attending a homebrew computer club meeting at MIT circa 1974 on this upcoming 8080A chip, and how much better it was than the 8008 which was unobtainable and you might be able to actually get your hands on an 8080A. Hey assembly and higher level languages? Shut up and take my money. I built an 8080A S-100 system, it's still operational. In college, I used the 8085 SDK boards in our microprocessor lab to control stepper motors and other small circuits, it had a breadboard built in. The 8085 was intended to reduce the cost of the support chip set and it was competitive with the Z80. But the whole 8 bit generation of processors quickly got skipped over for the 8086 and other 16 bit processors.

[Mos Eisley Trooper]

The Z80 determined my career path

For Christmas 1982, aged 10, I got a second hand Sinclair ZX-81 when all around me were getting Spectrums. Whilst they played games (which I admittedly and enthusiastically partook in when visiting), myself and one other ZX-81 owner I knew started to learn coding since the games didn't have the same appeal (even with the 16k ZX-Panda RAM Pack acquired 7 months later on my birthday...though I do still have a soft spot for 3D Monster Maze). Initially we typed in listings from Sinclair User which helped learn tricks and techniques, but then I started writing my own things. I loved programming that little computer.

After that I moved onto an Acorn Electron (parents couldn't afford the BBC Micro I desired) so got into BBC Basic and 6502 assembly, and then by 1987 I was a PC owner (Amstrad PC1512, two 5.25" drives and no HDD until I added a 32mb hard-card after a couple of years), which set me down the Pascal then C route.

I'm now into my 35th year as an IT developer - and it's all thanks to the Z80.

Cheers, little chip !

[werdsmith]

Re: The Z80 determined my career path

I had no chance of getting my own. A teacher brought one into school and allowed us lunchtimes on it. Otherwise I don't know what I would have done.

I keep a ZX81 on the shelf here to remind me how lucky was with that teacher.

[Caver_Dave]

Re: The Z80 determined my career path

Ah "Sinclair User". I had assembler code produced in there when I was still at school.

Went on to Higher Education and only lasted half a lecture on the Assembler Programming course which happened to be on the Z80.

I produced code (for payment) for 6 different people on that course - each with their own foibles, which I had written down, so all of the code was slightly different in consistent ways.

I only earned 99% for that course - I was docked 1% for not attending the lectures!

[rlightbody]

We got our ZX81 around 1982. I was 9, and was blown away. I learned basic programming (because the tape drive didn't work, so if you didn't programme it, it didn't do anything). The BASIC books that came with it were an excellent introduction to programming. I'm sure it led directly to my career in IT. The wobbly 16K RAM pack that would reset the computer if nudged, taught me lots too!

The ZX Spectrum + that followed it in 1985 was, by comparison, mostly used for games.

[Missing Semicolon]

Lower-case ZX81

I succeeded in making an external I/O memory board, with 8K of SRAM, and an 8255 for I/O.

By adding an extra ribbon cable connecting to the processor pins, and a bit of finagling of (I think) the IX register, I succeeded in convincing the ULA to get characters out of a block of the SRAM.

So, user-defined characters! And, by preventing the MSB being used to trigger inversion of the matrix bits, the inverse characters became lower case.

I built a noddy video mixer on the input to the modulator, so that by probing around the circuit, I got lighter-white areas on the screen for signals that were high.

[HPCJohn]

Lower case TRS80 and sound

"machine with no lower case, sound or graphics doesn't look like much"

I have a TRS80 Model 1 which may be one of the few to have lower case.

How come? The character generator can generate lower case characters. However the memory was seven ships wide.

You solder an 8th chip on 'piggy back' to an onboard memory chip then bend the addresss pin up and solder it to the correct trace.

Viola - lower case TRS80

As for sound you hook a speaker up to the Cassette Put plug and pulse that to make sounds.

[Old Cynic]

Fortunately these will never really be gone (just like the 6502 et al), they will remain, be prodded, poked and programmed for for eons to come.

The benefits of these simple-processor computers to individuals as well as industry is, in my opinion, incalculable, almost like the steam engine to the industrial revolution. But, the onboard BASIC interpreter ,I believe, was the gotcha that really hooked beginners. The ability to turn on a print "Hello" all over the screen with a couple of lines, leading to more complicated things was the spark that led some people into the real heart of the machine, into the Z80, 6502 and to explore the support chips and other hardware.

Without that initial gotcha, I might have just played games and not bothered trying to conquer machine code and understand the nuances of the VIC2 & SID chip or learn how to mask software sprites on the weird screen memory of the old Speccy. Most 'Developers' these days would benefit from the experience of trying to create efficient code on one of these old machines where there were no libraries, no internet, no sample code. Just you, a flimsy reference guide and the desire to create something.

[Healeyman]

A Classic

I know the Z80 from my reading Ciarca's 'Circuit Cellar' and my career--long ago--at HP. I spent years, both in school and professionally, programming on HP's 2500/1000 computer, which was initially developed to control HP's instruments (using HPIB, of course). The later version of the 1000 I worked on had a 16-bit CPU--earliest versions used 4x4-bit CPUs 'bit-sliced' together--but its power came from its OS, called 'RTE' for Real-Time Executive and the architecture. All I/O operations were handled by all-but standalone Z80-based microcomputers on a bus, to which the CPU would issue read/write commands then be on its way. The I/O board, for example the serial board, with several UARTs, would send or collect the data, DMA it to the CPU and interrupt the CPU via dedicated circuitry. With this architecture, it was a data collecting machine par excellence; when HP went to RISC ('PA-RISC') for its business machines they couldn't even come close to 'real-time' as all I/O was handled by polling relatively dumb I/O interfaces. In benchmarks, the lowest-powered HP1000s--by then on a single CPU chip--would blow the doors off the most 'powerful' PA-RISC/HP3000 business machines in benchmarks (you cannot ever call a computer which requires polling 'real-time,' as response cannot be deterministic by definition but, with enough CPU speed they can get close).

My career at HP started in the early 80s. At the time, HP had been contractually committed to supporting the HP1000 through, IIRC, 2010 as it was the primary controller for the USAF's AWACS radar surveillance aircraft.

[Invisible Code]

6502

Weren't the Superboard 2 and Compukit UK101 6502 based (kit) single board computers that were around before the ZX81? Couple of guys had those at uni in the late 70s early 80s. Those were preceded by the KIM variants if memory serves correctly.

I went down the Z80 route and started work at a lab that had Intel MDS systems developing in PL/M for industrial controllers.

[elaar]

Even when better processors were used in Video Gaming, the Z80 often still took a place as the sound CPU in Arcade pcbs all the way through to the mid 90s.

If it died, you replaced with a new one (costing peanuts).

[archie99]

NSC800 anyone

I was working with the NSC800 which was a CMOS 5V version of the Z80 in 1984. It was great for creating low-power embedded control units. I'm surprised it doesn't get a mention

[jezza99]

Exidy Sorcerer

My first computer was an Exidy Sorcerer. This little known computer was Z80 based, supported full ASCII and, by using RAM chips for the top 128 characters of the character generator, supported programmable graphics.

It was possible to add an S100 (bus) controller via a proprietary expansion connector, and the computer ran CP/M very successfully with the addition of floppy disks. The original, of course, used a cassette interface to record programs and data.

It supported plug-in ROM packs with software, using 8 track cartridge shells which were cheap at the time. While the design limit for RAM was 32K, it was possible to expand to 48K by piggy backing an extra row of RAM chips and connecting the right address lines. Ah, the things I got up to in my younger years!

Exidy was an arcade games manufacturer and didn't know how to market the Sorcerer. Dick Smith in Australia sold more than Exidy did in the US.

[Damion2]

Nostalgia!

This article almost brought me to tears remembering my childhood. Like many others I attribute my playing on z80 based systems, to a successful career in tech.

I started with a second hand zx81, we didn't have much money, so anything like a BBC was completely out of the question. I was only about 8 but enjoyed writing software rather than playing. I think learning basic helped me academically, for example I think I grokked algebra faster then my peers by the time we covered it in maths. Later I got crazy into coding in z80 assembly after I'd saved up for a spectrum, I eventually got an assembler but was adept coding direct in hex, often via the multiface (shout out to romantic robot!). I payed with interrupt handling and opcode timing tricks to push the speccy to its limits.

GCSE computer studies was hilariously trivial for those of us already into computers. For my project I massively over did it, I wrote a z80 disassembler in speccy basic, a friend of mine from a richer family, had a BBC, he wrote a circuit board designer in 6502 assembly for his project. He now works at Intel. The only other person to get an A in the class wrote a calculator in BBC basic, so we had a baseline of all that was needed :)

On the PCs school I wrote a z80 emulator in 8086 assembly during my A-levels. It was years later during my degree did I finally cease using a z80 based system at my primary machine, getting a £20 80186 RM nimbus 2nd hand from my uni (I bypassed the 16-bit gaming machines). I finally saved up enough to get a 386 to run Linux (another reason for career success). Of course I played with all the x11 speccy emulators.

As I now approach retirement (I may have mentioned my career was quite successful in computing), I honestly plan to go back to coding for the speccy on unfinished projects from my teens!

Long live the z80 even if they stop manufacturing.

[Anonymal coward]

Luxury!

I was introduced to IT when (having screwed up my A-levels) went to work in a bank in the mid 70's. Eventually, I was trusted to key in the day's cheques and credits on the 'computer' AKA a huge Burroughs teletype device (maybe a TC500). Then there was that special day that we had to deal with the processing centre at Wythenshaw using reels of paper tape called 'algorithms'. After the trusty TC500 had hammered out the eagle logo in text graphics, I had to know more and went to uni to learn about computers. Only then, in the dying days of the 70's, did I discover the new-fangled microprocessors....Luxury! Yah, you try an tell the young people of today that, and they won't believe you...

[Paul Cooper]

From 1983 to 1987 I spent many hours writing, maintaining and adapting data logging software written for a Z80 S100 bus single-card computer. I wrote in assembler, and my development system was an Osborne 1 with Wordstar for an editor and ASM and LINK to create executables. The executable was then burnt onto an EPROM for testing and deployment. My development cycles was hours long!

Main lesson was how much you could do with so little; I am left with an abiding dislike of bloated software! My code fitted into a 2k EPROM without any difficulty, and I understood how every part of it worked; no libraries, and given the nature of the editor, I needed to keep a lot of code in my head as well as on listings.

[Bonzo_red]

Is the Zed 80 the British cousin of the Zee 80?

[ocelot]

I remember it being said when the Z80 came out, that people would still be writing new code for it in 50 years time.

And so it has become.

I was able to do software development on a Research Machines 380Z that the school bought, and hardware and software development on a friend's Z80 powered NASCOM 2 back in 1978.

But for various reasons, my own hardware was 6800 then 6809 based then I bought a BBC Micro with a 6502, then ended up as an ARM assembly language programmer . So the Z80 gave me my start.

[Chris Walsh]

It was the Sinclair ZX80, not the ZX81 that was available in kit form.

[MrRimmerSIR!]

https://www.computinghistory.org.uk/det/21611/Sinclair-ZX81-Kit