RIP Z80

[tchicago]

Zilog announced end of life of Z80, but it is still being manufactured as of now! 

https://www.mouser.com/PCN/Littelfuse_PCN_Z84C00.pdf
--- "Last time buy orders will be accepted until June 14th, 2024."

Maybe I should order a couple of them with a cool last production date code of 2024 as the memorabilia items.

Zilog brand apparently belongs to Littelfuse these days.

[brucehoult]

The votes are in. 6502 outlasts Z80 [1].

Z80 cores will still be in other chips -- it's just the DIP40 (etc?) bare CPU that is going away.

[1] To be fair that's 65C02, the NMOS chip went away a long time ago. And there will still be eZ80.

[SiliconWizard]

Incredible lifetime!

[lunix]

My very first commercial embedded project used a Z80. Then I fell in love with the 6809. Then the PIC.  A couple of decades after that first one, Zilog was pushing a Z80-based microcontroller, and had a giveaway for a design contest. I got a kit. Then I wrote to them that it had a dead short across the power supply, and they sent a replacement.  That one didn't work either.  And that was my last Zilog experience. I'm still sad about that. It was a better 8080, and those days are long in the past now.

[woofy]

Zilog could at least wait until March 2026, just to give the Z80 a 50 year lifetime.
Federico Faggin first came up with the idea for the Z80 in 1974, so I suppose it just about has its 50 years, though the first actual Z80's were born in 1976.
I cut my teeth on the Z80 and have since purchased many thousands for control products.

R.I.P Z80. 

[coppice]

The votes are in. 6502 outlasts Z80 [1].

Z80 cores will still be in other chips -- it's just the DIP40 (etc?) bare CPU that is going away.

[1] To be fair that's 65C02, the NMOS chip went away a long time ago. And there will still be eZ80.

Surely the Z80 still being made has as little in common with the original silicon as the current 65C02.

I played with a chip some years ago with a 1GHz Z80 core in it. I can't remember what the actual chip was, but cranking a Z80 up to 1GHz seemed amusing enough for me to have a short tinker with it.

[RoGeorge]

My first Z80 design was a home computer, compatible with both ZX Spectrum and CP/M.  That was happening sometime during the late 80's.  https://hackaday.io/project/1411-xor-hobby-a-vintage-z80-computer-prototype

[tchicago]

My first Z80 design was a home computer, compatible with both ZX Spectrum and CP/M.  That was happening sometime during the late 80's.  https://hackaday.io/project/1411-xor-hobby-a-vintage-z80-computer-prototype

Oh no, you breadboarded it! Must have been an immense amount of work.
My Spectrum was the second homebuilt computer - using a common Lviv variant PCB. It looked just like this one https://speccy.info/%D0%A4%D0%B0%D0%B9%D0%BB:Lvov1.jpg

But in general, the Spectrum was just for games, and not serious learning of computer science. For the latter, I kept using my first homebuilt computer based on i8080 (https://en.wikipedia.org/wiki/Radio-86RK). It had those big serious interface chips like i8257, i8275, and i8255 which were fun to learn and program.

[RoGeorge]

Some games for the ZX Spectrum were incredibly good for an 8-bit 3.5MHz Z80 that was also bit-banging the sound, reading the keyboard, refreshing the DRAM, and updating the image for each frame, all at the same time.  I remember Knight Lore was a total shock back then.  It was looking amazingly good:



The next jaw-dropping game for me was to be many years later, around the year 2000 when Half-Life came out (for PC).



By that time Z80 was rather rare, the IBM-PC standard was everywhere, and Z80 was only used for automations and such.  Z80 has had so far an outstandingly long life, considering with how many microcontrollers it was competing over the last 30 years or so. 

[woofy]

DRAM refresh was built in, with a 7-bit counter output during the M1 cycles of an instruction read, but yeah, it was quite amazing what programmers could do with a few MHz in and 8-bit processor. 

[MathWizard]

Was/Is the Z80 copied a lot, by counterfeiter's? I wonder where all the ebay ones come from.

[nctnico]

That is a long lifetime indeed. There was a time where I knew most of the Z80's instruction set by opcode. Goodbeye old friend... 

[Howardlong]

DRAM refresh was built in, with a 7-bit counter output during the M1 cycles of an instruction read, but yeah, it was quite amazing what programmers could do with a few MHz in and 8-bit processor.

Yes, DRAM refresh was "built in" in that there was a counter and notification that it's safe to do a row refresh.

Not so great was that to get the timings into spec that the DRAM of the day required, a fair bit of fettling and empirical work with glue logic & RC networks was required to get the RAS/CAS and setup/hold timings to work within tolerances, particularly if you were doing a production run.

It became easier and cheaper to stick in static RAM as time proceeded, what with higher densities, substantially reduced glue logic, and dropping prices.

[woofy]

DRAM refresh was built in, with a 7-bit counter output during the M1 cycles of an instruction read, but yeah, it was quite amazing what programmers could do with a few MHz in and 8-bit processor.

Yes, DRAM refresh was "built in" in that there was a counter and notification that it's safe to do a row refresh.

Not so great was that to get the timings into spec that the DRAM of the day required, a fair bit of fettling and empirical work with glue logic & RC networks was required to get the RAS/CAS and setup/hold timings to work within tolerances, particularly if you were doing a production run.

It became easier and cheaper to stick in static RAM as time proceeded, what with higher densities, substantially reduced glue logic, and dropping prices.

Yeah, my first Z80 design used 8 x 4164 dram's. It never sold many, nobody ever needed that much ram on a controller in those days and subsequent revisions used 6116 and 6164 sram's.
That was my only time putting dram's on the Z80.

[iMo]

I got my first Z80 inside the ZX81 kit when it appeared, added a DIY 16kB module made of 8x4116. Then ZX Spectrum. Messed for while with U880D. Then switched to 68k and never went back to the Z80. It was a nice time, indeed.. 

[coppice]

DRAM refresh was built in, with a 7-bit counter output during the M1 cycles of an instruction read, but yeah, it was quite amazing what programmers could do with a few MHz in and 8-bit processor.

Yes, DRAM refresh was "built in" in that there was a counter and notification that it's safe to do a row refresh.

Not so great was that to get the timings into spec that the DRAM of the day required, a fair bit of fettling and empirical work with glue logic & RC networks was required to get the RAS/CAS and setup/hold timings to work within tolerances, particularly if you were doing a production run.

It became easier and cheaper to stick in static RAM as time proceeded, what with higher densities, substantially reduced glue logic, and dropping prices.

Controlling DRAM took quite a few chips in those days, yet the total logic involved was not that great. Zilog threw the refresh counter into the MPU, but left the rest out. I was always puzzled that nobody tried throwing the whole DRAM control thing into their MPU. Its not like getting the feature set right was a problem. From the earliest days practically all DRAMs were drop in replacements for each other, with a well defined path to the needs of future generations.

[brucehoult]

A couple of friend and I designed and built a wire-wrapped 6809 machine in 1983. We thought about 68000 but it was expensive and the 6809's 8x8->16 multiply was faster than the 68k's (wider) multiply, and the 8x8 suited the music synthesis one of us wanted to do.

We did DRAM refresh in software, with an interrupt causing a page of NOPs to be executed. Or that could be turned off when you knew the software would be doing sufficient memory references itself -- or just masked for a short while for a critical loop.

[HwAoRrDk]

Was/Is the Z80 copied a lot, by counterfeiter's? I wonder where all the ebay ones come from.

In period there were several clones by Warsaw Pact states - East Germany, Soviet Union, Romania.

It's doubtful that's what ebay counterfeiters are selling today. I believe they mostly are selling chips that aren't Z80s at all, or remarking older chips to appear newer.

[coppice]

A couple of friend and I designed and built a wire-wrapped 6809 machine in 1983. We thought about 68000 but it was expensive and the 6809's 8x8->16 multiply was faster than the 68k's (wider) multiply, and the 8x8 suited the music synthesis one of us wanted to do.

We did DRAM refresh in software, with an interrupt causing a page of NOPs to be executed. Or that could be turned off when you knew the software would be doing sufficient memory references itself -- or just masked for a short while for a critical loop.

A lot of machines performed DRAM refresh in software. I've also seen a DMA channel committed to refresh. You still have to do things like RAS/CAS muxing in hardware. The refresh counting is only a modest part of the DRAM support problem.

[SeanB]

The Zilog one might be stopping, but the Toshiba and Sharp units are likely to stay in production for a good number of years still, as they also have all the peripherals of the Z80 family inside the package, and thus you can make a complete SOC with them, as they all now have 32k of RAM and 32k of flash built in, though there are still a fair number made with older OTP Eprom memory, which is in a lot of industrial equipment, copiers and such, along with lots of other applications where they have a very long code base life, and are familiar with the design.

[GromBeestje]

They have the peripherals inside the package? Something like the Zilog's eZ80? That one stays in production. It's the classic chips that go out of production. I was wondering if there was any of the other manufacturers still make the classic chips?

[amyk]

I think almost everyone has used a device with a Z80 core at some point in their lives, even after the 8-bit minicomputers faded away. The https://en.wikipedia.org/wiki/S1_MP3_player and the Texas Instruments calculators are the most prominent examples I can think of. (GameBoy doesn't count, as it's not truly compatible.)

[wek]

The votes are in. 6502 outlasts Z80.

There are still plenty of 8051 made, most of them binary-, some pin-, and maybe still a few of them clock- compatible with the original (1980).

JW

[peter-h]

I started on the Z80 too. A great chip and really versatile in its time.

Today, most embedded stuff (except low end) is done in C and if the clock is fast enough the CPU architecture becomes almost irrelevant. And those who cut their teeth on the Z80 are mostly retired now.

And few of the original products will be selling today; I still nominally sell a Z180 based box designed in 1991 but the volume is very low. There are still GA avionics (like the JPI EDM700 family) which use the original Z80, EPROM, SRAM but their sales will be very low too, even in that extremely slowly innovating field.

How long a product can run is always a fun debate but the truly original DIP40 Z80 must have virtually ended. Even in the military sphere, things have moved on. The other day I saw the internals of the Javelin missile; a 30 year old design but still current on the Ukraine battlefield, which uses all kinds of expensive chips.

Littelfuse owning Zilog is the ultimate insult; a shit company making boring stuff owning a one-time real innovator.

[coppice]

Today, most embedded stuff (except low end) is done in C and if the clock is fast enough the CPU architecture becomes almost irrelevant.

If you try to sell an MCU to many large customers now they show no interest unless it has an ARM core, or perhaps increasingly a RiSC/V core. You can talk about a totally unsuitable chip for their needs, and they'll listen if is has an ARM core. You can talk about a great fit for their needs, and they are not listening simply because it doesn't have an ARM core. But the core is almost irrelevant, because almost all new MCU code is developed in C. The MCU's value is all about its unique qualities - interesting peripherals, special memory qualities, like error detection/correction or non-volatility, partitioning to meet regulatory requirements, etc.