My experience with Zilog was a less than happy one. The company I worked for was a big user of Zilog microcontrollers, one version of which developed SERIOUS reliability issues.

The issue was sufficiently significant that a group of us travelled to Zilog headquarters and had a couple of high level meetings which included then CEO Ed Sack.

The lawyers became involved but it was resolved with an out of court settlement, with the provision that the details could not be disclosed.
It is unfortunate I can’t share, because the technical aspects of the investigation are thoroughly fascinating. One of the major highlights of my whole career.

 

My experience with Zilog was a less than happy one. The company I worked for was a big user of Zilog microcontrollers, one version of which developed SERIOUS reliability issues.

The issue was sufficiently significant that a group of us travelled to Zilog headquarters and had a couple of high level meetings which included then CEO Ed Sack.

The lawyers became involved but it was resolved with an out of court settlement, with the provision that the details could not be disclosed.
It is unfortunate I can’t share, because the technical aspects of the investigation are thoroughly fascinating. One of the major highlights of my whole career.

Considering that it's likely more z80s (and derivatives) were shipped than any other 8 bit processor, I'd be surprised if there weren't a few hiccoughs along the way.

I actually enjoyed pretty much all of my history with Zilog.

 

Z80 board
View attachment 320428
View attachment 320429
If you've ever built a 8080 cpu, the Z80 was bacon wrapped with bacon on a thick bed of bacon.
View attachment 320427
8080 main and graphics processor boards with sram and eeprom board.

Back in the day I build several prototypes using those same boards, even some featuring the Z80.

Someone mentioned the TRS80. I designed a graphics card for that. Ran into a huge issue as RS messed up the memory control lines, and when you knew the address being accessed the time to hit the WAIT line had passed. This was due to converting the RD and WR lines to the "other standard" when they were combined with the MREQ and IORQ lines. My fix was to do a software work-around where any memory write to my card was followed by a read/verify cycle. This was the memory my video card used, mapped into a high area of RAM, with this memory switched between the computer and a video controller chip.

 

Someone mentioned the TRS80. I designed a graphics card for that.

In my early youth (9 to 13 yo), I spent a great deal of time customizing my TRS-80 Model 1. 48K RAM in the keyboard, upper/lower case, inverse video, interrupt heartbeat, A/D for joystick, etc. This was all inside the keyboard unit, which became a rats-nest of piggybacked ICs and wire-wrap wire. I still remember the trepidation I felt when I first punctured the warranty warning label (the cost of a Model I in those days was about $3K in today's dollars!)



I even wrote my own tape-based operating system. Since EDTASM required all source assembly code to be resident in memory, I had no room for comments in the code, even with 48K. I still have a copy that I review occasionally for fun, and wonder how I did that. The labels were limited to 6 characters, so even those are cryptic.

I had a full collection of the "...and Other Mysteries" books (for which I worked my butt off to afford!), plus a complete library of 80 Microcomputing/80 Micro magazines. I even submitted one of my programs to 80 Micro, for which I was paid $250 (a fortune for me at the time!). The manuscript for the article was written using a word processor that I also wrote.

Anyway, great fun, and it set me on the course of a profession in engineering and embedded programming. No amount of money could have purchased such an education.

It's a shame today's hardware is so highly integrated and intractable. Today's youth are missing out on a lot.

 

It's a shame today's hardware is so highly integrated and intractable. Today's youth are missing out on a lot.

Well... when any kid can go and buy a development board and using a USB cable they already own, using freebie compilers and libraries and can create a Wi-Fi controlled LED in a few hours or days, I'd say the bar can be as low as you want.

When I built my first Z80 computer the task was to drive a bunch of flip flops to drive relays. Think of a test stand. I was. I needed to program an EPROM (yes, one "E"), and I first built a programmer driven off a Sinclair computer. The code was hand assembled using pencil and paper and typed into the Sinclair. Then I could unplug and re-plug the EEPROM into my target. Lots of wire wrapping!

I do not miss those old days... well, I kinda do.

Now today, my last big project was a 4 channel RGBW LED controller. The "W" part was the trick driving this project. Based on an ESP32 dev board I could check out the hardware and initial code on a solderless breadboard (what 20 bucks now?), collect the parts off EBay for a song, and get PCBs using Kicad and JCLPCB for 30 bucks (and the PCB re-spin for the same). 3D printed the case in my dining room. Now I can ask Alexa to light my kitchen cabinets in any color plus a warm white wash.

 

Well... when any kid can go and buy a development board and using a USB cable they already own, using freebie compilers and libraries and can create a Wi-Fi controlled LED in a few hours or days, I'd say the bar can be as low as you want.

I did what I did because it was hard -- and rewarding -- not because it was easy.

I'd be bored out of my mind if I was a kid today.

 

I programmed a TRS80 Model 2 to perform two tasks. We used that model because it used 8” floppy disks. A co-worker wrote an RPG report on an IBM system that extracted certain orders from their ordering system and wrote them to an 8” floppy disk. The Model 2 read them, formatted them into the custom format required by their vendors and then emulated a TWX terminal to automatically place the orders. A couple of the vendors were Clorox and Starkist. I wrote the TRS code using assembly and BASIC on Pickles & Trout CP/M OS.

 

I recently built this for fun. It's a bare Z80 with an Arduino Nano providing all the Z80's hardware needs.


The GitHub link in the user guide only works for collaborators at the moment. I've not made it public yet.

 

I recently built this for fun. It's a bare Z80 with an Arduino Nano providing all the Z80's hardware needs.
View attachment 329888

The GitHub link in the user guide only works for collaborators at the moment. I've not made it public yet.

I did much the same thing a few years ago with a PIC18f46k22. I used SDCC for the Z80 coding. https://sdcc.sourceforge.net/doc/sdccman.pdf
https://forum.allaboutcircuits.com/threads/memio-emulator-for-z80.117003/post-916798

https://github.com/nsaspook/z80memio/tree/master

 

I did what I did because it was hard -- and rewarding -- not because it was easy.

This reminds me of a simiar quote:

We choose to go to the Moon in this decade and do the other things, not because they are easy, but because they are hard.

 

This reminds me of a simiar quote:

Me too. But I thought of my reply before I recalled JFK.

 

Should be ample Z80 variants available from China for decades to come. I have no trouble getting Z86 variants. Of course anything claimed new is likely not, so you have to have a means of testing the devices. That can be a challeng!

 

Is this a good news?

Open source Z80 clone seeks to help bring classic chip back from the dead.

 

Let it stay retired. It was a good run

 

Let it stay retired. It was a good run

Yeah! Everyone who's invested 50 years in code development should throw it all away and start over again, because, The Future!

It's going to be fun when you run the financial system and tell everyone to throw away their COBOL code.

 

Yeah! Everyone who's invested 50 years in code development should throw it all away and start over again, because, The Future!

It's going to be fun when you run the financial system and tell everyone to throw away their COBOL code.

Yes, The Future. It happens, like it or not. You don't throw it away, you use that past experience and time investment to move to something better (as in the ability to use modern tools to make it better), if it's needed. The Z80 (as a hardware abstraction/implementation of a compute machine) today is not needed but COBOL (as a software abstraction of business practices) still is.

Places that run COBOL seldom run it on 50 year old native hardware.

 

Yes, The Future. It happens, like it or not. You don't throw it away, you use that past experience and time investment to move to something better (as in the ability to use modern tools to make it better), if it's needed. The Z80 (as a hardware abstraction/implementation of a compute machine) today is not needed but COBOL (as a software abstraction of business practices) still is.

Places that run COBOL seldom run it on 50 year old native hardware.

Well the Z80 may be aged in its original form, but there are myriad offshoots in current silicon technology and likely even more versions used as IP blocks in ASIC devices or libraries.

 

Well the Z80 may be aged in its original form, but there are myriad offshoots in current silicon technology and likely even more versions used as IP blocks in ASIC devices or libraries.

Sure, people pick things because they like them, lots to like about the old guy, even if it wouldn't be a first choice for a new project.