Steve WetherillChapter 2 of my memoir is up: "The Locked Room."
A computer no one could touch, a lift that never stopped, and the Z80 chip that turned out to be more important than anyone told me.
https://stevewetherill.substack.com/p/chapter-2-the-locked-room
Quazar@stevewetherill Reading through your chapters this morning and enjoying them.
In 1995 the Z80 was still being taught as a module at St. Andrews University where I was doing an Electronics + Physics joint degree. I was told just to skip all the lab work for the module after correcting the lecturer that OUT actually uses all 16 bits of the address bus, not just the lower 8 as per the course notes.
(I'd finally taught myself Z80 assembly the year before on the SAM Coupe computer. Despite getting a 48K Speccy in 1983 I always had the belief that machine code would be too hard to learn, so spent my years as a kid tinkering with BASIC and various pieces of game maker software.)
@quazarsamcoupe the Z80 was and is quite a complicated CPU, certainly compared to its peers. In fact, I grabbed the Zilog IDE for their EZ80 and related chips about 20 years or so ago, thinking it'd be interesting to test the behavior of my Z80 emulator by running the same code in theirs. Well (IIRC), they got the DAA flags behavior wrong. Even Zilog can get it wrong!
@stevewetherill I've got a Zilog eZ80 Dev Kit sitting on my shelf here - in a past job back in 2004/5 I chose the 50MHz Acclaim MPU part for some industrial control designs, and have used it since in my own retro endeavours.
Zilog's initial silicon for the eZ80 parts was quite badly bugged, with a slow trickle of errata datasheets appearing in those early years, although not all the bugs could be worked around via software tricks. Revised silicon, under the 'AcclaimPlus' name, fixed the majority of problems.