westom wrote:
On Mar 14, 5:01 pm, mlawson wrote:
What I want to accomplish is to homebrew one. Many years ago I
wirewrapped a TRS Model 1 PC and had a blast. Yep, I'm an old fart! And
it's a nice little hobby. I'll probably re-configure it to support a
Z-80.

Now I am completely inspired. Somebody who actually wants to learn
something rather than shotgun computer boards and call himself
knowledgeable.

Unfortunately, the hardware configurations of the Z-80 and 8080 are
different. And the Z-80 had a slightly larger instruction set.
Closer to the Z-80 was the 8085. I also designed hardware and wrote
software for that computer chip.

Because I am inspired, I dug through those piles from history. I
found multibus manuals for the 80/10 (8080 microprocessor) and 80/24
(an 8085 micro).

8085 was closer to the Z-80 (including no -5 volts required). But
Intel never got that chip to work fully correct. However Intel also
realized that the 8085 was wasting time playing catchup with the
superior Z-80. So Intel jumped forward to promoting the
simultaneously developed 8086 - which became the basis of what?

Zilog never recovered by trying to promote Z-80. Eventually Zilog
became part of Exxon when Exxon was foolishly doing what business
school graduates recommend. Z-80 was the first of many companies that
would attempt to challenge Intel's greatest power - its abilities to
innovate.

One neat feature of the 80/xx series were peripheral ports that
could be changed simply by replacing chips in 16 pin sockets. For
example, I reconfigured one design using especially powerful drivers
from Silicon Solid State Scientific (if I remember) that were more
powerful and transient hardened AND that were pin for pin compatible
with TTL chips for those sockets.

If I remember correctly, those multibus cards required better
regulation on its power supply as well as a master reset provided by a
multibus one backplane.

Eprom - largest was a 2732 on 80/10 and 2764 on 80/24. But again,
for memories over twenty years ago. Eproms from that other computer
would not work. Code must first setup peripherals unique to the
Multibus card. For example, Intel used its own serial port chip.
Many computers used a better (and different) serial port chip. Setup
required different bytes in different registers. Not that I am trying
to discourage you. But those multibus cards require writing code and
burn those newly programmed Eproms.
Well, truth be told I already have a Z-80 design done that I plan on
using to control my model train set. I have a Z-80 assembler that runs
under Windows and I plan to take the compiled code (in a hex file) and
read the file into a Basic program for output to the parallel printer
port. That will connect to the Z-80 system directly into memory. The
printer port strobe signal will write the byte into memory and also
advance a series of binary counters to the next memory location. I am
using the BUSRQ control line to switch the address & data bus from the
Z-80 to this input method. Then, once the code has been transfered,
release the BUSRQ and let it rip.... I hope it works. I'm going to start
wiring it up this summer. Mike