[Intel-devsys] Tauntek/Grieb Z80 and 8080 ICEs (Herb Johnson)
hjohnson at retrotechnology.com
Wed Aug 7 10:48:13 MDT 2019
Thanks for joining the discussion, Gary; and confirming the state of
your previous project and available boards. Again, your post in
retrobrew was also informative. Here's my thoughts for what they are worth.
First: a microprocessor ICE is useful and productive. I've used Bob's
product also, it was a great and useful thing to have been produced at
that time. I was an early user, and people through the years have asked
me about the product due to my Web page about it.
I read the evidence as given. That evidence makes it clear to me, that
this particular implementation of a microprocessor ICE, is not
productive to replicate going forward. (Gary's stock and further
interest are his call of course.) I'll take down or modify my Web page
on the subject accordingly.
What remains, is the principle of using one programmable device to
intercept instructions and signals of a 8-bit processor; and then take
actions and report them by serial link; and also the converse of
serial-link actions to take actions upon the 8-bit processor. ICE is an
old concept, what's recent is to implement it in the physical space of
One apparent design challenge, is to use an intercepting programmed
device that can manage 5-volt logic, versus the predominance of 3.3V
signaling today. Beyond that: it's "designer's choice" for another
programmed-logic device. I don't know what those choices are.
The alternative, as I found and posted here, is to completely simulate
the 8-bit processor at real-time speeds (several megahertz) with a
faster modern processor. I posted that someone apparently used an ARM
and other logic to ICE-emulate a Z80, and also potentially other 8-bit
processors, and even features of additional hardware. It's not clear, if
that particular result or its design will become available.
Point being: again it's "designer's choice" if someone wants to build a
completely synthetic Z80, 8085, 8086 processor ICE. (The 8080 has some
non-5-volt signaling to deal with.) ARM processors are fast, cheap, and
small. That choice offers "one ring to rule them all", theoretically.
There's another alternative, which is simple to create but limited in
results. I've done it myself: Lee Hart introduced it to me. The same
principle was used in the MITS Altair and IMSAI 8080 front-panel boards.
One disconnects the processor's data pins from the board-under-test; and
straps them as needed to force the processor to execute specific
single-byte instructions. All this takes is a processor chip to mangle
up, maybe a socket.
Less fancy description: force a NOP on the processor and watch as it
races through the address space after reset. One can force other
instructions; with simple logic one can force a pair of instructions by
using A0 etc. to select two rows of diodes/drivers. You get the idea.
It's crude but.. no CPLDs, no device programmers, no
search-ebay-for-misbranded-parts, no PC-board-layouts-sent-to-China.
It's a reasonable choice, because it's easy for an individual to DO, on
their bench. "A bird in hand, is worth two in the bush."
Those are my thoughts.
PS: Gary, my apologies for describing my look into
retrobrewcomputers.org as "down the rabbit hole". But that is my
experience as an outsider, who is not part of the conversations there,
and who occasionally looks to see what is available-for-purchase-today.
I appreciate, the site and related discussion sites, are not there for
that purpose, it and the participants have other purposes which are
commendable. Without elaborating: many people are able to create
quickly, microprocessor items for their own or other's interests; and
then move on to something else. I'm sorry if I'm impatient about that
fact of modern technology.
more regards, Herb
On 8/6/2019 9:44 PM, Gary Kaufman wrote:
> Herb -
> I did a run of the boards a few years ago and posted about the
> experience here:
> I have some of the boards and CPLD's available for both 8080A and Z80
> should anyone be interested.
> I fixed up the PCB to include a solder mask and silk screen
> Be aware that functional XC9536's are very difficult to source. The
> XC9536XL is a 3.3v part and wont work. Some of the available parts have
> very old date codes (ASJ) and in my experience wouldn't program properly
> with any version of the Xilinx software I could locate. All of the Ebay
> parts seem to be used, and remarked. In my experience less than 50%
> were good and many batches were entirely junk. I eventually purchased a
> fairly expensive ZIF socket so that I could program and test the parts
> prior to soldering onto a board!
> -= Gary
Herbert R. Johnson, New Jersey in the USA
http://www.retrotechnology.com OR .net
preserve, recover, restore 1970's computing
email: hjohnson AT retrotechnology DOT com
or try later herbjohnson AT retrotechnology DOT info
More information about the Intel-devsys