H8-7 Breadboard I/O address decoding

46 views
Skip to first unread message

Terry Smedley

unread,
Sep 7, 2021, 11:45:58 PM9/7/21
to SEBHC
I've been working on a TMS9918 prototype board, and decided to use the H8-7 breadboard (thanks, Dan!) to temporarily handle the bus interface and address decoding. 

Heath was clever with the address decoding circuitry on this board - they decoded an 8-bit address using only the high byte of the address (A8-A15) so that (at least for an 8080) the board could be used for either memory or I/O addresses.  The 8080 duplicates the 8-bit I/O address on the low byte and high byte of the 16-bit address bus, so for an 8080 it doesn't matter which byte you decode for I/O addressing.  

But, the Z80 doesn't do that - the high byte of the address bus on a Z80 I/O instruction contains the contents of one of the registers (which register appears on the high byte depends upon the particular I/O instruction).  To force compatibility, Heath and Trionyx (and perhaps others) designed their Z80 CPU boards to look like an 8080 by forcing the lower 8-bits onto the upper 8-bits during an I/O cycle.  Newer "pure Z80" boards, like the v4 CPU, don't do that - the address bus contains whatever the Z80 puts there.

Which means the H8-7 board can't properly decode I/O addresses with a newer CPU board since it ends up decoding a register value rather than an I/O address.  There are several ways to fix this.  I don't think I will ever need memory address decoding on the breadboard, so I just rewired the address decoders to look at A0-A7 instead of A8-A15.  If you wanted the breadboard to work for both memory and I/O addresses, the fix would be a little more involved.  You can also just use the unbuffered connections to A0-A7 that are on the board, and do you own decoding with a GAL or other circuitry.  The H8 bus should be able to tolerate one more TTL load on the unbuffered address lines.  

I am not aware of any other board for the H8 that attempts to decode I/O addresses using the high byte of the address bus, so this is probably a problem only for the H8-7 breadboard.

Terry


Norberto Collado

unread,
Sep 8, 2021, 12:17:10 AM9/8/21
to se...@googlegroups.com

You are 100% correct. The Heath Z80 boards & Trionyx Z8 will copy by HW the A7-A0 to A15-A8 to do as the 8080A does. Les’s Z80 board that I support do not do that. All the Z80 board that I support do follow the Z80 design.

 

On Heath Z80 board, U10 will copy U16 A7-A0 to A15-A8 lines. All done in HW.

 

Diagram, schematic

Description automatically generated

--
You received this message because you are subscribed to the Google Groups "SEBHC" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sebhc+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/e184bf48-16d7-4ca3-a2a8-09baa5ca9f18n%40googlegroups.com.

Glenn Roberts

unread,
Sep 8, 2021, 5:48:55 AM9/8/21
to se...@googlegroups.com
Thanks for the tutorial. I never understood this ‘til now. 

The h-8-7 is a nice board. Last summer I did my “rescue” from Ed Aumiller in Virginia. He used to develop boards for the H8 by starting with a bare board (he bought them from Heath spare parts dept) and literally “hacking” the board - cutting out a rectangular shape to insert a wire wrap board, then using the decoding logic on the H-8-7 to interface to the bus. I have some of his leftovers from these experiments.

Sent from my iPad

On Sep 7, 2021, at 11:46 PM, Terry Smedley <terry....@gmail.com> wrote:

I've been working on a TMS9918 prototype board, and decided to use the H8-7 breadboard (thanks, Dan!) to temporarily handle the bus interface and address decoding. 

Paul Birkel

unread,
Sep 8, 2021, 6:18:12 AM9/8/21
to se...@googlegroups.com, Glenn Roberts

Glenn:  Would you please post a few photos?  How did he stabilize the insert?

 

From: se...@googlegroups.com [mailto:se...@googlegroups.com] On Behalf Of Glenn Roberts
Sent: Wednesday, September 08, 2021 5:49 AM
To: se...@googlegroups.com
Subject: Re: [sebhc] H8-7 Breadboard I/O address decoding

 

Thanks for the tutorial. I never understood this ‘til now. 

 

The h-8-7 is a nice board. Last summer I did my “rescue” from Ed Aumiller in Virginia. He used to develop boards for the H8 by starting with a bare board (he bought them from Heath spare parts dept) and literally “hacking” the board - cutting out a rectangular shape to insert a wire wrap board, then using the decoding logic on the H-8-7 to interface to the bus. I have some of his leftovers from these experiments.

--

You received this message because you are subscribed to the Google Groups "SEBHC" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sebhc+un...@googlegroups.com.

Les Bird

unread,
Sep 8, 2021, 7:42:41 AM9/8/21
to SEBHC
Oh, when I made the original Z80 board design I took the address decoding logic from the 8080 board schematic. I had no idea there was a difference. Good to know for the future.

Les

Glenn Roberts

unread,
Sep 8, 2021, 10:00:59 AM9/8/21
to Paul Birkel, se...@googlegroups.com

Paul Birkel

unread,
Sep 8, 2021, 10:46:21 AM9/8/21
to Glenn Roberts, se...@googlegroups.com

Ahh, THX.  They’re not coplanar; instead simply a big cutout for the field of WW pins and then attachment points at the overlapped edges.

 

How does that work with normal interboard-spacing?  Does one have to leave the next slot open?

Glenn Roberts

unread,
Sep 8, 2021, 11:28:02 AM9/8/21
to Paul Birkel, se...@googlegroups.com

By my measurement there’s about ½” (~12mm) from the back of a board to the top of the chips on the board behind it.  with the approach Ed used the wire wrap pins protrude about 1cm (10mm) behind the board so they would clear the top of the chips in the board behind, but if anything else was sticking up higher there would be an interference issue.

Norberto Collado

unread,
Sep 8, 2021, 11:33:30 AM9/8/21
to se...@googlegroups.com, Glenn Roberts

I’m talking the H8-Wire-wrap board and adding the point breadboard to avoid wire-wrap sockets as they are very expensive. It is in my To-Do-List. I think this makes more sense.

 

Background pattern

Description automatically generated

 

Norby

Steven Feinsmith

unread,
Sep 8, 2021, 11:53:32 AM9/8/21
to se...@googlegroups.com
Yes, I used a wire wrapping method on my past projects. The problem was that the company that provided the materials had pumped into the unreasonable cost. I had to give up and use a breadboard for the last 10 years. Recently, I saw several tools from eBay tried to sell at an enormous cost. For example, the wire wrap strip unwrap tool for AWG 30 wire is selling at eBay for $20.89 apiece. I bought the same one in 1978 for only four bucks. I still have it in my tool cabinet. I cannot imagine how they create something so expensive that no one is using it nowadays.

I gave up with the wire wrap method and am going to sell off items soon. The only thing I keep is some wires for bypass purposes.

Steven


Dave McGuire

unread,
Sep 8, 2021, 12:02:12 PM9/8/21
to se...@googlegroups.com
On 9/8/21 11:52 AM, Steven Feinsmith wrote:
> Yes, I used a wire wrapping method on my past projects. The problem was
> that the company that provided the materials had pumped into the
> unreasonable cost. I had to give up and use a breadboard for the last 10
> years. Recently, I saw several tools from eBay tried to sell at an
> enormous cost. For example, the wire wrap strip unwrap tool for AWG 30
> wire is selling at eBay for $20.89 apiece. I bought the same one in 1978
> for only four bucks. I still have it in my tool cabinet. I cannot
> imagine how they create something so expensive that no one is using it
> nowadays.

Lots of people use wire wrap today. It's mostly hobbyists now, but
don't underestimate the number of hobbyists out there.

Further, smaller production volumes always mean higher prices, so
what you're seeing makes sense.

-Dave

--
Dave McGuire, AK4HZ
New Kensington, PA

Les Bird

unread,
Sep 8, 2021, 12:17:05 PM9/8/21
to SEBHC
Oh, and Terry, I'm very interested in this TMS9918 prototype you're working on. Is this to replace the HA-8-3 Color Graphics Card? Are you also planning on adding the AY-3-8910 sound chip as well?

Les
Message has been deleted

Terry Smedley

unread,
Sep 9, 2021, 1:31:41 AM9/9/21
to se...@googlegroups.com
Les,

Ultimately, yes.  We found a TMS9918 circuit design for the RC2014 Z80 computer that used the static RAM scheme that Norberto suggested.  It looked like we could glue that on to the H8 bus without too much difficulty.  So this first step is just to shake down the static RAM and understand how that would work in a HA-8-3 clone.

I do have some AY-3-8910 chips coming that will be part of this experiment.  

I am looking forward to seeing your games, Les.

Terry


--
You received this message because you are subscribed to a topic in the Google Groups "SEBHC" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/sebhc/1SK4lwTlCOo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to sebhc+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/92512107-e67e-46ce-8bf5-d5620b05ecccn%40googlegroups.com.

Les Bird

unread,
Sep 9, 2021, 9:30:21 AM9/9/21
to SEBHC
Nice find Terry. Looks like such a simple design, hope it translates to H8 well.

Definitely looking forward to the results of this project.

Les
Reply all
Reply to author
Forward
0 new messages