Zilog8000

[Glenn Roberts]

Just watched Scott baker’s YouTube video on his z8000 board for the h8. And he built a PAM-8000 font panel monitor too! Great work Scott!

https://youtu.be/iS4Z_j-Y_NU

Sent from my iPad

[Scott Baker]

Thanks Glenn, I was just going to post that to the group this evening!

I learned a few things during the experience:

1) Take an extra few days when designing a complicated board before hitting the <submit> button. The piggyback bodge I had to implement I realized about 48 hours after jlcpcb started fabbing the board. By then it was too late and cheaper to just piggyback the change.

2) Implementing my own interrupt controller may well be superior to what Heath did, but it's going to complicate my life if I ever want to use the H37. Need to understand what they did and why they did it with the H37.

3) The front panel "Interrupt Enable" line didn't do what I thought it did. Or maybe it does, but I didn't appreciate how it relates to single stepping. I need to go back and read the theory of operation a few more times.

4) When using a CPU that's not compatible with Z80/8080, the ratio is going to be at least 10:1 of software development time to hardware development time. If I wanted to make the H17 work with this thing, it could easily take me 20 hours or more.

Writing the panel monitor, and seeing the display come to life for the first time was pretty fun.

Scott

--
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/9A0279F4-FA76-4C86-BB98-E7E8CD38C024%40gmail.com.

[Douglas Miller]

The key to the H37 interrupt "hack" is that they needed a way to wait for data to start to arrive, and for each byte to become available, but could not meet that timing with "conventional" instructions at 2MHz. So, they HLT and when the DRQ interrupt occurs they force an EI instruction instead of the normal RST instruction. This way the execution resumes fast enough to meet the timing requirements. But, it means they have to take over the normal H8 interrupt processing during the time that data is coming from, or going to, the floppy.

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/CA%2B8qVMO1rXpi%2Bua28MjiQB6AeLknevy7pRgaOQR6JUOiH3jXQg%40mail.gmail.com.

[norberto.collado koyado.com]

Great job Scott in getting the Z8000 working on the H8. For sure it is a very complicated design. I’m impressed and thank you!

 

Do you think an 8086 CPU design could be next as Trionyx did?

 

Norberto.

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/CA%2B8qVMO1rXpi%2Bua28MjiQB6AeLknevy7pRgaOQR6JUOiH3jXQg%40mail.gmail.com.

[smb...@gmail.com]

Thanks Douglas, I'm going to have to study the H37 board as I'd like to find a way to incorporate the DRQ into my revision of the Z8000 board. I don't really want to reproduce the H37 hack as-is if I can avoid it (since all the software needs to be rewritten anyway. for the Z8000...). I'm running the Z8000 at 6 MHz currently. I can run it up to 10 MHz, but ultimately it depends on what frequency the bus and peripherals can withstand.

Norberto, I'd like to know more about the Trionyx 8086 board. From a few pieces of literature I've come across in the past, they extended the bus to 16-bit. I've avoided doing that due to the complexity -- I'm assuming we could throw RAM onboard like we do with everything else these days and avoid bus modifications. What I'd really like to know is what software exists for the Trionyx 8086 board. If software exists for it already, then it's a lot more tempting than trying to write new software from scratch. The logical historical move would have been to release CP/M-86 for it, with the necessary drivers to support the Heathkit storage devices.

On the subject of the Z8000, I continued to improve the monitor this weekend. I now have it able to display and edit registers (all 16 general-purpose Z8000 registers, plus the PC and the FCW), and I can break and resume the program. It's a fun effort, but it's also a lot of Z8000 assembly code.

Scott

[norberto.collado koyado.com]

Scott,

 

I only have the following and no schematics or BIOS ROM’s…. Trionyx supported the 8088 and 8086 on same board. I think we should wait until we can get more documentation on such design.

 

 

 

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/f7bfc145-3402-44a8-b174-4fecfc36f0d8n%40googlegroups.com.

[Robert Litz]

I have a board with parts installed from Trionyx 8086 or 8088 board.  Never got it working. 

[Robert Litz]

Did you mean 8087 floating point coprocessor not 8088?

[Robert Litz]

I don’t have the Trionyx disks for the 8086/8088 board I do have an 8” set of CPM 86 disk maybe for Z100.  

On Sunday, January 22, 2023 at 3:17:10 PM UTC-6 Norby wrote:

[norberto.collado koyado.com]

Hello Robert,

 

Can I get/borrow such board along with any floppies to add live to the board that I have without any parts?

 

Norberto

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/253e4c69-2045-4efe-9576-7a9f041782c9n%40googlegroups.com.

[smb...@gmail.com]

Follow up on the Z8000 Board. My second revision arrived from Jlcpcb and I built it up today. Pictures is attached:

* Fixed the upper/lower data latches --  no more ugly piggyback bodge pcboard

* Added a socket for the Z37 cable. Have some jumpers to hook the DRQ and/or INT lines up to the Z8000's NMI or NVI interrupts. Utterly untested.

* Reworked the bus control lines (IOR, IOW, ...)
* Added an optional 5x2-pin serial header, in parallel with the 15x1-pin headers.

* Minor tweaks here and there

It's coming along well, though I'll need some time to write a Z37 driver before I can really call it ready for release. Until I use a "real" peripheral, i.e. something other than just the front panel, the bus interface is fairly unproven.

Scott

[norberto.collado koyado.com]

Scott,

 

Very nice work and very impressive design. Which OS are you planning to run from the H37?

 

Norberto

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/62d979db-3cc5-4765-9f8e-2df695949bf6n%40googlegroups.com.

[smb...@gmail.com]

I would probably just use the H37 for data storage. 

Right now there's enough on-board Flash (1 MB) to store CP/M-8K 1.1 and a plethora of utilities to go with it, and that's the only thing that will run (and perhaps will ever run) on this board. The floppies could be used to store programs. There's not necessarily any reason why I couldn't boot CP/M-8K from a floppy, but it doesn't seem like a high priority.

I've previously interfaced a WD37C65 and I've interfaced IDE/Compactflash to the Z8000. Writing a driver for the WD2797 shouldn't be too hard. The H17, on the other hand, would be painful.

Scott