Close

Design Complete

A project log for Novasaur CP/M TTL Retrocomputer

Retrocomputer built from TTL logic running CP/M with no CPU or ALU

alastair-hewittAlastair Hewitt 04/02/2019 at 18:185 Comments

However... complete doesn't mean finished. There will be continued iteration and refinement and then unknown problems to resolve during the prototype phase. After 37 days I have arrived at a design that should work and meet the current specification.

So the final numbers are in with a total of 54 chips. These break down as follows:

The gates still need to be shuffled around a bit and I may be able to consolidate with one less TTL chip. It's just as possible that this process could push the count up by one, so the TTL chip count is plus/minus one at this point.

I've also been working on some PCB layout ideas. I like the one-board approach of the Gigatron and it should be possible to cram the YATAC on to a board of similar size. I want to use a COTS enclosure though and this could become quite large and costly with a single PCB.

What became more apparent with a completed design is the natural demarcation between the GPU and CPU. The GPU requires a similar number of chips to the CPU and it seems a logical way to split the design to layout two PCBs.

The current plan is to place all the GPU logic, RAM, analog circuitry, clock generator, and connectors on one PCB. The CPU logic and ROM would be on a second PCB that mounts on top of the GPU using two headers. Each board would hold 27 chips in a target area of 18x12 cm. The entire assembly would then fit in a Polycase ZN-40.

Next up is a BOM and then placing an order for the prototype.

Discussions

Alastair Hewitt wrote 04/04/2019 at 13:25 point

@roelh I'll probably redraw what I have this weekend and I can scan it in to a PDF so it will be actually readable. This will then be drawn up in CAD and edited as I move forward.

  Are you sure? yes | no

roelh wrote 04/04/2019 at 17:30 point

Your nice pencil drawing is typical for the first stages in a design, where inspiration and creativity find their way to the paper....  but an explanation would certainly be needed for us HAD visitors to understand the design. Writing down how the circuit operates will also help yourself, you might find errors or improvements during that process. A description of the intended instruction set, however short, would also help.

  Are you sure? yes | no

Alastair Hewitt wrote 04/04/2019 at 23:49 point

I'll start to invest some time in explaining things now the velocity of change has dropped. The instruction set is shown in the project background photo... but yes, it needs explanation!

  Are you sure? yes | no

Alastair Hewitt wrote 04/04/2019 at 00:31 point

Hmmm... It's kind of a mess at this point. The schematic is strewn across multiple pieces of paper and still under some revision. I'll be starting on a proper CAD drawing in about a week and refine it as I validate the prototype. I pieced together most of the current drawings and took a photo. Be warned! It is a mess and may be hard to read. Hopefully I'll have a real drawing by the end of the month.

  Are you sure? yes | no

roelh wrote 04/03/2019 at 20:11 point

Hi Alastair !  Can you show the current schematic  ?  And thanks for following and liking  #RISC Relay CPU !

  Are you sure? yes | no