Friday, September 11, 2020

MAG-85 Update

MAG-85 Isn't Dead
MAG-85 circuit schematic

While I haven't posted much about my MAG-85 project or blogged about it lately, I have continued to work with it until late 2018, when all my activity shifted toward making a cross-country move from the Sierra Foothills of northern California to Florida. Electronics got roughly pushed aside for activities like packing boxes, painting, minor home repairs, and so on. As it happens, a lot has happened with the MAG-85 that never got posted simply because posting what I do multiplies the time involved in doing it by about 4x, so I wanted to get to a stable point, collect all the information, then post it in one big update.

That was probably a bad choice, versus the do a little, post a little approach I started with.

Presently, I am in my new home and still getting my electronics stuff set up again. I have found the boxes with my MAG-85 stuff and organized them, and my soldering iron is out and in use, though I will not really have an electronics desk till next week or so while I reorganize some things in my "laboratory". I broke down my breadboard MAG-85 that was assembled on my HP Logic Lab--it wouldn't have survived the move as-is, and most of my work has been on my vectorboard assembly lately anyway. I will be reassembling a solderless breadboard version of the MAG-85 again, though, and I'll document that as I do it. I have the breadboard ready to go, in fact.

The major item of work hanging over me is the software--the monitor program. I have all the necessary pieces--the keyboard routines, the I/O routines for the display, the various functions such as reading memory, writing to memory, displaying blocks of memory to the display. It needs to be assembled into a complete program, though, along with the current software to perform setup of the system at power-on. I hope to assemble this into a cohesive "operating system" as I am building up the breadboard. I will do this initially with the NOVRAM as the sole memory, but will expect to adapt that circuit to the use of both EPROM and RAM, so that users can have a system with a secure section of memory. Since I also want to accommodate a variety of memory devices, and later memory expansion beyond the 8K of the original MAG-85, this gets complex enough that I don't want to hold up the development of a monitor for the current design.

Beyond the desktop versions of the system--breadboard or vectorboard--I hope to create a battery-powered handheld version of the design. I have always wanted to have a small 8080-compatible handheld system that I can plink away at while sitting in my easy chair. My current vectorboard version was a step in this direction. It's bulky, but can be used handheld. I can make it smaller, though. The provisions I made for expansion, along with being overly conservative about its thermal design, made it larger and more unweildy than it had to be. So beyond getting a working monitor, I intend to refine the design for handheld use.

Once I'm at the point of having a PCB, I expect to release it as an open design, so that anyone can get a board made without relying on me to produce them. I will release the software under an open license as well, so that a community can run with it.

Among the design changes I am considering (after reaching the point of getting a full monitor and stable basic system) are a display with 4 lines rather than two, to allow a more sophisticated interface for both the machine language monitor, and for a possible Tiny BASIC or Forth addition to the base software.