Here is a dump of the ROM in a text-based format. I couldn't think of a good way to represent the 4-bit words in a normal binary format with the order being ambiguous.

Connecting a logic analyzer up to the ROM and triggering an "autoload" sequence, the processor reads ROM address 0, followed by ROM address 1, and then seems to lock up. I'm curious if the processor is attempting to store the 8-bit word into RAM for some reason? My RAM board is in very poor condition and I will need to devise a way to troubleshoot it. It'd also be helpful to have some of those control lines hooked up to the logic analyzer while it is happening.


I'm working on a disassembler that should let me shuffle the order of the 4-bit words around until I get something that looks sane.

‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On Wednesday, May 13, 2020 3:27 PM, Ken Thompson <ken@google.com> wrote:

can you send me the bits of the rom.
i will take a look.


On Tue, May 12, 2020 at 12:44 PM TangentDelta via TUHS <tuhs@minnie.tuhs.org> wrote:
Hello.

I have a pair of controller card cages out of Mergenthaler Linotron 202 photo-typesetting machines. Sadly the machines themselves were scrapped, and these card cages are all I was able to save.

The controllers use Computer Automation Naked Mini processors, which are relatively small 16-bit minicomputers designed for embedded control applications. I've been hacking on these for a few months now and have built up a system bus pinout diagram and several schematics. I haven't been able to find any technical information online in regards to the specific model of Naked Mini processor used in the 202, but I have found a trove of documents for other Naked Mini models on Bitsavers.

I pulled the 512x4bit "boot ROM" mentioned in the "Experience with the Mergenthaler Linotron 202 Phototypesetter, or, How We Spent Our Summer Vacation" paper and dumped it, but the resulting binary doesn't produce any sane-looking code when manually disassembled using the documents on Bitsavers and reference, no matter how I arrange the nybbles. The processor also does not appear to respect the control opcodes issued by the Computer Automation LSI series programming console that I obtained. This has led me to the hypothesis that this is not a stock "Naked Mini" or later "Naked Milli" processor, but something specific to Mergenthaler.

My goal is to get the processor to run my own code, and eventually design my own MaxiBus peripherals to use with it.

If anyone knows where I can look for more information in regards to the 202 and the Naked Mini processor, or has any stories of working on these machines, I would greatly appreciate it!

Thanks.