Atomic Phoenix - Brown Gas generator project

Glad to hear the progress SysEngineer.

GlobalBEM 2013 Conference Presentation by Moray B. King

youtu.be/GhZ1reCoyrY

The CPU board is a relatively simple 2 layers PCB with very straight traces, so it would not be too difficult to derive the schematic: this would simplify a lot the work and will give some clues for understanding the code. To this end it would suffice a good quality high resolution picture of both sides of the board.
My guess is that the program control loop runs on polling for most of the external resources, due to non critical timings, and perhaps employs the external interrupt for the ADC only. If we can get the schematic we can figure out the peripheral addresses and so link the code to its activity and purpose. Without the schematic this task is quite hard, IMO.

Quoting Mats002:

Ok, so we have a measure of electrolyzer current
going into the ADC going to the microprocessor as
8 bit(?) resolution. What can output be? A control
signal to the DC electrolyzer? DA conversion can
be constructed very simple, no need for a chip.

Ok. So speculation just slows us down until we can get a good schematic down. Current is key measure, you have to have the overall picture before understanding and interpreting the details.

Thanks for the LS28 note -- for some reason, I was not able to find it is a 74LS family list.

The ADC is an 8-channel ADC. I know that the panel board gets the current (voltage across a shunt) and voltage signals from the heavy current electrolyzer. However, they are both directly displayed by a meter chip tied to the panel displays. So far, just the current signal (conditioned by op-amps) goes to the ADC. The voltage signal seems to be ignored.

I have sent photos to Greenyer till I can figure out how to post pics here.

The 74LS28 is a quad 2 input positive NOR bufers. Can send phoro of document on the part. How?

I based the idea that there is only one interrupt on the decoded assembly listing. It is now apparent that I need to trace the board more thoroughly to verify that. It also seems to use memory-mapped IO (based on addresses that I se dis-assembled).

I have now powered up the system and get some activity, but except for reset, do not see any response to pushbuttons on the panel.

I don't see any comments here yet. I saw some of the developer's video on Radare, yes steep learning curve.

Saw mention of only a single interrupt.
This indicates to me the developer established his own external interrupt architecture. There will be some logic to capture which event occurred and some value associated with it.
When the circuit is chased down to some level of a schematic, let me know if you need help deciphering the logic. I designed a number of systems using TTL logic back in that era.