This was a long fought victory. Two pairs of loops were added, the first to fill the video memory with 64 columns of text repeating every 4 lines with the enumeration of code page 437. The second repeats the background colors as 8 columns and foreground colors and fonts on alternating rows.

There were various issues getting this working. One was a bug in the build script that generates the list of valid instructions. Most of the other issues related to stability. The supply voltage sags from one side of the breadboard to the other. The difference is over 0.3V, but the most stable range of operation falls within a narrow 0.1-0.2V range. Adjusting the supply to stabilize one side will destabilize the other. It also appears the NOR flash is aging each time new code is flashed. The access time is slowly increasing and the once stable 25MHz is now glitching (spot the glitch in the last sequence of 0-9 above).

There are a few more tests to run, but the value of developing and testing on the breadboard is rapidly diminishing. The PCB is starting to progress with a final component layout and routing strategy. The first revision will be a simple two-sided board to test the layout and power distribution. Further revisions will likely switch to 4-layers depending on actual stability.