(I couldn’t let you go… v0.2 is born)

The first real working circuit was built on a breadboard around Christmas 2019. Not everything was working, e.g. the LCD doesn’t work yet. I think I know where to look (waiting for command to be processed by LCD controller), but I had made UART my priority.
Also I had added an additional EPROM to test the access to ROM in higher banks.

This circuit was then rebuilt in Kicad.
I should warn you, the image below still has the errors I made!
The v0.3 version of the schematics should have these resolved. (And might also introduce new ones)

WARNING: Schematics of v0.2 have some errors!

I did some more debugging in the last few months, which resulted in a PCB, also designed in Kicad, with the dimensions of 300mm x 200mm. (This should fit a 1U stripped switch I have).

The gerbers for this project were send to JLCPCB to produce the boards. (I moved the GAL22V10s to a daughterboard, as I intend to move the chip select and bank select logic to a CPLD in the, near, future).

I received the board within about a week (large boards, with red solder mask!) and I started to solder the components. Nearly all ICs are socketed, but the AS6C1008. I didn’t have a 32-pin wide DIP, so that one is soldered directly to the board.

I did rewrite/copy the monitor of SBCOS, by Daryl Rictor, as a starting point for my tests. Somehow I could get my SBC to boot. I had made an error on the GAL22v10 daughterboard. This was easily solved by cutting a trace and soldering two wires. Now I had a working SBC, sort of….

Looked like I couldn’t return from a subroutine. Was there something wrong with my SRAM? Nope, I had connected the second chip select (the one which is active HIGH!) to ground by mistake. So I had to cut a pin. I soldered a 1k resistor between the pin and the Vcc pin and I had a working system.

Running at 250KHz (as I use a 1MHz oscillator and divide the frequency by 4, by using both flipflops of the ‘HCT74)