Tandy CoCo 3
I have three CoCo 3s. One has a Hitachi 6309 in place of the stock 68B09E and was always a little flaky. The other two have been collecting dust: a clean one that’s still in the box, with the foam, the booklets and a Radio Shack flyer, and an old unmodified one. The old one is getting 512K and a 6309. The boxed one stays boxed for now. Maybe I’ll upgrade it, maybe not.

Mostly the CoCo has been an appliance for me. Boot NitrOS-9, poke around, turn it off. These notes are me trying to make it something I actually write code for, which on this machine means BASIC09 or an assembler, since there’s no Turbo C or Merlin to lean on.
If you have something interesting of equal or greater value, reach out.
Current state
As of 2026-09-25: The 6309 machine boots and survives tapping. An undocumented 32 MHz clock mod came out, the oscillator went back to what the schematic calls for, and the GIME socket got cleaned. It hasn’t had a long burn-in yet. BASIC09 runs. The native asm assembles by hand, but make does nothing. The Speech/Sound Cartridge is still untested on the repaired machine.
Architecture
Nothing yet.
Hardware
| Item | Notes |
|---|---|
| CoCo 3 with 6309 | NTSC, HD63C09E. The one that had the clock mod. |
| CoCo 3, old | Unmodified. Next up for 512K and a 6309. |
| CoCo 3, boxed | Clean, still in the box with foam, booklets and flyer |
| NitrOS-9 | Runs the CoCo 3 in its 1.79 MHz double-speed mode |
| CoCo SDC | Storage. The dev loop through it is slower than I hoped. |
| RS-232 Pak | Takes the only cartridge slot |
| Multi-Pak | Four slots, bulky, rarely used |
| Dual floppy drives + controller | Matching set |
| Speech/Sound Cartridge (26-3144A) | Has never worked on the CoCo 3 |
The single cartridge slot is the real bottleneck.
Gateware and firmware
Nothing yet. I’ll probably get a prototype board for the cartridge slot, throw an FPGA on the bus and tinker, since that’s what I end up doing with everything.
Open problems
- Burn-in on the repaired machine, with a loop that exercises video and disk, and a tap test once it’s warm.
- 512K and a 6309 for the old unmodified machine.
makeunder NitrOS-9 stays silent. Getting it working normally is next.- The Speech/Sound Cartridge has never worked, but the machine flaked out every time I tried it. It gets another go now that the machine is stable.
- Ideas I haven’t started: DriveWire over the bit-banger port to free up the slot, NitrOS-9 drivers, an FPGA cartridge like Byte Hamr on the IIe, and using a real 6809 as ground truth for Simple6809.