Whilst it may work, it will almost certainly be stressing the 74LS04, which may result in it failing prematurely. I can see two possible solutions here, though...Oops! Yes they are fighting. But I think it might be OK and it does actually work.
Because the pico can sink considerably more current than the 74LS04 can source, the pico wins. This is the reset code - the output of the 74LS04 is effectively shorted to GND for 10ms. Do you think this will break anything?
The quick and dirty solution might be to lift the leg 4 of IC9, install a link between pin 13 & pin 4 on the socket of IC9, and make the Pico output open collector. R3 on the Atom will keep nRST high until either BREAK is pressed OR the Pico output is driven low. The other potential benefit of doing this is that IC25 (8255) will also be held in reset by the Pico; although it does mean that the CPU will be taken out of reset marginally ahead of IC25. I don't know if that might have an impact on any code that expects keys to be pressed on BREAK (eg shift or Ctrl).
The alternative would be to replace the 74LS04 with a small custom PCB. By using a smaller SMD version of the 74LS04, you could probably make the PCB fit the footprint of the existing through hole 74LS04. With the custom PCB, instead of using the 2nd inverter on the reset circuit, you would add a small transistor / MOSFET (with appropriate pull up) to the PCB, and use that to drive nRST. As with the previous option, the Pico output would need to be open collector, but in this case the Pico would not be able to reset IC25.
Statistics: Posted by KenLowe — Thu Mar 27, 2025 12:54 am
