"I built my first ISP and prototyping board/circuit or the 420 this past weekend)"

That chip was my first ISP part too. I got a sample and tried it out about two years ago. My opinion of the part is pretty low now, though. The errata sheet is pretty long and the documentation is crappy. The datasheet omitted the pinout for port 0, plus it had the XTAL pins backwards for the TQFP version. I think they updated the sheet in the past six months, though. The read-modify-write glitch screwed up one of my projects. The glitch caused an interrupt to trigger every once in a while. What I don't understand, though, is that I rewrote the code in assembly and avoided the R-M-W instructions and the glitch was still there!?! I will not use it again. The Cygnal parts are superior in EVERY WAY, so I will use them in the future.


"and all the examples I have tracked down seem to suggest it as a safety feature to prevent overloading the pin (?)"

Sounds like bunk. And it results poor operation. The output structure cannot drive a typical transistor into saturation. The 1.5mA of output is only during 0-1 transition. Even if it seems to work on your bench, that doesn't mean it can turn on the transistor under all conditions, and with marginal micros and marginal transistors.

If you want an external discrete, check out logic FETs (nds355 from digikey, for example). Open collector buffers/inverters also work, especially for entire ports. Good luck on the hardware design. It is definitely an interesting hobby. So many details to worry about.

- Lee