No need to worry about your sanity. It's the lack of an oscilloscope that would worry me.

I doubt a properly functioning power supply could cause a loader malfunction, unless the thing is not really operating properly. Now, you probably should be really careful about the capacitors on the Vcc-to-GND portion of your circuit. There should be some (three or four) small ones on the order of 0.01 microfarads, scattered about, one at the power entry, one at the MCU Vcc pin, one at the HC125's Vcc pin, and one at the MAX232's Vcc pin, as well as the required 5 cap's to make the Max232 work. (Two for the charge pump, two for the output supplies V= and V-, and one to decouple the MAX 232 from the rest of Vcc) A between-22-and-50-ohm resistor in series with that device's Vcc connection wouldn't be a bad idea, either. The big bypass cap for the MAX232 should be on the MAX232 side, BTW. You should also have about a 10 uF Tantalum cap at the power entry.

15 MHz isn't on the table on page 134 of the "Ultra-High-Speed Flash Microcontroller User’s Guide", but, if you're getting an intelligible message from the MCU, then it must be close enough to something that works with their autobaud scheme. I'd try to get something that is supposed to work, though, if it were up to me.

I'm going to wire up a programming circuit in the morning (I'm in GMT-6) which will be approaching close-of-business for you, I suppose. I'll let you know how that turns out.

If you're really determined to make this work, you'd best stick to what's in their recommendations.

I think you were on the right track leaving out the "reset circuit". Since it's driven by the PC, the R and C are irrelevant.

How long does it take to produce those prompts you're seeing? Is that a continuous burst, or is it one every 15 seconds ... or???

RE