Ok, here's my next thing. I have this arduino on a breadboard experiment. Used a sketch or two to program the ATmega328. It works. That is, it works when I pull the chip from my Diavolino and run it in there. The blink sketch is running on it. So, I put it in a breadboard with a pull-up resistor on the reset pin, and basically the breadboard looks like the one shown here
, but with no other board attached. I have an LED on what I believe to be pin 13; that is, from the picture, fifth pin from the right in the top row. If I apply 5v directly to the wire that's connected here, the LED lights. The pull-up resistor is probably working ok, but the behavior is the same even if I directly wire reset to 5v.
This leads me to think that something is wrong with my clock. I have wires attached to the clock pins as shown, to plates in the breadboard, on one side of which rests each of the two crystal pins, one per wire. On the opposite side are load capacitors. Somewhat higher than people generally use because that's what I have, 33pf, each of which are connected directly to ground on the other side. Basically it's as shown in the picture, but the load capacitors are not the usual 22pf ones. It's my understanding that a larger capacitor should just slow the clock down some, but I don't know exactly how much breathing room I have here. Is there an obvious problem with this setup? Is there anything good to check if not?