Alright guys, what did I fry?

Man, winter is here. Ground yourself.

Anyway, I have a working circuit with a DS1307 breakout board, a 7-digit serial display, and an Duemilanove. I'm using the Arduino Time Library for the clock.

Anyway, I fried something. I touched my Arduino near the power adapter socket and created a huge spark, and my board promptly reset. It still works fine, but for some reason after about 10 minutes or so the clock resets to 12:00.

Which is odd, because nowhere in my code do I set it to 12:00. Further, the RTC itself isin't even resetting--once I reset the Arduino, the right time displays for another 10 or so minutes. I've tried a 9v and through USB, same issue. Once it hits 12:00, it still counts every minute.

Im stumped. What happened?

Hard to say. Why not put some serialPrint statements in your code and monitor on a PC and see what its doing? Could be you need to reload the program, could be something is overheating and causing the arduino to start with a voltage drop. Do you have an o-scope or a multimeter or any debug tools? This little guy from ($89) has come in really handy for me for looking at signals when something isn't static enough to look at using a multimeter.

I don't know if these guys are around still they used to have a soundcard based oscilloscope that was okay for lower frequencies - like catching power cutting out.

I just realized that there are two big caps near where I touched. If one went bad, that could cause the voltage drop you describe after a certain period of time.

I do have an Oscilloscope and multimeters, et, but I'm not sure what it's SUPPOSED to look like in the first place.

The voltage on the caps should be steady (or maybe sitting at like 8.5V with some AC ripple if are powered from a wallwart for example). If you monitor them and see some anything funny going on, that would be a sign. Pull up the Deumilanove schematic from the hardware section. C6 & C7 are the big caps.

One is a steady 3.33v, the other continuously fluctuates between ~3.8v and 0 constantly. If its supposed to be steady, that's a bad sign...but I don't get how its related to my problem.

What is your power source? If you are seeing C7 go up & down like that I am surprised anything is running. If it appears to work outside of the occasional reset, I'd go back to adding serial prints or something to track what the code is doing.

If it is a big cap then it is an electrolytic. If you over voltaged it you could have put a hole in the dielectric. A small hole will heal up with the electrolyte action but if it is too big this will break down again and short out, punch another hole, which will heal. This is a bit like an oscillator. When it short's out you can get a reset or a corruption of the ram.