Well, I printed out the datasheet last night and went over it. The calculator they have is pretty cool as well. I think I'm going to stick with the LM317's though because the NCP3066 is 92% efficient while the 317 circuit is around 88 - 90% efficient. I'm driving 7 led's per driver from 22.75 volts (after 317 drop) at 1.25 amps ~3.2 volts. The other reason is I wouldn't need to add on a separate led bank to do moonlight because I can dim the full range.
I figured I'd give a little bit of an update as well since I've been busy with the board the last few days.
the DS1307 RTC is now onboard so no need to buy one.
Also added in an eeprom chip to store values (in case of power outage the aquarium doesn't screw up).
Got the board almost routed so it can be done with 1oz copper instead of 4oz (led driver traces are giving me headaches)
Giving up one float sensor in favor of a salinity (conductivity) sensor. I need to order the parts and breadboard it still, but I want it more than the second float.
I have a question for those who are smarter than me. The PH/ORP/Salinity sensors all need a dual power supply. Currently I have been using a TLE2426 in a virtual ground circuit. It's a rail splitter that provides the reference. Last night I found a negative voltage regulator and my question is this: which is better? I'm assuming the negative regulator will give me a true dual power supply, but is that better than using a rail splitter? Any help would be appreciated since I don't have that firm of a grasp on opamps yet.