OK, so here's a VERY detailed discription of the system... hopefully you can cross refference it to the pics to get a beter idea. or, you can just ignore this novel entirly, and enjoy my tank withthe knowledge that it works...period.

The sytem is made up of three tanks: a 2.5 gallon minibow sump, a 2.5 gallon custom cube sps display tank, and a matching 2.5 cube refugium. The refugium sits above the level of the sps tank, to allow it to drain into the main tank by gravity, maximizing pod transfer (hopefully :

: ). The main tank has an external overflow box with a Hoffer Gurgle Buster standpipe (as seen in the third pic). The overflow water enters the sump via flexable bilge-pump discharge hose from home depot (great material for overflow lines, as it flexes without kinking, and yet is ridged enough to not flop around). The end of the overflow line goes to the bottom of the sump, and has a bag of Purigen right at its outlet. Next in the sump, is the UTR/Wetworx skimmer (check out his thread in the DIY library), then the Iceprobe, then the 100w AGA heater (made by hydor, but not crap like the old Hydor heaters). The microbubbles generated by the skimmer and the overflow output are taken care of by lots of very porus LR chunks placed stratigicly about the sump (in all the little nooks and cranies, but not right up against the heater or chiller). Return flow to the tanks is acomplished by a Mag 5 (500gph) ran externally. There is a very noisy 4" Hamilton fan pushing intot he stand right behind the return pump... this only runs durring the day when the heat from the MH needs to be negated. I plumbed the Mag 5 to the sump with a tru-union to ease maintanence (and it has already come in handy... I can take the hole sump out in about 2 minutes when needed). On the inside of the sump, a 90* elbow pointed down is on the intake bulkhead so the waterr is being puled from the very bottom of the sump. This design helps decrease microbubbles also, and ensures that incase the water level drops dramaticly for some awful reason, (say, the skimmer outlet gets pluged, water is pumped out the skimmate outlet, overflowing the skimmate cup, spilling all over the inside ofthe stand, and seeping out onto the floor?... um, yeah

... oops!) the pump does not run dry untill the sump is nearly empty. The return line is flex vianal tubing, and splits via a Y fitting right after the pump, with one line going to the main tank, the other to the refugium. Ball valves allow me to controll how much water goes to each tank, so the main tank gets about 3/4 of the flow, and the refugium gets the rest. 1/2" vinal tubiung fits very snugly on 1/2" CPVC, so I used 1/2" CPVC to plumb the return lines after the split. On the main tank side, the retun water enters the tank via a three outlet Calfo style manifold (4th pic). On the refugium side, the water enters via a simple "spraybar" that extends down the center of the back wall, and T's off at the floor and extends to both sides. the holes in this are 3/16", and create a nice distribultion of the weak flow into this tank. From the refugium, the water simply flows down to the main tank by gravity though the pipe you can see in the 2nd pic. The closed loop system is run by a mag 2 (though I have a Mag 3.5 I'm considereing putting in its place

), and runs through a SCWD. I am aware of the problems some people have had with SCWDs, but IME, they just have to be set up and maintained properly, and they will run for years. The plumbing from the intake bulkhead in the tank to the pump is quite crazy, with eight 45* bends and a piece of PVC called an EZ-span coupling (an adjustable PVC coupling with a union at one end... very niffty) to get around the Iceprobe and the sump. I plumbed the CLS outlets using 1/2" Loc-Line, since the compact Loc-Line ball valves would fit where standard PVC ball valves wouldn't, and the poseable bending action of the Loc-Line allowed me to get the ouputs from the SCWD in the idea positions to maximize flow and reduce backpressure. Inside the tank, the intake starainer for the CLS is in the center/back of the tank floor, and is hidden by the rockwork, and the outlets are at each back corner, with three 1/2" Loc-Line segments ending in a Y fitting. The ATO system consists of a flotswitch from www.californiareefs.com, an aqualifter pump, some sexy black tubing, and a 2 gallon bucket form Lowe's. A hole is drilled in the lid of the bucket to allow the suction line from the aqulifter in, and a couple 1/16 holes are drilled elswhere in the lid to allow air in as the bucket empties. A hole is drilled in the top edge of the sump for the aqulifter outlet line to enter, and the tubing is positioned just above the waterline. Temporaraly, I have a VERY RELIABLE suction cup holding the floatswitch in place, but I will make a little acrylic bracket for it this weekend (I do not recomend mounting a floatswitch for an ATO with a suction cup, because if it fails, your tank is either gonna get all your topoff water, or none at all). The big green cup is the skimmate cup. The turkey baster siting on the ATO bucket is the most versitle aqaurium tool I know of: blows off detritus, picks up untouchable critters (like bristleworms), transfers pods, empties skimmer cup, prime's siphons. Thge jumble of wires is really not a jumble at all, it just looks that way. both powerstrips are securly mounted, and all cords have driploops... there's just 11 cords, so it's a little crazy.. Istill need to instal a GFCI in placeo fthe standard outlet on that wall... hope the house doesn't burn down before I get to that this weekend!

.
The last pic is of my 70w MH DIY Regent fixture mounted in the canopy and shielded by the slpash gaurd. It is actually a perfect fit for the 7" tall canopy, and ismounted quite squrely, I'm not sure why it looks crooked

. It is cooled by two 3.25" Coralife fans, which actually put out a really decent amount of air, and keep it amazingly cool... cool enough to touch the fixture in the middle of the day!

The controller for the Iceprobe is mounted on the outer left side of the stand so the probe can reach into the tank itself (the overflow chamber, actually, right next to the digital thermometer probe), and not sit in the sump with the chiller unit and the heater. I found that having the probe in the sump so near the two temperature control devices caused "feedback," and the temp constantly varried, but by putting the probe in the tank, the real temperature of the tank water is meassured, and the Iceprobe can respond acordingly, and the temp stays between 80-80.5* (right where I want it).There is a little fighting between the heater and the chiller

, but they are workign things out

. I should also note that the chiller heatsink/fan are actually outside of the cabinet, so the heat it draws from the sytem doesn't just hang around in the cabinet and get absorbed back into the system.
So, that's the system... any other questions?

Jared