A three gallon box (approx 15"x 12"x 12") was build and elevated above main tank to create a surge every 60 secs. This design is known as a Carlson Surge Box. The box was build out of 3/4" plywood, jointed using wood glue and pocket holes, the inside edges sealed with GOOP, then painted with three layers of 2-part epoxy to seal out water. I used wood instead of a common plastic container. This lets me custom size it to fit my need and most importantly, lessen the noise of fill and drain cycles. The 3/4" thickness dampens the noise a lot better than a typical thin plastic trashcan.

The overkill of pocket holes with square head screws were used to make 110% sure that this box will never ever come apart. Better safe than sorry.... These pics show three openings into the box. The big empty hole will accomodate a 1" bulkhead and will be the outlet of the box. The 1/2" PVC pipe is where the box will drain in case the outlet is stuck and fail to surge. The original design was to use a 1/8" line to feed the box. That turned out to be too slow in filling and did not provide enough water fast enough for siphon to start. I modified it later to a 3/4" line on the top cover. The rabbet cut on top is to provide a ledge for the top cover to sit.

The Fill pipe extends 95% to the bottom in order to minimize the splashing of water. The top, even though is removable has a very tight fit. This will ensure noise from fill and drain cycles is contained. So far this method has proven to be quite effective.

Updates: (Mar 22, 2002) Even though the box worked 100% correctly during initial tests, after all components are installed (PVC bends, fill pipe, etc.) the surge cycle will not restart after the first "turn-on". This is probably due to the cheap valve used to control water fill rate. It is a small 3/4" ABS valve found at the irrigation section at Home Depot. In the next couple of days, I'll replace it with a PVC gate valve for more precise control to match the correct fill rate. Bubbles from the water returned is a common problem to surge boxes such as this. That is another problem I'll try to solve and will document it here later.

Updates: (April 2,2002) A PVC gate valve replaced the original valve for better water fill-rate control. However, water did not stop completely after each surge. This gradually prevented a surge from occurring at all. The bottom of the Drain pipe was cut at a 2 degree angle to help suction of air at the end of each cycle. This proofed to be unsatisfactory. I increased the angle to 5 degrees and the problem persisted. To ease the troubleshooting, I added an union to the Drain pipe. After 10-15 trys with mediocre results (bigger angle; mulltiple 1/4" high 1/64" slits; etc.) I realized that something similar to anti-siphon holes can be installed on top of the bend. This is when two threaded barbs were tapped and installed onto the top of the Drain pipe. Regular airlines were installed and adjusted about 1/2" above the bottom of the Drain pipe. Air was able to suck in before the water bottomed out. When water was drained to the bottom, enough air sucked in and the siphon stopped completely. Not only did this solve the problem of water not stopping at the end of the cycle, but no air purged into the tank at the end also. This is something I have not figured out why yet...but can't complain with this added benefit.

Updates: (Oct 2002) After several months of usage, more problems developed with the Carlson Surge Box. The water input rate was critical in the continual cycle of the surge. I had to adjust the water rate once every two days to insure the continuous operation of the unit. Even though it only takes 5 min everytime to readjust, the trouble was more than what I want to handle. I build a new and slightly bigger surge box. This time I used a standard toilet flush stem and valve. As soon as enough water accumulated inside the box, the bouyancy from the build-in float, forced the valve to open for a surge. After 6 months of usage, I didn't have to adjust the unit once.

The first pic shows the top of the surge box. Notice the condensation collecting on the bottom of the acrylic panel. I used arcylic so the surge cycle can be inspected without removing any component. The second pic shows the outlet of the surge box. It was a 90 degree 1" to 3/4" reducer. I've notice that having it reduced created a better surge than just leaving it at 1". However, if I downsize it further to 1/2", the surge will not be forceful. That would be counter-productive with a "surge box".

The fill tube can be angled straight down. However during a power outage, water will back siphon and cause the sump to overflow. The tube is about 1" away from the box wall, so water will not "fall" in creating unwanted fill noise. The drain stem cost about $8 at Home Depot. This particular brand has a build-in styrofoam block. This eliminated the need for an additional toilet float. Enough water in the box will lift the block and cause the stem to flush. Since it was designed for toilet application, the stem had a 2" drain (if I remember correctly). I used several reducers to bring it down to 1".

The pictures shows the water output and bubble amount at the labeled time frame. I tuned the surge session to take about 65 seconds to complete. Water fill rate affects the time between each surge and also the length of each surge. Obviously, slower water input equals to longer time between each surge. Regardless of time, the same amount of water is required to start the surge. Unavoidable air bubble is introduced at the beginning to every cycle. However, the benefit seen in the reaction of the corals to the surge outweighs the unsightly bubbles.