I have built 2 ATS units and am including both here. These use a somewhat different approach in order to reduce the size and be more appropriate for a small tank. Many thanks to SantaMonica for providing a starting point and encouraging information exchange.
I have one of the bio wheel filters with 2 cartridges sized for 30 to 50 gal on a 29 gal tank but with a 7" fantail, 2" ryunkin, 7" pleco, and lots of food it is under powered. We also put several philodendrons in but the roots started snagging tails so my wife removed them. Then the tank was out of balance and turned green. We added a gravel filter with little effect and replacement philodendron but they take time to establish. At least the roots can be confined to the bubble tubes now. The tank requires daily 20% water changes to limit green water so i decided to try building an ATS. I believe the ammonia is very high but my wife who is not color blind assured me that the test color is not on the test chart. I have tried multiple brands in the past and found they all provide different readings. Our fish sit on the bottom too much and i have had to medicate for white fungus on fins so i know there is a problem. Ammonia absorbing chips have no apparent effect on this tank or maybe I didn't put enough in. Reduced feedings helped stop the fungus but the fish are still not happy. Photo of daily food on a post-it note:
I couldn't find much info about fresh water ATS so I decided to experiment as cheaply as possible. I also wanted a small, efficient system so I came up with the idea of using a single lamp surrounded by a screen on 4 sides. In order for it to be efficient, the water must be limited to one side of the screen so that all of the water is lit. I decided to build it out of a Costco nut jar. I used 2" pvc central vacuum tube to direct water around along with some pvc pipe and bulkheads made from sprinkler parts. The pvc parts cut well but I had too much space between pvc parts and the curved jar sides so i tried gluing on some thin plastic from a salad container. Well nothing i have would reliably hold the different plastics together so I sandwiched on another piece of pvc and it held.
These jars can be roughened but it is not easy and will never look like cactus because the plastic is brittle.
I hooked it up and it ran perfectly and silent. It was quieter than the whisper air pump and had nice even flow.
After a few days the green water started to disappear. The green water algae had started to block the filter sponge and by cleaning it every day, it got completely clear. All was not good though. The ATS was not growing algae on its sides as intended. After a 2 week trial, there was almost no growth on the rough plastic sides but there was 1" long strands growing on the pvc at the top and bottom. I siliconed on some craft mesh but it splashed and was noisy. I gave up on this scrubber because the opening was too small especially after installing the pvc and cut up my hands. In these photos, a light bulb socket extends through the hole in the lid. It is the same socket used in the 2nd try.
Had it ever worked, cleaning would require disconnecting the fill tube from the pump. The entire ATS then would be brushed with a clean dish brush.
sorry, my pictures are limited.
------------------------
2nd try:
I wanted to be able to experiment with different material and angles if splashing happened again so i set it up to use solid material of any type. All sides are clear 1/4" Acrylite FF acrylic except the lid which is mirror acrylic. FF won't hold up in a large aquarium but should be fine for this and it was all out of the bargain bin. I decided to use cement/fiberglass boards that i had lying around because I heard algae likes cement. This is the stuff used in shower walls and before tiling. In order to avoid bloody knuckles, i decided to carry the water in a trough around the outside. My wife hated the visible white pvc in the first setup so most of it is gone on this version. I used some thin bent acrylic to press the boards against the acrylic weir and avoid water going behind where it is not lit. I ended up making it taller than originally intended because the acrylic scraps were a little larger than my design and I saw no reason to waste it. The water returns to the tank down a waterfall with an acrylic landing just below the water surface similar to many hang on back filters. The light is sealed with silicone but may be enclosed in acrylic or glass later. It was wired through a jam lid for that purpose. I will decide after trying other light bulbs if necessary. A brace down the back of the tank prevents tipping and helps with leveling. At the pump connection, a small piece of acrylic is used to break water momentum that would shoot out the top otherwise. A number of changes were made during construction so the design is only an approximation of the actual. The boards are vertical instead of slanted and the weir is not notched. Otherwise it is very similar.
system volume = 29 gal
pump (Maxijet900) flow =200 gph approximate
"screen" width (1 sided) = 20 in
"screen" area = 140 sq in
light = 14w cfl 5500K on 16 hours
projected power use =14 kWh/month = $4/month (mostly offset if other filter is abandoned)
PG&E rate at 3rd tier is $0.295/kWh. This is usually where we fall for additional usage.
acrylic cost = $10
lamp fixture = $5
timer = $15
pump = $18
all other items were surplus from other projects.
It appears mostly professional from a distance, even though my first acrylic project in 15 years is somewhat sloppy with drips and bubbles in joints. Amazingly, there are no leaks and no silicone was needed except on the bulb. This version is much less bright in the room since the light is surrounded by cement board and a mirror. Light leakage is bright enough to act as a night light for our stairs and foyer but indirect and not blinding while looking at the tank.
The water trough catches detritus because it is too deep and slow on the side away from the pump but I am not sure what to do, if anything.
I was concerned that the ATS would not sit flat on the tank but it is near level and water does flow on all 4 sides. I may make adjustments if growth is uneven. For now it adds to the experiment.
The internal flow is smooth with no splashing. The return is not as quiet because it is undersized but i believe i can fix it by placing another small piece of acrylic below the release. I may also have to raise the bottom cement board a little more. Alternatively I may enlarge the opening. It runs nearly silent when I put an additional piece of acrylic to expand the landing just below the tank water line Until the tank is clear enough to see the problem, i will just wait.
If another version is necessary, I may use black acrylic on the outside. This version may get paint.
I had some water on the floor the first day. It turned out to be splashing where water returns to the tank. water wicked between the lid and ATS, then down the back of the tank. I found the problem within 3 hours so there was less than a quarter cup and no damage. The fix was to fill the tank to replace the water now in the trough. Raising the water stopped the big splashing and made it much quieter. I will make a few more adjustments to eliminate any chance of this happening again once the water clears.
After 3 days, algae is starting to grow on the cement. Since the green water had a head start, it took off first and the tank looks very green. At least I know the lights will grow algae. I hope the hair algae starts soon so the water will clear. The fish seem to have more energy.
After 6 days, the cement is looking greener but still most of the growth is in the water column. There is no noticeable increase in evaporation with the lid on. Interestingly, the side with the most water has the least algae. All sides are 1.5" from the light. I think maybe I should have scratched up the cement. I didn't because it was already somewhat rough. I think I will roughen one of the sides that has the least growth this weekend. Ammonia is still very high on the test strip but the fish are not sitting on the bottom any more. I am going to have to think about other possible materials to try growing algae on and maybe find a higher wattage light or a different spectrum.