Originally Posted by
rcomeau
dkeef -
I think I decided to keep discus to build an automatic water changer, not the other way around. An Arduino runs it. It is good fun. Good luck with your solution.
My changer is currently configured to automatically change 50% of my 50g tank every other day. I do not need to touch it while it performs these steps:
1. 20 min to siphon 50% from the tank. 14 ft vertical pipe x 0.433 psi/ft = 6 psi of siphon vacuum (no pump). This step is silent.
2. 15 sec to drain cold water from the fill pump to replace it with warm conditioned water from the bins. This avoids pumping that dead water into the tank.
3. 180 sec to refill the tank. The pump is loud but it is in the basement. The water splashing into the tank is louder because it is up where we are.
4. 4 sec to top off the tank to keep the float switch up
5. Send me a text message of how long it actually took to refill the tank. This gives me peace of mind. Old failures had the symptom of taking way too little time to refill.
6. 140 sec to drain unused water from the bins. I don't like/want the old water in there before filling with new tap water.
7. 800 sec to refill the bins until the float trips
8. 25 sec to top off the bins to keep the float switch up
9. 2 sec to pump water back up into the siphon line because a prior step drained it. This solved a problem that I had where the line was not primed for adhoc vacuuming.
10. 30 hours of doing nothing other than aging the bin water (no agitation nor heating)
11. Begin agitating the bin water
12. 48 hours after the start of the last water change, start the next step
13. 6 sec to dose 6 ml of Prime with a peristaltic pump into a bin because my tap water has chloramine in it. Agitating continues which mixes the Prime.
14. Heat the bin water until over 81 deg F then start step 1. Continue heating until step 3. The heaters are set to max heating. The arduino turns them on/off once at 84 deg F.
Debris vacuuming is manual. However, I simplified it down to pressing a button to trigger arduino to siphon for 90 seconds. It uses the same tubing that stays clipped inside of the tank for water change siphons. The tubing extends halfway down into the water when clipped to the wall and it is also clipped up under the entire length of the top. I pull it out of the clips to reach everywhere along the bare bottom of the tank then put it back in the clips when done. I do not disconnect the tubing nor remove it from the tank. That reduces debris vacuuming to a little more than the 90 seconds. (Except that the siphon is stopped/started with a diaphragm valve. Debris held the diaphram valve open once or twice. I solved that by putting a filter before it. The debris needs to be cleaned out of that filter every month or so.)
The walls still need to be wiped down and filters need to be cleaned about once a week. I try to do that before the water changes so that 50% of the disturbed water is replaced during the next automatic water change.
It ages and agitates the water in two bins. I believe that the purpose is to maximize water/air exchange by maximizing the surface area and maximizing the agitation of the water surface. Two bins have more surface area than I found with any one deep bin. A power head pumps the least agitated water from one bin to drop it onto the surface of the other bin. Ph rises to match the tank at about 7. The way I think of it, the water in the tank is agitated by the filters. Agitation of water change water in the bins conditions it to be closer to the tank water.
The bins are behind the hot water heater in the basement. That made easy access to the cold water supply and the sump pump for accidents. It is also about 14 feet below the tank which provides 6 psi of siphon vacuum. The more vertical distance the better for siphoning. The pump is strong enough to pump the water up that distance from the bins to the tank.
I cannot comment on how this would work with a planted tank although I did keep a planted tank with injected CO2 to control PH at 7 in the past. I stopped that before going bare bottom with water changes for discus. I do not think that most people work out how to maintain CO2 levels when doing high water changes regardless of doing it automatically.
I am glad that I didn't keep it simple. However, to avoid misleading anyone, my approach involved custom electronic circuit design and construction, custom programming, and custom plumbing. The ability to completely customize it was a key success factor. I did not start the project until I was sure that I was capable of building out the entire complex concept. I called it Project Rube after Rube Goldberg. I doubt that most people would succeed with it even if it were packaged as a product with a manual. Perhaps those points are answers your subject question.
The effort obviously does not need to be all or nothing. Some of the harder manual steps can be automated without automating the rest.
Good luck getting suggestions. I will start my own thread to get improvement suggestions at some point. Oddly, some suggestions seem to be biased by how much more work it will add. The changer patiently does the work! It can be configured to perform the most ideal water change regardless of difficulty. I also found it hard to get ideal suggestions in the past.
I hope my post is what you expect and honorably insisted enough to keep requesting. I am glad that you stuck around.