If you’ve ever had a water leak in your home, you know the sinking feeling that comes over you as you walk through the door to the sound of running water. [Greg] knows this feeling quite well, having returned home to a sopping wet floor and an overflowing reef aquarium on more than one occasion.
Both of the overflows he experienced were due to a clogged drain in his display, but there was little he could do as far as walling off the drain from potential blockages. With all of the delicate creatures living in the tank, the only possible solution that came to mind was monitoring the aquarium’s water level.
Unfortunately he had no idea how to get this done aside from using probes (which would rust in the salt water) or expensive off the shelf systems. [erich_7719] from the All About Circuits forums helped [Greg] out and designed a circuit for him which would monitor the water level using an IR sensor. The circuit simply shuts off the pump if the water level gets precariously high. As you can see in the video below it works quite well, and as a safety measure, requires a manual restart of the pump once the high water sensor has been tripped.
If you have a need for the same sort of setup, swing by his site for a detailed schematic as well as a bill of materials.
fish is food dogs are pets
IR seems a bit overkill, I figured a pair of wires hooked to a transistor could flip a relay to shut off the pump.
even if he is worried about rust or corrosion, stainless steel would have worked fine.
I probably would have gone with a cork and a lever switch (like DigiKey CKN9913-ND). If you go with a normally-closed switch, you can just hook up your pump directly to the switch without extra bits.
@therian
Trolling is silly hacks are worthy :)
i would use a float. like in the toilet tank. connect to a switch to shut the pump off when the water rises. and the pump will also turn on automatically when the water level goes back down.
YOWWW!! this idea is good though, and alot smaller.
Use graphite probes to prevent rust.
Teflon coated capacitive sensors can be immersed in the water and never ever corrode, even in corrosive conditions. But for salt water, plastic ones work too and for much less cost.
Better still, just mount them outside of the tank and sense the water level directly through the glass.
You can get 12 volt ones used off of ebay no problem for awfully cheap. Interface to an Arduino that controls a valve and presto. If too full, turn off valve and or e-stop the pump and sound an alarm. Simple, no moving parts.
Another idea is to use float switches. Made from polypropylene so rust again isn’t an issue. They can get gummed up though with algae or gunk or debris so capacitive is preferred in my book.
So what is actually happening? Is light reflecting back up to the ir sensor when it’s out of the water, then not when it’s underwater?
The cat genie uses a similar method of water level detection. My understanding is their version isn’t very robust.
By similar method, I mean IR.
Blah, this is stupid. This is just adding ANOTHER point of failure to a poorly designed system.
Proper aquarium design means you design your system so failures CAN’T happen. Poor design means you design to lessen the impacts failures.
The proper setup for this system would be to have two drains; one for regular use and one for backup. It’s called the herbie method and is VERY well known in aquarium circles.
At the very least, design the sump to allow the return pump to suck air before pumping enough water to overflow the tank.
It does seem like a normally closed float switch would be a much easier and reliable way of handling things. Once the float is lifted, power is cut on the pump until the user returns to figure out the problem.
Like Bruno said, there should be some kind of overflow drain installed regardless.
I would have gone with a contactless capacitance sensor. Put a pair of metal strips on the outside of the glass, and it can sense the water level with surprising accuracy. Definitely add an overflow drain, too. If you’re really paranoid, set the entire aquarium on a tray with its own drain to catch any leaks.
I hope Greg is properly ashamed of himself for setting up his aquarium in a non-hackaday approved manner. It would be best if he gave away his tanks and sat in the corner thinking about how he’s failed us.
Ha
http://www.sparkfun.com/products/9072
this product may also work well for measuring the water level
hmmm no, i was thinking of http://www.sparkfun.com/products/10221
posted the wrong one in the other comment
Capacitive sensor on the outside of the glass or above water level. Set to signal at appropriate level. Problem solved.
I appreciate all the feedback. Love the Monday morning quarterbacking too. How many featured projects have you had, Bruno…?
Anyway, the only legit criticism I have seen in the comments so far is the sensor used. The guy who designed this for me suggested this sensor and so I used it. Pretty simple logic. Any mechanical sensor is prone to failure, already tried those. And honestly I just wasn’t comfortable with the probe idea.
The normal water level is above the trim on the tank so monitoring it_through glass_ is not possible.
Lastly, this tank is tiny. Two overflows would be impossible. One was hard enough. The water is being pumped up from a sump, which is not clear in the video. There is already one overflow plumbed into the back glass. That’s where the water goes when it drops over the black ledge seen right below the sensor.
Greg, i am in need for a similar sensor for my nano tank. please, email me when you get this message.
This is cool, but a closed feedback loop system that doesn’t require you to reset it would have been even cooler and just about the same difficulty.
having just adjusted my toilet tank 10 minutes ago, the float jumped into mind, (although that would probably be bigger).
CutThroughStuffGuy (and anyone else interested in catgenie sensor details):
The cat genie optical sensor uses two clear plastic light pipes (xmitter and rcvr) both connected at the far to a clear corner reflector (it’s really all one piece of clear polystyrene in an elongated “U” shape with a pointy end).
When the corner reflector is immersed in water — which has a different index of refraction than air — the amount of light returned to the sensor side changes.
The lack of robustness you refer to in that system is due to the fact that lime scale or other stuff found in litter boxes can stick to the end of the sensor, blocking it from sensing the rising water. I requires regular cleaning to operate properly.
I don’t think that a fish tank would have nearly the amount of fouling on a sensor that a litter box would, but if it did, regular cleaning would be a good idea.
@griffon
stainless steel WILL rust in salt water
Really good idea using IR. Mechanical linkages and parts are prone to failure, and having a float on the surface is not ideal. First problem that comes to mind is salt buildup on the pivot of the float preventing it from raising.
What did you use to prevent corrosion to the leads and circuitry? (other than tinning the wires)
I would have used a constrained plastic float with a reed switch personally. But without over-engineering things we’d still be happy sitting around our fire-warmed caves.
Corrosion is prevented by two methods. First, most wires are far enough removed from the water to be safe. Second, any wires close are heat-shrink-spliced. This seals them completely.
Over-engineering this is not. If you know saltwater, you know saltcreep. It infiltrates everything! It will jam a mechanical solution no matter how it is maintained. Solid state, not corrosive is the only solution here.
I’m with Greg. Saltwater and anything mechanical will fail without constant maintenance. If not from salt creep from little creatures that will attach and grow. I had a float/reed switch for an auto top off that would always have little white hard creatures attach to the shaft prohibiting smooth operation of the float.
A lot of the critics of the above project do not really understand the issues involved in keeping an aquarium. Firstly, floats are out because this is a display tank. It’s supposed to look like a piece of coral reef, not a toilet cistern. The necessary equipment is already a pain without adding more. Secondly metals in probes are toxic to marine life, so corrosion is not the only issue with that.
Greg has dealt with some of the other issues himself.
For those of you asking for the pump to automaticlly restart…if you did auto restart the pump what could potentially happen is the drain is only partially blocked…the water rises and trips the overflow sensor…the water slowly drains out and the pump is reset. but as the pump pumps more water in the tank than can be drained we creat a cycle of the pump turning off and on repeatedly. Now if you watch his video there is a large burst of air when the pump turns on. If that were to happen say once a minute for a few hours your going to have a massive problem…ie dead fish, dead coral. All that excessive air is going to trigger a change in the dynamics of the water such as PH, Temp, and oxygen content of which allot of fish and corals cannot handle. i can also see the point that leaving it off as well can be a problem due to no circulation.
I saw on your website about top off shut off switch for overflow tanks and I look at the list there was there to order what you need to make one but some says unavailable do I have to contact the company to see when it could be available the total price are all this was about $55 and some cents well those have a plug-in for the motor pump plug into and switches on top of the tank or sensor or any updates are different now