vacuum break gravity feed ATO

[HankB]

This has come up in a couple of threads but this idea never seems to gain traction. I'm curious why. I've started to build one and it seems to work fine so far.

 

The concept is very simple - fill a bottle with water, connect a tube to it and suspend it above the tank with the end of the hose at the level at which the level should be maintained. When tank level drops, it admits air to the line which rises to the bottle and allows water to drain into the tank until the tank level reaches the end of the tube. Without the possibility for air to get into the hose, vacuum holds water in the bottle until tank level drops (*)

 

I'm aware of some prior work on this. It was discussed in this thread, but no actual testing was reported until I did some. I put RO/DI water in one glass and 1.025 SG salt water in another (filled to the same level.) I then connected the two with a piece of 1/2" ID siphon tube 18" long. The tube was prefilled with RO/DI water. I wanted to see if enough salt would diffuse up into the RO/DI water to affect the salinity in the vessel (which for this test contained a pint of water.) After a week, the salinity had not dropped by more than about 0.001 so I concluded that diffusion would not be a problem.

 

More recently someone pointed out this thread (or at least gave me a term to search for - tea jar ATO ) It looks like there were several tries but no reports of long term success. The problems they ran across are exactly what I anticipate: The system is sensitive to air leaks that compromise the vacuum. Also I did not see a particular advantage to using two hoses. It only adds complexity and allows additional connections that can leak.

 

For a reservoir I chose a clear plastic juice bottle with a volume of about 3/4 gallon. (My plan is to use this on a 10g tank so that should cover about a week's evaporation from my experience.) That was more rigid than polyethylene and just as cheap. It needs to be somewhat rigid so that suction generated by weight of water in the line does not cause it to collapse.

 

The feed tube was chosen as the smallest diameter that would admit air when filled with water and inverted - 7/16" ID clear vinyl hose. I could probably use a larger diameter, but with a smaller diameter there would be no flow.

 

The feed tube was fitted through a slightly smaller hole in the bottle cap. Attaching it was a bit of a problem. I could not find any glue that would actually adhere to the bottle cap which I believe is made from polyester. Super glue gel, some plastic glue and silicone sealer would all stick but could be peeled off. I did find that the super glue adhered pretty well to the vinyl hose, so I glued that to the cap, putting a thin bead on each side of the cap itself. That would provide a mechanical connection to the cap. After that dried I added a layer of silicone aquarium sealer to provide an air tight seal. With the super glue providing a mechanical support, I think the silicone will hopefully not be pulled off the cap or the hose.

 

Next I need to figure out how to attach the hose to the tank in a manner that will reliably hold it in place. My plan is to get some vinyl line just big enough to fit the feed line inside. I can glue a piece just inside the tank and another piece on the feed line that will act as a stop. I anticipate that the level in the sump on the tank will be low enough to allow a decent mount using this. Another alternative would be the using an airline hose clamp as was done my airman_verde04.

 

That's it. No float switch, no pump. No problem with the equipment if it runs dry. The concept is simplicity in itself.

 

It does have disadvantages:

- The reservoir must be mounted above the tank. The feed line must be run downhill to the tank. This makes an aesthetic mount a lot more difficult than with a pump driven ATO that can be mounted below the tank.

- The system must be air tight or it will simply drain into the tank.

- The end of the hose must be fixed within the tank. If it can move, that would be equivalent to a loose float switch in a conventional ATO.

- There is no equivalent of a safety fallback switch as is frequently used as a backup to the primary sensor in a conventional ATO.

- This would not scale well to a larger reservoir. Well... maybe if a rigid enough large bottle could be employed.

 

I'm curious if anyone has gotten something like this to work long term. I've had it maintaining level in a plastic dish on our kitchen counter. The test installation uses a drop of 32" from the mouth of the bottle to the water level. I plan to reduce that to about a foot on the actual installation which will be on an as of yet undisclosed project.

 

Thanks for reading this far! I'd appreciate hearing about any other projects along these lines, particularly long term experience.

 

thanks,

hank

 

(*) Technically, atmospheric pressure holds the water in the bottle. The atmosphere can support water to a height of about 32 feet at sea level, so the reservoir must be below that point relative to the tank - ignoring vapor pressure of water at room temperature.

[masterbuilder]

Here are your fittings:

http://www3.towerhobbies.com/cgi-bin/wti00...LXD708&P=ML

 

They work well and are cheap. They wok through the cap or the bottle itself. I have used them before, many times and never had a leak. For others go to towerhobbies.com and do a search for "fuel fittings", you will find dozens and dozens of them.

[HankB]

Here are your fittings:

http://www3.towerhobbies.com/cgi-bin/wti00...LXD708&P=ML

 

They work well and are cheap. They wok through the cap or the bottle itself. I have used them before, many times and never had a leak. For others go to towerhobbies.com and do a search for "fuel fittings", you will find dozens and dozens of them.

 

I probably should have checked them out. I'm using the fourmost fittings for a DIY CO2 setup on my planted tank. Unfortunately I need something at least 7/16" ID and I didn't see anything that large. If I could find it, a bulkhead fitting like that would be great!

 

thanks,

hank

[evilc66]

Check McMaster Carr. They have almost anything you could need.

 

Sounds like a good project. Pictures?

[tyty22]

im trying this as well so i will be tagging along

[Zombo]

Tagging along.

[HankB]

Check McMaster Carr. They have almost anything you could need.

 

Sounds like a good project. Pictures?

 

That's an awesome site! But I didn't find what I want - a bulkhead fitting with a 1/2" ID barbed connection.

 

Pictures will come as I get a chance, though there's not much to see. Imagine a plastic bottle with a hose glued to the cap suspended upside down above a tank. There's a makeshift holding device consisting of an oversize spring loaded clamp at the tank end.

 

There was some discussion of reservoirs in another thread and I'll respond to them here (so as not to clutter someone else' tank thread.)

 

One concern with the reservoir is that it be fairly rigid. The weight of water in the hose will tend to collapse the reservoir. When air admits to the hose (as tank level drops) it will allow this water to drain and the reservoir will suck in enough air to return to normal size. Then the hose will fill with water and once again try to collapse the reservoir. If there is a lot of flex, it may allow too much water to drain into the tank.

 

 

FWIW I'm putting my money where my mouth is. I installed the ATO on my QT tank yesterday and I'll be watching it today. I only have about a quart of water in the reservoir at this time, but so far it is working well.

 

-hank

[HankB]

That's an awesome site! But I didn't find what I want - a bulkhead fitting with a 1/2" ID barbed connection.

Yes I Did! Part 2974K861:

 

(Not sure if the picture will link, but you can go to mcmaster.com and search. I've ordered them and they should ship today.

 

thanks,

hank

[HankB]

One week update. Actually it has been running over a week. And it has been doing fine, maintaining an even level.

 

I'll add two disadvantages:

 

- Not as easy to stop/start as an electrically driven unit. I take the bottle down and at that point it sucks water from the hose back to the reservoir. When I put it back in place, it drops some makeup water into the tank as pressure equalizes. If done repeatedly, this would add some make up each time it was done and eventually add too much water.

 

- The outlet at the tank must be firmly located or the level will not be consistent when refilling the bottle. The 'level setting' is more likely to be disturbed than a float switch during maintenance. This can be overcome by a decent mount for the end of the hose - not the temporary clamp and zip tie I'm using now.

 

The next generation will use the barbed fitting installed in the bottle cap. That also requires larger 1/2" ID hose. I plan to seal that with silicone sealer as I did with this one, but because it will only require a very small amount of silicone sealer, I will do that when I have some other work to do with silicone sealer (in the near future.)

 

Overall I'm still pleased with the performance. It maintained level over the Labor Day weekend when we were out of town. I had to refill it on our return and it continues to work without any problems.

 

One problem I need to deal with down the line is the mount for the hose on the next tank with which this ATO will be used. That will have a relatively short distance from water surface to the edge of the tank based on initial flow tests and that leaves little space for the mount. Perhaps a below water mount with a vent hole in the hose will meet that need.

 

Incidentally, I had to buy 10 of the fittings pictured above so if anyone wants one, let me know and I'll sell spares for cost ($1.53) plus postage.

 

thanks for looking,

hank

[tyty22]

post some pictures please

 

i have a idea of my own. im going to use pvc piping.

[HankB]

post some pictures please

 

i have a idea of my own. im going to use pvc piping.

 

I will soon be developing v1.1 and I'll get pictures of that. The upgrades include:

- the fitting pictured above installed in the bottle cap.

- larger (1/2" ID) tubing

- vent hold drilled in the tube at water level. That way the tube can extend deeper into the tank to make for easier fastening.

 

How do you plan to use PVC? I was thinking that a piece of vertical PVC several inches in diameter would make a good reservoir because it would be rigid. Actually, there would be no reason it would have to be vertical. It could be run horizontally inside or behind a hood with the hose dropping down to water level.

 

-hank

[yardboy]

Sounds interesting.

Possible resolutions to a couple of your problems. How about a valve in the hose so that it could be valved out when you are ready to add water to the reservoir?

Also, how about using a piece of acrylic that is bent so that it has to be flexed slightly to clamp to the side of the tank. then a hole that is a tight fit that the hose could be slipped inside of and at the proper level ?

I've got something similar in my sump so the hose doesn't slip out when water is pumping, keeps it from slipping out and flooding the floor with topoff.

[HankB]

Sounds interesting.

Possible resolutions to a couple of your problems. How about a valve in the hose so that it could be valved out when you are ready to add water to the reservoir?

Also, how about using a piece of acrylic that is bent so that it has to be flexed slightly to clamp to the side of the tank. then a hole that is a tight fit that the hose could be slipped inside of and at the proper level ?

I've got something similar in my sump so the hose doesn't slip out when water is pumping, keeps it from slipping out and flooding the floor with topoff.

 

 

I don't like the idea of a valve because it provides additional locations that can leak and provides little value in exchange.

 

I like the idea of something that slips over the edge. I wish I knew more about working with acrylic.

 

thanks,

hank

[suziq21]

One week update. Actually it has been running over a week. And it has been doing fine, maintaining an ev

 

Incidentally, I had to buy 10 of the fittings pictured above so if anyone wants one, let me know and I'll sell spares for cost ($1.53) plus postage.

PMing you about an extra one

[rickh]

How about some pictures?? R

[BKtomodachi]

How do you overcome the osmotic gradient?

[supraeli]

How do you overcome the osmotic gradient?

 

i dont think there is a membrane for there to be any osmosis. however, there may be some saline diffusing through the system along the concentration gradient, but diffusion is inversely related to distance, which in this case is quite far in molecular terms. also, the cross sectional area of the tube is small, which is directly proportional to the rate of diffusion. lastly the difference in concentration is pretty minimal (0.035 molecules of salt per water molecular in 1.025 SG at 75 F) which will decrease as distance from the tank increases, so there will be little driving force behind the diffusion after some point along the tube.

[HankB]

i dont think there is a membrane for there to be any osmosis. however, there may be some saline diffusing through the system along the concentration gradient, but diffusion is inversely related to distance, which in this case is quite far in molecular terms. also, the cross sectional area of the tube is small, which is directly proportional to the rate of diffusion. lastly the difference in concentration is pretty minimal (0.035 molecules of salt per water molecular in 1.025 SG at 75 F) which will decrease as distance from the tank increases, so there will be little driving force behind the diffusion after some point along the tube.

 

What he said.

 

I was a bit more empirical. I connected a glass containing 1 pint of RO/DI water with another containing 1.025 SG salt water mix using a 1/2" ID tube filled with RO/DI water. After a week I could not detect a significant drop (~0.0005) in salinity so I concluded that diffusion was insignificant in this application.

 

I'll get pix when I set up version 1.1 in my new 10G build.

 

thanks,

hank

[HankB]

post some pictures please

 

Here you go.

 

This is V1.1. I have gone to a barbed bulkhead fitting that is attached to the bottle cap and sealed with silicone sealer. I put a thin smear of silicone on the fitting's flange and tightened just enough to extrude some. When it had time to cure - weeks in fact - I tightened it some more to put mechanical pressure on the silicone. Then I switched to 1/2 ID clear vinyl tubing to fit the barbed fitting. I clamped it to the barb using a zip-tie.

 

At the tank end I tried to fasten it using a short piece of 5/8" ID tubing siliconed to a baffle. That popped right off because the silicone doesn't really bind to the clear vinyl. Instead I zip-tied it to the leg of a cooling fan. About an inch from the tank end of the hose I drilled a 3/16" hole. That admits air and determines the water level. With the original plan, the 5/8" ID sleeve was going to be below water level so this technique can be used if it is convenient to mount the hose with the outlet below water level.

 

Here's an image of the tank end:

 

And the reservoir end:

 

And the overall layout:

(Apologies for the crappy exposure on that last one, but it's not going into the contest!

 

I've been using this for several weeks now on two 10g tanks. I find that both do a great job of maintaining accurate levels. I thought it would be awkward to "turn them off" during water changes and such, but I just unhitch the bottle and set it down next to the tank.

 

Thanks for looking,

hank

[freddy4130]

This should do just fine. Just fill her up and turn her over.Prolly have to snip the tubing on the inside of the tank that extends to the bottom, but it is airtight and hard as a rock. I actually have one that I will try out in to an ice chest for a week or so. I'll let you know.

http://www.homedepot.com/webapp/wcs/stores...uctId=100637818

[freddy4130]

Ok. I lied. I don't have one. Well, I do but it has a valve for releasing the pressure. It leaks like mad so it would be useless. For $11, I could use a new one anyway. I may go get the cheapie, no valve having one and see if it would work for ATO.

[jmt]

This was done multiple times in different forms with mixed results.

 

Osmosis occurred in every case.

 

The most functional way was to actually make a "U" at the bottom of the tube so that the opening is pointed up, not down. The whole "U" has to be submerged. This yielded the most consistent refill with the least amount of mixing.

[gandhii]

I'm confused about this osmosis thing being spoke of. Isn't the intent to have the outgoing tube to drip into the tank from above the water level? How is there to be osmosis when there is no connection between the two bodies of water except when the water is pouring into the tank?

[c_k_kuehne]

Isn't this the basic principle behind a pet water bottle for ATO in small nano's or pico reefs?? Although you are taking it to a bigger scale.

[yardboy]

I'm confused about this osmosis thing being spoke of. Isn't the intent to have the outgoing tube to drip into the tank from above the water level? How is there to be osmosis when there is no connection between the two bodies of water except when the water is pouring into the tank?

If the outgoing tube dripped, what would stop it from emptying the reservoir of top off water?