Introduction: Modular, Windowsill, Air Powered, Hydroponics Herb Garden/Experimenter

Although this is an evolution of previously entered work, there has been significant changes to the whole modular idea as well as some control mechanisms added.

I have learned along the way that the back flow of air through the pump IS definitely a problem, especially when working with air tight systems or relatively low height water columns. For this reason, I have added two mechanism to aid with the control of the fill and drain cycles. I have also added a central "manifold" through which all air flows, allowing for expansion/experimentation.

The air escaping through the planters, when the reservoir empties, was also a problem, so I added a separate bottle that I call a bubbler. The bubbler redirects the escaping air and reduces the bubble size, reducing splashing, vibrations and effectively extending the pump on time. I suppose it could be used as a planter as well. Don't know what you'd call the bubble method.

The MODULARITY has been expanded so that the "HUB" of the system is the Bubbler and the TEE below. Everything else is swappable. You can change reservoirs and/or reservoir types as well as "planter lines" and types of planters at your whim. This allows for experimentation and expansion without having to disturb that which does need to be disturbed.

I do plan on improving/screwing with stuff as time goes on and required a "productive" window sill garden that accommodates different designs and methods, without starting from scratch each time. This what I've come up with and hope to be adding "Modules" as time goes on.

Step 1: Modular Reservoir


1 ea 2L pop bottle with cap
Approx. 18 inch length of 3/8 inch OD Vinyl tubing (trim as required in final setup)
1 ea 7/8 inch length of 7/16 inch OD tubing
1 ea 5/8 inch length of 1/2 inch CPVC tubing
Hot melt glue (Silicone would probably be best)


1. Drill one each 5/16 inch and 13/64 inch holes in the cap as shown.
2. Cut both ends of the 2 ft length of tube on an angle as shown.
3. Thread the 2 ft length of tube through the 5/16 inch hole in the cap as shown.
4. On the end of the same tube, that will be outside of the bottle (cap orientation), slip the 7/16 and 1/2 inch pieces over each other as shown to form an adapter.
5. This adapter will leak without tension applied, so hot melt glue as shown to seal.
5.Place cap on bottle and adjust tube so it reaches the bottom of the inside of the bottle.

Step 2: Bubbler Assy


1 ea narrow and relatively tall soft drink bottle w/cap
Plastic mesh bag(s)
1 ea 1 inch length of 1/2 inch CPVC tubing
1 ea 1/2 inch CPVC TEE connector


1. Drill 13/64 inch "vent" hole in the bottom of the bottle as shown.
2. Strip mesh bag(s) of anything other than mess, such as labels or string.
3. Loosely stuff mesh into bottle as shown. Evenly distribute inside bottle to avoid channels where large bubbles can form and run.
4. Drill a 9/16 inch hole into cap and expand with a piece of 1/2 inch CPVC tubing as shown. Place cap on bottle.
5. Place the TEE on one end of the 1 inch piece of CPVC tubing and the bubbler bottle on the other as shown.

Step 3: DWC Module - Deep Water Culture


1 ea 2L pop bottle
1 ea small pop bottle w/cap
1 ea 13 inch piece of 7/16 inch Vinyl tubing (mine was a 12 inch scrap and was minimally long enough)
Hot melt glue
Approx. 4 inch piece of coat hanger wire
1 ea aquarium air stone
1/4 inch aquarium air tubing as required... at least 1 foot


1. Cut a 5/16 inch slot in one of the recessed groves on the bottom of the 2L bottle as shown.
2. Cut one end of the 7/16 inch tubing on an angle and insert same end just inside the slot. Put tension on the tube and glue in place as shown. I used two elastics to secure will glue cooled, which are in the rest of the pics and are not necessary, but could be used as min/max markers.
3. Bend wire as shown and use it as a clip, at the neck of the bottle, to secure other end of 7/16 inch tubing in a vertical position.
4. Cut a 1 inch hole near the top of the bottle, opposite the tube.
5. Using scizzors, cut the neck off of the small bottle and insert into 1 inch hole as shown. Use a small bottle as they are not "flared" like the 2L are and it makes for easier insertion into the 1 " hole.
6. Drill a 13/64 inch hole into the cap.
7. Insert air tube through through 13/64 inch hole and put the air stone on the end that will be inside the bottle (again, cap orientation).
8. Put cap on the "side" cap and adjust tube so that the air stone rests on the bottom.

NOTE : Stones, and any submerged hole for that matter, will clog over time changing the associated rates. Routine observations of the full cycle wouldn't hurt to keep wandering variants under control if necessary.

Step 4: Flood and Drain Module


1 ea 19 inch length of 1X4 pine (same length as a pop case)
3 ea 1 inch thread detergent bottle caps (convenient)
1 ea pop bottle lid
1 ea 1 inch length of 1/2 inch Vinyl tubing
3 ea 1/2 inch lengths of 1/2 inch Vinyl tubing
3 ea 1/2 inch CPVC TEE connectors
1 ea 1/2 inch CPVC 90 degree elbow


1. Drill four equally spaced 1.25 inch holes at 4.5 inch centers w.r.t. each other. (same spacing as a pop case)
2. Build "interface" from bottle cap to I/2 inch CPVC connectors. Please see pics below for details. I made 3 using detergent caps and 1 from a pop bottle cap.
3. Secure bottles and "interfaces" through holes as shown.
4. Interconnect the bottles using 1/2 inch CPVC or Vinyl tubing. I use Vinyl for this because of its' "flexibility", excuse the pun, but would probably use CPVC in a larger or permanent setup for its' rigidity.

Step 5: Manifold Assy

Note - Manifold only necessary if multiple lines required... aka DWC unit. Should the DWC unit not be desired to take advantage of excess air, as in my first unit, Expandable Hydroponics System from Junk - Flood and Drain, the "self tapping screw valve" can be "installed" directly on the reservoir bottle cap. The soft plastic will provide a good seal and it will cost nothing to fix if you screw up again and again and again..... Gotta love seemingly endless, FREE, building materials, eh?

Please note that a gang valve, with it's individual line controls, would be preferred for multiple lines, but for those who do not require such controls................


1 ea several inch length of 1/2 inch CPVC tube
2 ea end caps for same
1 ea 13/64 inch drill bit
1 ea 1/64 inch drill bit
1 ea small, self tapping screw
Air tubing as required during assembly


2. Install end caps on each end.
1. Drill 1 ea 1/64 inch hole, near one end, into, but not through tube.
3. Tap 1/64 hole using small, self tapping screw and screw in fully when complete.
1. Drill 3 ea 13/64 inch holes into, but not through, tube as shown.
3. Insert aquarium air tubing securely into holes as required.(not shown)

PRESTO - A one input / two output (at least in this case) air pressure manifold with a passive, adjustable, pressure relief valve. Please excuse any incorrect terminology. Additional holes (ports) could be added as required until maximum air usage limit is reached or reservoir no longer empties completely.

Step 6: Modular Planters


4 ea 2L pop bottles


1. Cut the bottom off the bottles.
2. Drill several small drainage holes in the bottom piece.
3. Invert bottom piece and force into the top piece as shown. Trim if required, but it should be a snug fit.

In the photos, only one is complete, the other three are for demonstration purposes only.

Step 7: Assembly

You are on your own for a support structure. I used what I had on hand and the rules are simply dictated by gravity.

Hopefully the pics speak for themselves.

Step 8: Operation and Operating Parameters

Turn on and inspect for and repair any leaks, both air or liquid.


Pump comes on and pressurizes the manifold, initiating the emptying of the reservoir up the tube and into the Flood and Drain unit, as well as the bubbling of the air stone in the DWC unit. This continues until the reservoir completely empties or equilibrium is reached between the opposing forces. It can be "tuned", and will be discussed seperately. As the reservoir empties, the bubbles will be redirected, largely, up the tube and into the bubbler rather than the air stone.

The greater the resistance from the air stone, the faster and more completely the reservoir will empty.


Pump goes off and the water that has evacuated the reservoir returns under the force of gravity. The returning nutrient pushes the air up and out, into the manifold, and out through the DWC air stone as well as any other leak available. The bubbling of the air stone will reduce as the weight of the returning nutrient diminishes, until all of the nutrient returns or an equilibrium of the opposing forces is reached. This can be easily tuned as well.

Please see video for demo of unit w/DWC only as a vent -

Please see video for demo of reservoirs only-


Tuning is achieved by varying the opposing forces, such as gravity and air pressure. Generally speaking, gravity affects various "thresholds", while air pressure affects rates and the ability to reach the thresholds or a state of equilibrium. At least in my understanding and my ability to explain same. Lets just say, it's a balancing act.

The effects of gravity can easily be varied, and observed, by increasing or decreasing the height of the planters w.r.t. the reservoir or by varying the depth of the air stone in the DWC unit.

The effects of air pressure, acting against gravity, can be regulated by the porosity of the air stone in the DWC unit, by varying the depth of the "self tapping screw valve" in the manifold or reservoir lid, or any other method of controlling the "air tightness" between the air pump and the reservoir.

Using air tightness only, one can overcome the issues with the slow drain rates regardless of pump used, as long as it can supply enough air and pressure that is. With the DWC unit eliminated, the excess air will simply vent out the screw valve thingy.

NOTE: Only one complete system can go onto any one individual pump because of the pressure loss that will occur when/if the reservoir empties. Any slower systems will stop filling, and start draining, when the first one starts bubbling. One larger, shallow reservoir would be preferred over multiple reservoirs in larger systems.


cortes96 (author)2013-06-16

How much did this cost?

mkimmerle (author)2011-10-21

How much would you say this would cost assuming you would use recyclables for the bottles?

Groove22 (author)2009-01-27

Would there be any way of making the system automated? It would be nice to have the pump kick on at preset times through out the day and shut off after X amount of minutes

wiley coyote (author)Groove222009-01-27

There is a 6 cycle/day Christmas light timer running the pump. You can see it behind the tea stained reservoir. I keep screwing with the stuff, but generally, it runs 6 times a day for 2 minutes at a time. Things keep growing .... cats keep eating them.

runningman53 (author)2009-01-15

Being ignorant in this field I have to ask what is the purpose of having to move the water from one bottle to another and then back again? Couldn't you just bubble it where it's at? And to what advantage does one get by doing this verses planting in the soil and not useing any energy?


The water moves so that the roots do not drown and stay oxygenated. I've thought about the bubbling thing and I don't see why not. That last one is a question that I'm not going to touch, other than automation. There are a lot of opinions and pros always come with cons.

prox (author)2008-12-18

what is the bubblers purpose? is it a place for the air in the resevoir to go when the water returns? does the bubbler eliminate back flow into the air pump? this is my main concern. ps. very cool, ive been trying to come up with something similar without much luck.

wiley coyote (author)prox2008-12-18

The bubbler's purpose is to give "most" of the air a place to escape through once the nutrient has evacuated the reservoir. The air will continue up the nutrient line and through the planters, otherwise. When the nutrient returns, it escapes through the screw hole and any other leaks in the caps. It should not siphon back because the air line tip is above the reservoir water line and the "screw valve vent" will be the path of least resistance for the escaping air. To eliminate siphoning completely, I would recommend installing a one way (check) valve between the reservoir and the pump as well as mounting the air pump above the water line. It probably isn't actually necessary, and for all I know, the bubbles may be good for the roots, but I put it in to reduce vibrations and noise from large bubbles making it through the system. It took an awful lot of boredom and anticipated future incompetence (not to mention present day incompetence) to motivate me this far, so don't give up...

prox (author)wiley coyote2008-12-19

gotcha, thank you. one think to consider and you mention it--aerating the resevoir water, the oxygen being good for the roots. i think simply running the air pump hose into the resevoir water should do the trick. what do you think?

wiley coyote (author)prox2008-12-19

Yes it would. You would then have the potential of siphoning and would, therefore have to add the other precautions. I did not do this to avoid the siphoning effect AS WELL AS avoid clogging of an air stone. On that note, like you said, just run the hose in w/o an air stone. I also do not believe that it is necessary for two reasons. 1. The roots are oxygenated by the drain cycle's vacuum effect within the grow medium. 2. The gas transfer takes place at the surface, so other than agitation, it would have little effect.

bwpatton1 (author)2008-12-17

Three questions 1. Do you have any new modules b/c i am interested to build them. 2. Could you post an instructable about a larger system you talked about. I think I would like to line my room with plants :) 3. What kind of nutrients do you use, I am relatively new to hydroponics and could use a few tips.

wiley coyote (author)bwpatton12008-12-17

Some ideas I had that you may want to incorporate.

MosEvo1 001.jpgMosEvo1 002.jpgMosEvo1 003.jpgMosEvo1 004.jpgMosEvo1 005.jpgMosEvo1 006.jpgMosMod 001.jpg
wiley coyote (author)bwpatton12008-12-17

1. Always working on next module, at least in my head. 2. Just posted what was going to be a larger one, but ended up having to cut it twice, so I stuck with my previous scale. It can easily be made as long as you want and go around corners as long as you keep it level and tap it in various places instead of at the end. A "Sucker Stick" if you will, or individual planters, which could be mounted right on the pipe itself or above it. That's how I've envisioned it. That sucker, made into a single solid unit is then stackable up the wall. Somebody make it stop.......... I'd love to see pics when completed. 3. I just use store bought generic, but "complete" liquid hydroponic food. I am just a beginner too and am trying to work out a system prior to next spring. I have found this site to be the BEST source of info and useful links, by far.

awakebyjava (author)2008-12-15

Have to be honest man, you lost me on this one.

wiley coyote (author)awakebyjava2008-12-15

It's the same system, with a vent to help the drain cycle (either the DWC or the screw), and I made so that the individual modules (based on function) are easily removable to allow for further experimentation/expansion on/with/to this unit/plants. The vent and bubbler were added to address the problems and questions that people had with the first unit. It automates the drain cycle for those who were manually releasing the air. The rest is just me learning how to "do it the hard way" as I go. I've since removed the DWC from the system and am venting through a screw. It may be re-introduced in the future. It was more of a Eureka moment and an experiment than anything else.... but it does work. I try to exploit every potential I see, to see if it's practical.

chucky705 (author)2008-12-07

Very nice I built a 14 plant system that uses a cheap pond pump almost the same with the 2l bottles but out of 1 inch pvc with a 30 gallon drum for the resivor sorry about the spelling. yours is very cool

wiley coyote (author)chucky7052008-12-07

Thanks. I'm curious if you had any problems with pump/filter failure. I'm hoping to have several systems set up in the spring for comparison purposes. I'm expanding the reservoir to a 4 footer as we speak, uhhh... type. Dunno how many plants I'll be able to "push". I think the pump should be able to handle "horizontal" expansion with ease, but that's tomorrows module. Off topic.......but......The prototype sprung a leak, at a hot melt glue joint, from moving it yesterday, so I just swapped out the feed tube and "shelf" and plants (top half) in less than 5 minutes, before the pump came on. Worked great w/no leaks or apparent plant stress. Pressure fitting rocks!

tabletopphoto (author)2008-12-06

Have you tried bigger reservoirs like 10 gallons or bigger?


I mean no, 10 litres or bigger is what I meant. See below. The theory is the same though, at least in my head.


Yes, but in a horizontal reservoir made of 4 inch drain pipe. The horizontal nature of the reservoir allows you to tap it anywhere as well as reduce the vertical forces at odds with each other. I thought about larger bottles, but they would have to be shallow to be effective and could only be tapped, easily, in one location. The horizontal reservoirs could then be shelved vertically without the need of expensive water pumps pushing great heights of water. The "shelving" would have to support the weight of the nutrient anyway....... I could go on and on and on... Sorry you asked, eh? One pump..... MANY plants would be my goal I guess.


This link will explain it better. The original size doc is easily readable.

Larger reservoir type thingy

tabletopphoto (author)2008-12-05

It would be so nice to see a video of that working. I just don't get it.


I'll be editing it into the Instructable momentarily. I didn't because of quality, but I managed to get enough detail on the reservoir and DWC unit to be useful. Please try to extrapolate everything that is not in the frame.


Okay. I'm starting to get it. Atleast I believe I am. The dwc is a means to reuse air that would other wise be lost. That is efficient.! 'm not a person who can read plans well and actually need to see the finished product in action to completely understand. I hope you could post a video showing the whole system in action, and for those of us who start out a bit slow, explain what is going on. That would be really useful for pple like me who don't always get something unless you draw us a picture (or vid) Thanks again! I appreciate.


I've added another video of one complete unit, using only the DWC as a vent (and any lack of air tightness), as an edit into the final step. Hope it helps.


I'm kinda suprised that the plant doesn't drown in the DWC since it only gets air durring the cycle. The pump runs how many times a day?


Just set it up, but the plant that's in there is virtually indestructable. It's been sitting in a virtually dry kitty litter container DWC unit for ever, the air stone was not even fully covered and the thing just does not die. These philodendrons are great for testing conditions because they show the signs of stress immediately, but haaaannnnng onnnnn forever and then bounce right back when given favorable conditions. It does look a bit better and there is a rootless piece that broke off in there too. I want to see if the roots will take under these conditions as well. Anyway, I just set it up to go off six times a day for 4 minutes which should drive the flood and drain cycles near perfectly. The plants will be the proof of anything, although I do have a habit of stressing them on purpose.


That is exactly correct. No problem. I'll post something, if I can produce something of useful quality.

nolte919 (author)2008-12-05

I have to say I am quite confused which is a shame because you definitely have something spectacular going on here. First of all I don't see the point of the DWC at all. It looks like a complicated way of letting air escape. And why did you put a hole in the side when you could just have easily put the tubing down the neck that's wide open. And why did you attach tubing to the bottom of the DWC. You mentioned draining and checking the level in there but can't you just look at it and tip it upside down to drain? That tubing at the bottom seems like a leak waiting to happen. And why does the DWC need min and max level markers? If air is just bubbling through it how would the water level ever change? Also, I don't understand if air pressure is rising, pushing up the water in the planters, why would the bubbles be decreasing in the DWC? Wouldn't you get more and more bubbles as the water level in the planters increased? You've clearly put a lot of effort in explaining how you built everything and did a wonderful job I might add. But maybe a little more explanation as to why decided to do some of the things in the first place. I understand the bubbler just not the DWC. I don't mean to be a downer. I actually think this project is brilliant and I don't even know what that DWC is for. After I figure that out I'm sure I'll be blown away.

wiley coyote (author)nolte9192008-12-05

I will try to answer as clearly as possible, in the order you ask and I don't blame you for being confused because you are correct. The DWC is absolutely unnecessary. The screw valve would have been enough, but that actually came later. In my previous designs posted, people were having problems getting the drain cycle working due to air tightness. I did eventually as well, requiring me to solve my own problem. Having a DWC unit available, I tried it out and it worked. It was only an afterthought to add the manifold and screw valve as a simpler fix. All of my designs will be moving toward accessing everything for maintenance/replacement w/o disturbing the rest. With that in mind, I will explain the DWC unit's design when it could've just been an unmodified bottle. The hole in the side is for filling with a funnel and for access to the air stone, separate from the opening that the plant is in. Once the plant is secured, I use polyester fibre for aquarium pumps, you would not be able to access the nutrient w/o the second hole. I attached the tube at the bottom so that I could drain the waste nutrient, again w/o disturbing anything else. It also would allow me to view the nutrient level once the bottle is covered back up to protect the roots from light. Plants also might not like being tipped upside down once past the seedling stage. You are correct, it is a leak waiting to happen. The DWC unit is a Deep Water Culture is unit meant to grow plants, not just bubble air through it. Once the air starts travelling up the tube, it becomes the path of least resistance to the air. The pressure increases until that point, as do the bubbles. At that point, the larger bubbles travelling up the tube MUST be easier. It surprised me too. I would like someone to explain to me why I do the things I do. Looks like were both out of luck on that one...... You're not being a downer. I expected every question you asked. Although I think I've already answered this, I'll be clear. The DWC is actually a separate planter using a different growing technique. It is unnecessary, but was available to be added to the beast to solve a problem so I MacGyvered it in. I also plan on building more completely useless "modules" to add to it whenever I can. For no reason other than, I can and virtually for free. Someone somewhere said something about air powered aquaponics. Hmmm.....