Lost in the mist, or Twisted journeys in the AAA world (strawberry version)

Hi, guys!

I’ve been reading and commenting in @Larry3215 Adventures in Aero threads for a while growing my own desire to build one of such systems. The only difference is that I’m going to do strawberry, not weed.

Just to catch up your attention here is the video on YouTube and a picture:

It contains:

  1. A quiet oil-free compressor (160 liters per minute on input, 54 dBa)
  2. The main pressure regulator with 5 micron filter and dryer
  3. A nutrition tank made from a plastic beer keg, pre-pressurized by one of the air line leafs with its own pressure regulator
  4. Air atomizing line with its own pressure regulator,
  5. An only one Chinese nozzle so far with…
    6, 7. …two normally closed 12V electric valves (one per nutrition and air supply lines),
  6. A hand made controller
  7. A root chamber 1250 x 700 x 500 mm (~ 49" x 28" x 20").
  8. 240W LED lights (Samsung lm301b 3000K + 660 nm) from China.

The first pieces I’ve got made this “Young grower kit”:

A compressor, a beer keg with additional parts, a few pressure regulators, nozzles and some tubing.

The root chamber become the most difficult part of the build. I wanted to have a big enough one to be covered by a light I have and to test different plants density. It took around two weeks to build it.

Its made from aluminium shapes and PVC film 0.4 mm for furniture faces.

If you are interested in build pics here they are. Actually, A LOT of them. May be even too many.

Everything has started from a few metal lines:

First of all - the storage stand:

Looks crooked, right? I’ll fix it later.

Then I started making a frame. So cut the metal down to parts. Every piece got its own mark:

Metal work:

Then time for inner space came.

Some math:

Cutting out:

Trying to whip parts in, but no luck - this PVC film doesn’t like glue:

So, hello rivetter and rivets again!

Broke my beloved automated rivetter in the process:

But at the end it got some shape:

Not without problems…


The next is a Montessori board for adults:

And, finally, preliminary results!

The controller consist an esp32 devkit v1 board, a step-down power converter (12V to 6V) to power the esp32 and a solid state relay to open valves. It connects to the home wifi and then to mqtt server and waits for commands. The controller and valves are powered by some local brand 220V to 12V@5A LED power adapter I bought via local BBS.

I’ve written a separate small web-based control panel which works outside of the controller (currently on my laptop. It will be rewritten, of course, to be more high-level: like to allow to set timers and to control sensors (which are not available yet), but the controller will work independently. Now it works independently too, but requires external command to start and stop misting. Anyway, it is for test only and to be improved in the future.

I have a lot to do with pressure configuration as well: I don’t like the mist at the moment. I might be wrong but the mist what is going up are too dry.

At the same time I need to figure out how to make the film light-proof: it appears that it is not.

But in overall it looks very promising and I’m really excited!

Hope it wasn’t very boring (just a little). If you have any comments or questions or need any details: please feel free to ask. And any advice pretty much appreciated!

Some discussions and QA are in the @Larry3215’s thread I’ve linked earlier and in an another one as well so check them out please.

P.S. As usually sorry for my English.
P.P.S. And special “Thank you” for the @Larry3215, it is him and his invaluable AAA threads hooked me here. Very appreciate!


I love the title!

I have questions, but Im too stoned so Im off to bed.

Oh - yes on the build pics. You can never have too many pics :wink:

Dont forget to tag @Northern_Loki about moving some of those other posts over here if you want. Im fine leaving them where they are, so its up to you and how you want to do things.


Have changed the nozzle air cap to an another one - external mix type. The mist quality seems to be much better.
Here is an example: now it at least doesn’t look dry.

But I need to do something with precision pressure regulators: seems they drawn air all the time. They are relief type regulators, but I was under impression that they must seep only when downstream pressure are higher than set by regulator. And I hear constant seeping even though they sits right before closed valves. Don’t understand why is this happening. Receiver pressure drops too fast: I would say about 2 - 2,5 bars per 30 minutes.


That does look better to me.

What volume and dimentions is the root chamber?

Have you decided on a timing schedule yet as far as ON/OFF times? How finely can you control both?

Have you checked hang times yet?

Id recommend using your cell phone and its LED light. Seal up the chamber as much as possible so you have zero air leaks and light leaks. Then let the camera shoot down through a small hole. In the side or top should work fine. Let the timer work through a few ON/OFF cycles to be sure the chamber is full of mist, then start recording and see what you get.

It would be good to take pics of the nozzle itself to check for spits and spurts when the solenoids kick on/off.

Oh - have you tried to do any flow rate measurements on air or nutes?

Yeah, that doesnt sound right to me. They should only bleed off any excess air in the lines that is above the set pressure - then stop. They are going to use way too much air.


The root chamber is 1240 x 700 x 500 mm (~ 49" x 28" x 20"), but it is not exactly a right parallelepiped as I’ve tried to make a slope (this is why I needed to do some calculations: to correctly cut a film to achieve this goal), so its height is 480 from one side and an additional triangle from the other side. So its slightly more than 416 L. I can calculate an exact volume but not sure it worth the hassle. :slight_smile:

As I’m using a micro-controller and write code for it by myself I can control it with literally any precision (from microseconds to years). But I didn’t decide timings yet. I don’t fully understand how to check if the mist quality good enough (we talked about it before, I remember) and without it it has no sense to measure hang times and without hang time I can’t choose on\off timings.

I didn’t do any rate measurements as well. It’s a good idea to start, I’ll do it right now. Don’t have precise scales but I believe it’s ok to just mist for a long enough time range and check changes with 1 gram precision scales. And with air, I’m don’t have any precise tools to measure air flow or even pressure, just some analog gauges, but I think I can try the same way: to fully charge the air receiver, shut off the compressor, get full pressure approximation value then start misting and check when pressure on outer regulators starts to sink. At this very moment I need to stop misting and check the air receiver pressure. As I know receiver volume, I can calculate amount of compressed air at the beginning and at the end and the difference, and with time amount - flow rates.

Seems they need air for their own purposes: I’m using SMC IR2000 regulators and their specs has this option:

Air consumption (At supply pressure of 1.0 MPa): 4.4 l/min or less.

As I’ve set air supply line pressure at an around 3 bar I believe this bleeding amount should be much less than 4.4 liters per minute. But still it is very inconvenient: now fully loaded receiver loses around 2 bars in 23-25 minutes, so it is about 1.5 - 2 liters per minute. I’m thinking to put an another solenoid before these precision regs at the moment, but it will give some gap on the nozzle air supply line (nutes are pre-pressurized and nutes tank has non-return valve for air so it keeps its pressure all the time) so it is necessary to measure consequences. In theory I can open this new main valve first and after a little while open the closes ones. Have sense, actually, I need to think about it. Probably I don’t even need an additional relay and code changes, just RC circuit should be enough.


Thats close enough.

Mist quality is hard to judge or measure. Hang time is one pretty good one though. Here is how I would do it.
You dont know flow rates, so you cant calculate how long an ON cycle will “fill the chamber”. So you kind of need to work backwards. You want a hang time of at least one minute to two minutes. Thats going to be a judgment call also.

My chamber is about 210 liters, and Im currently running three nozzles for 1.1 seconds ON time. That more than fills the chamber.

With a single nozzle, I would start with 1 second of ON time x 60 seconds off time. Make a short video of the cycle as I described above.

Then repeat with 2 seconds ON time and compare the two videos. Make additional videos with different ON, OFF times - making adjustments depending on the results of the first two. This will be tedious, but its the only way I know to get in the ball park. See how much hang time you can get. The ideal setting in my system seems to be roughly 1/60. One second ON time to 60 seconds of OFF time. That is also factoring in the dry areas that Im trying to get more mist too. With better nozzle placement and or better root ball placement, I think a better ratio would be closer to 1/100 - but I have not made it that far yet.

Be sure to post them to Youtube so I can see what they look like.

I probably wouldnt bother to do the measurements your thinking about, but they will work. I am lazy and I would just wait to see how long it took to empty the rez, and the air tank. Then I would convert that to cycles per minute or hw ever you measure it. Of course, that will only work after you have a set ON/OFF time.

Do you have a supply of sacrificial seeds to stick in your torture chamber? Have you thought about how your going to start the seeds in the system? :wink:

1 Like

Almost forgot - is there a way to turn that off of disable it? A screw or something you can adjust?

The idea of using an additional solenoid should work, but that shouldn’t be necessary.

1 Like

The current problem I’m trying to solve is that even if I see the mist particles going out of the upper part of the chamber, its lid actually dry. If I put my arm inside low enough it is getting wet in a proper way (with tiny drops all over hairs) but if I keep it near the top then it stays totally dry. Looks like the big enough drops just can’t go up - which is weird. Or there is no enough drops of the size I need.


No, it looks like a design feature.


I had the same exact problem when I was shooting down from the top of the chamber. I just could not get enough droplets to the very top part of the chamber that were “wet” enough for the roots. Even with very long ON times, it stayed too dry.

Thats going to be hard to judge until you have some roots in there. The roots will let you know if they are too dry or not pretty fast. If its too dry, they will start to air prune in just a few hours time. If too wet, they will stay smooth and grow mostly down.

Unless you have some fancy ($$$) equipment, I dont know of any other way to accurately measure droplet sizes.

You can look at hang times, but if your nozzles are producing lots of very small droplets - sub 20 microns say - they will fill the chamber and stay floating long after the ones you are interested in have fallen to the floor.

If you look back at that video I did where the camera was looking straight up when the nozzle fired, its pretty clear the large droplets just fall within seconds. That leaves the smaller ones to “hang” around. They look good, but Im starting to think they are almost all too small to be useful.

Someone posted some spreadsheets earlier showing that the major portion - volume - of every ON cycle was composed of the larger droplets while the smaller ones made up only a small volume relative to the larger droplets. This makes sense because of the cube square law. If you double the diameter of a droplet, it is 8 times the volume of water.

In my situation, its becoming clear that I need to directly spry the roots to at least some degree. So far the best performance was when the nozzles were down low, shooting UP almost but not quite at the net pots. That setup made the best looking roots but it had a major problem. The roots went straight for the nozzles and covered them up before the grow was even 1/2 over. That forced me to move the nozzles UP to the top of the chamber. The roots still went for the nozzles, but it was at least a little slower progression.

I have no idea how big strawberry roots get, so you may have a lot longer to go or maybe its no problem at all. It will depend on how deep your chamber is compared to how long the roots get.

The other thing to remember is that droplets dont “bounce” off a surface. So any droplets hitting the sides or end of your chamber will just stick there rather than bouncing up to the top. Some of them will be carried by the air as it flows up and circulates in the chamber, but the largest ones will not make the curve. For the most part, only the smaller ones will go UP and float around.

You may just need to move the nozzle up a ways on the end board. Maybe try the middle or near the top to see what differences there are. You might also try pointing the nozzle up at an angle, but that might create some areas that are wetter than others along the lid.

One way I “measured” the number of droplets hitting the net pot was by using cardboard tubes from paper towels/toilet paper rolls as indicators. Put a cardboard tube in the net pot sticking down the same depth. Then run a few ON/OFF cycles and see how wet it gets and which sides are dry, etc.


It occurs to me that a lot depends on how long your strawberry roots get. You may be better off having the nozzle(s) on the floor of the chamber shooting up. Or you may need more than one nozzle spread out along the length of your chamber. Or you may need to change the shape to form a deeper chamber. OR?

I would strongly suggest you do a test run. Put a row of seeds/shoots/ or what ever works in the system and let it run for a few days. You will very quickly be able to see where changes are needed.

This way of proceeding isnt very satisfying if you are like me and like to calculate every detail based on precise measurements. Unfortunately, Ive found that the empirical method tends to work better when dealing with plants. After all, botany isnt a “real science” according to the movie “The Martian” :wink:


Been away for a while, was doing some experiments with the controller: made some improvements in its software and control panel, made a little video about it. Nothing special, everything exciting is inside: https://www.youtube.com/watch?v=Kc252TzBdRg . And still a lot to redesign and refactor.

But for now it finally can more or less reliably control solenoids with precise timings and I understand how can I extend its commands if I want to.

And - finally - returning to experiments with nozzles and misting - on a new level!

So new videos soon!


Two mist quality videos. Pretty boring, I believe.

  1. https://www.youtube.com/watch?v=2v8-0jpBjQU
  2. https://www.youtube.com/watch?v=GpxREO92IFE - these one with hang test

(videos just uploaded, so I their resolutions option set going to extend a little bit later)

Mist hangs pretty long time so I believe droplets too small. :frowning: Don’t understand how to make them bigger yet. I’ve tried to increase water pressure and reduce air pressure but haven’t noticed any difference so far.


(pretending to be a spammer)

This nozzle looks much more promising. It’s an external mix air cap with the same water cap: https://youtu.be/_DATlQJ-UYw
Long video with test.
But I’ve tried to put a paper napkin inside a lip hole and it stays absolutely dry even after a few cycles 5 sec on - 60 sec off.

And it’s obvious that the water side solenoid is very slow one: water continues to go out under pressure hundreds of milliseconds after air solenoid closed.

I’m going to try to use the same solenoid for the water line soon, just need to go to the store and buy a 1/4" female-female converter to connect it. Hope it is water proof: it eats 4 amps, so if it isn’t water proof I’ll have an unexpected surprise. :joy:


Yep, this is all about to get to know :slight_smile: this is why I’ve made the chamber so deep - because I don’t know root length. According to the nozzles number: I recall Atomizer post in one of the forums where he told that was able to fill 700L chamber with just one nozzle. And actually I believe it is possible - the only thing is a correct nozzle or at least correct nozzle settings (pressure, feeding type etc).

It is late already so I had to stop experiments for tonight. And I realized that I’m using 5 amps power supply unit and it is not enough for two 4 amps solenoids. I have a couple of 200 Amps PSUs but they are hot swap server ones, so have open contacts and I’ve already made a fountain once forgetting to shut the handle on the tank - it has pressure all the time :slight_smile:

At the same time I’m thinking to try siphon feed. But I would prefer to have pressurized option, it is more convenient.

Honestly I do. I prefer to prepare myself in advance and hate having things I did not choose by myself - like these nozzles, for example; if you remember I’ve got not the ones I chose. And the worst thing is that I don’t really know what exactly did I get :smiley:


It appears that the solenoid is good, and this is just some volume of water in 5-7 cm long tube after the solenoid. Live and learn.

Still trying to figure out what to do with mist: should I search for another nozzle? What type: I mean, what shape? And where to place it?

I’ve started to test different air\water pressures and measure flow rates.


Yeah, one of the lessons from doing HPA is that the water solenoids need to be as close to the nozzle as possible. Other wise, its like you found out - the water continues to flow after the solenoid closes because the pressure built up in the tubing will cause it to have a run-on flow after the solenoid closes. With a siphon feed that wont be an issue.

I agree - your mist seems to be too fin/dry. How low can you go with the air pressure? Lowering the air pressure and/or raising the water pressure is usually the way to increase droplet sizes.

One thing about your nozzles - the mist seems to be more uniform as far as droplet sizes. My nozzles seem to have a wider range with more of them being too big as well as having some too small. if you can get those dialed in for larger droplet sizes, I think they will work pretty well.

I thought of a test you might do to check for “wetness” and to see how well the droplets are being distributed through the chamber.

First - pick an ON/OFF time you think you want to look at. Then dry out the entire inside of the chamber so the floor, walls, ceiling are all dry. Then run a few cycles only - maybe just 2 or three to start - and check the floor, ends, walls, etc to see where it is getting wet and where it is staying dry.

Then try a longer ON time and maybe lower/higher air pressures etc.

This will be tedious drying out the chamber each time, but I think it could be very revealing. I suspect you will find that the droplets do NOT want to turn the corner at the far end of the chamber from the nozzle. At least the larger droplets wont turn the corner nearly as well.

It sucks not knowing what you will actually get when you order the cheap nozzles. That is exactly what I ran into - I got nozzles that didnt work at all as they were described. I would worry that even if you found one you liked, would you get the same thing next time you ordered? Or would I get the same thing if I ordered some? Very frustrating trying to do this kind of stuff on a budget.

Unfortunately, the name brand versions of the nozzles you are using are very $$$ in comparison.

Im going to pst this in my thread too. I am about to give up on trying to “fill the chamber with mist” in my setups. It just has not worked using ANY of the nozzles/positions/pressures/feeds/etc, etc that I have tried - HPA or AAA. I am seriously doubting some of the things atomizer has been proposing as gospel. It may be that you just cant do it the way he says unless you are using a $200 nozzle as he has sometimes claimed. Or - he is just leading us on a wild goose chase down the wrong path trying to reach an impossible goal. It occurs to me that he has never posted root pics of his setups that I have been able to find - other than one shot of a single root hair all by itself. Have you seen anything more than that from him? As far as Im concerned, if there are no pics, it didnt happen.

Im going to go back through my earlier grows looking at what worked best. Im already pretty sure that shooting UP from the bottom of the chamber works better than anything else, but I want to be sure.


You know, actually yesterday I’ve found his videos - not too many, but with there are some comments after them. I’ve written him a couple of questions in one of the videos, let’s see will he answer or not. The video shows misting and some roots at the end. But I think I’ve already seen these before in his posts, both video and roots.

I believe the idea of filling a chamber with mist is having sense. At least it has some logic under it. The only question is how to make it happen. I believe it is impossible with cone shapes or partially possible with hollow extra wide cone. But if it is possible to put 1 - 2 flat fan in parallel to the bottom at the bottom or in the middle but vertically orientated and near the walls - then it might work.

You know, actually I have an idea: I’ve found some Sparying systems air\water caps in eBay for pretty good prices, and I can use them with current nozzles basements: it has 1/8" fluid cap mounting hole - looks like the same as in other manufacturers nozzles. And as these nozzles are copies of the Spraying systems ones, I believe it is possible to buy fluid\water caps only and replace the ones I already have. A new Chinese nozzle costs $13-15 (I’ve paid more, but later found such prices on eBay) and even more on aliexpress. So one can buy a couple of chinese nozzles, replace caps with genuine ones and have a $150-200 nozzle for much less. And I even found an antigrip nozzle for $50-60 IIRC with the Best offer button.


Remember - the first thing that you need is a nozzle that produces the correct droplet sizes.

No placement details or angles, or number of nozzles, or shape of the chamber will compensate for nozzles that dont make the right size droplets. I can attest that nothing will help if you dont have the correct droplet sizes being delivered. Ive tried all sorts of variations trying to make up for that one critical detail.

And thats the one detail he never ever gives out.

1 Like

Yep, for sure. This is the reason I’m looking into another options.

For example, there are a couple lots on eBay with 4 x new 2850 air caps (1) and 3 x [seems to be used] 73335 fluid caps (2) which are parts of a Spraying systems SUN13 AA nozzles, and it is $12.99 and $15.99 respectively with free delivery. There are a number of new air\fluid caps on eBay as well.

SUN13 specs:

It gives like $9 for a SUN13 nozzle if you have a base (which I have) or plus $9 + $12-15 if you haven’t (for an any Chinese nozzle to disassemble and get a base only). One thing makes me precautious: this nozzle has pretty low water flow rates: if we believe Atom in flow raters (like from 2ml per 100L of volume) then I need from 8+ ml to 24+ ml per misting session, and I don’t want it to be too long, so 1-2 seconds, which gives me water flow rate beginning from 29 liters per hour.

The problem is that specs has only general information on drop sizes like this:

I’m going to write to them and ask for the detailed technical specs after week-end, both to the US and to the local Spraying Systems reps. Or may be will be better to call and ask. I believe this might be more reliable than any Internet-gathered empiric data.

Meanwhile, I’m doing some measurements of the flowrates of my nozzles.