External Fogponic system design


Hey Guys!

This is my first post here. I am building an external fogponic system. The system takes water and nutrients from a nearby reservoir, creates the fog with an ultrasonic fogger (orange bucket) , and expels it into a nearby growing chamber through the hose. The water that is condensed in the growing chamber can be returned to the reservoir.

To connect the system to openag I used the openag brain_box system by Jake Rye. Even though the box can connect to most sensors used in the Food Computer 2.0, I dont think I will connect all of them. Am not sure how reliable sensors can be in an external system, were they are not located in the growing chamber.

The fan is a Bucket Head dry/wet vacuum. The fan was too strong to work with a single head fogger. I tried lowering the speed down with a fan controller but it is still really strong. I am going to try to use a 12 head fogger. If the fan is still too strong, (which it will probably be) I will need to redesign the system.

I am planning on adding a chiller system to keep the water temperature down. There has been some research that the heat from the ultrasonic fogger increases the water temperature This leads to the droplets taking in fewer nutrients and weakening the roots.

Here is a sketch of the end product, let me know what you guys think! I will keep you posted with any new changes.





I built a smaller version of the fogcomputer. I finally made it work. Here is a picture:

The picture does not show well enough the fog, but is coming out of the hose. The fog is being produced inside the plastic bottle. Here is a pic of the inside of the bottle. The fogger is the metal cylinder with white center you see at the bottom:

We tested the water temperature for several minutes. We noticed that using a fogger increased our water temperature rather fast. In less than one hour, our water temp increased over 3 degrees C. Here is a screenpic of the first measurements.

So far, I have only connected the temperature sensor to the fogger. Here is a closer pic of the wires. In the pic we are starting to connect the fan to the pi. We connected the fan to a 3.3V pin in the pi, which is not recommended but it worked. For future versions we are going to use relays to control the chiller, fogger and fan. The chiller will be turned on or off according to the water temp and the fogger and fan will be controlled by a program. Or at least that is the idea, any suggestions?

We are using the brain_box design but we cant get the GUI working. If anyone could help that would be amazing!

  1. Why are you pursuing a fogponic system? It is my understanding these types of systems should only be used for propagation of cuttings due to the fact that the size droplet created by an ultrasonic mister is too small to carry any nutrient. Have you seen this system used elsewhere, or had any success previously growing a plant?

  2. What do you plan to utilise for a water chiller? Also, why do you think that the temperature of the water is critical, if you plan to turn it into fog anyway (which I would imagine lowers its temperature)?

  3. Have you found any examples of the brain_box being used recently? I will say that I know there are issues right now with the brain software according to this post. Additionally, depending on your goals it may be simpler to utilise our MVP code for the time being if you’re looking to actuate additional devices.


Hey Peter,

I have seen these systems on the internet basically. You are partly right, as of now these systems work better for propagation and cutting, in fact, they have been successfully used in Thailand to grow potato seedlings. My family’s company also grows potato seedlings but in a greenhouse. I want to design a system for that.

Fogponic systems do carry nutrients, tho. Right now there is a lot of speculation about why fogponic systems are not producing the expected results. Some say that the particle diameter produced by fogponics is too small to carry nutrients. There has been correlations between density, which changes with nutrient concentration, and particle size. Here is a link to the scientific paper I got theinfo if you are interested. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3829785/

Another possibility is the small size dropplets do not impinge on roots as well as bigger size dropplets. Both of these issues are because of the small droplets, but droplet size can be controlled by the frequency of the ultrasonic fogger.

We are planning to use a Coleman 40 quart thermolectric cooler to take advantage of te insulation. Now that you mention it, the chiller should not be necesary. The fogger does create heat an increases the water temperature by about 5 degrees C or more tho, and it people on the internet recommend to use a chiller. It will also provide more control on the water temperature, I need to do some more testing and weight the pros and cons of using a chiller, at $82 dls they are quite expensive.

I do not know anything about coding but thanks a lot for all the input. I will look into the MVP code for sure. We are almost finished with the prototype, I will keep you posted!

Thanks again for the input, it was very helpful.



Hi Guys,

Here is an update. We built a better prototype. Here are some pics:

A close up

A view from inside. You can see the thermosensor measuring the fog’s temperature.

This is the exterrnal fogponic system. It has an ultrasonic fogger, a coleman thermoelectric cooler, a bucket head fan and the brain box. The fan is too strong for what we need and we had to add a voltage regulator. We will probably replace it for a dc fan.

@Webb.Peter I tested the fog temperature to see if we really need a chiller as you suggested.
Here are the results. During the first hour the fog keeps rising in temperature until it plataus at 31 C. We need the water chiller unfortunately.

We are going to start growing food as soon as we automate the system with the brain box. To grow food you just connect the hose to any growing chamber you want. We are planning on using this as our growing chamber:

making holes in the squared area and grow the plants there using net cups.

If you guys know any programmers that can help me out with the code that would be great. I am willing to pay but it has to be open sourced here. I will keep you posted!



These are some incredible resources you’ve posted, I am very happy to see that you’re approaching this from a research point of view. I’m curious what @Atom, @adam, & @yusuf.khan.su think about this setup compared to traditional High Pressure Aeroponics.

Great looking growing chamber, did you build that? Is that a PVC material or what are you using?

I would be happy to help you with your code, what are the aspects you’re trying to actuate? I’m not super familiar with the brain_box but am interested to see where you’re at with the software. The MVP right now actuates a 120V plug to control the exhaust fan that is the thermostat, this could be adapted to exhaust a greenhouse (I plan to use the code for that exact purpose). Also, depending on how you build your chiller you could use the exact same software & hardware to cool the water as well as the air (temp sensor, and 120V relay actuation).

I’m curious if it matters how long the fog travels through the tube prior to reaching the roots. Does it condensate heavily on the sides? Also, with HPA you dose in small increments every few minutes, do you plan to do similar or have it always fogging?

I understand that your water temperature has increased, but what I am wondering is the temperature of the mist itself. Something that really hasn’t been researched probably because it is generally the same number with hydroponics is the difference between root temp, and water temp, perhaps water temp is irrelevant and what really matters is just keeping the foggy air cool. In the past, by just putting a 12V fan on a reservoir (this was 120 Gal) it dropped the water temp just by evaporative cooling by 6 degrees consistently as opposed to when it was off. That being said, I’m amazed that you’re seeing a 10C increase in temp from the fogger, that is definitely substantial (and 32C is too hot for hydroponics at least).

Love the commitment to keeping things open source. Let me know how I can help with the code. I definitely think it will meet your needs, and provide a great starting block for someone who really isn’t familiar with software dev to get in and add on some more sensors/relay actuation controls within just a few days.

Aeroponic System Categorization (flex/vertical/horizontal trays)

A small fridge equipped with an external stc 1000 thermostat could be used as an inline fog cooler. Itll be interesting to see how the fog performs as ive tried it multiple times without much success. The issue, apart from heat, is the droplet size, a 5 micron droplet has 1000x less water than a 50 micron droplet so you need 1000x as many. The resulting fog density tends to restrict gas exchange. Ultrasonic droplets tend to be a uniform size so they tend to agglomerate and coalesce on the roots. With a mixture of droplet sizes (5-80 micron) they tend to remain seperate and suspended in the chamber. With the external fogbox approach you should watch for signs of root damage caused by the airstream. typically itll manifest as a discolouration of the root tips, light tan to dark brown, basically its a sign the airflow is air pruning the roots. In soil,only the root tip dies, causing the root to branch back to the stem but without any medium to retreat back into, the aero root has no respite from the airflow and that results in a very stressed plant that doesnt want to grow ;) Heres an idea of the ideal min/max mist/fog density with 5-80 micron droplets. The maximum pic is just after the misting and minimum pic is just before the next misting. Note the chamber floor (15" down) is easily visible through the fog . Ultrasonic fog tends to sink to the lowest point and the “A” in the pics wouldnt be very easy to see, if at all.


This is a polycarbonate chamber with an aluminum frame on the inside. I am glad you liked it.

We are trying to actuate several things. Right now the setup is one 120V plug to control the fogger. We also have a relay connected to the fan and another one to the chiller. The idea is to use the relay connected to the plug as a timer that turns the fogger on and off every minute. The relay connected to the chiller turns on and off depending on the thermsensor readings to keep the water temp between 25-27 C.
From what I understand, modifying the mvp could work great.

When I did the experiment I hanged the sensor, as shown in the pic, so that it was not in contact with water and could measure mist temperature. The coleman has a chiller and a fan so I am going to play a bit with that and experiment with evaporative cooling too, it would be great to get rid of the chiller.

Thanks a lot with offering help with the code!! So far I have downloaded the brain_box api to the pi but the gui does not seem to be working. I can read the temp sensor, which is connected directly to the gpio, and I get connection for both atlas sensors, when I run the i2c test I get no results back. Instead of getting 4d like in the brain_box example below, It only shows lines.

Are you in Boston? If you help me with the code I am buying the beers.


Atom thanks for your tips we will sure take that into account when designing this. Airflow is very tricky in a system like this. Apart from root damage, we need to consider condensation, we are noticing that water condensates in the hose that connects the fogger to the chamber. We were thinking on cranking up the airflow to counter that, but it looks like a bad option.

Thanks for the pics. So if I understand correctly, a mixture of droplet sizes is more desireable?


Ultrasonics generate a very narrow range of droplet size, typically 3-5 micron which creates more of a pea soup fog than a light haze. The maximum fog density pic above contains less than 1ml of water in a 100L chamber. The ultrasonic foggers i have deliver/convert around 600ml of water per hour but running it for 1 minute (10ml at 6 seconds per ml) in the same 100L chamber doesnt produce the same result as the pic. It creates a thin layer of very dense white fog on the chamber floor that you cant see through at all.
The condensate in the tube shouldnt be a major issue. I`d run the delivery tube uphill so the fog can reach the root chamber and fit a small drain tube at the lowest point in the delivery tube to remove the condensate which cant physically flow uphill :wink:


You seem to have a lot of knowledge on the topic. Do you have any recommendations if I want to still use an external fog system? Maybe run the fogger in longer intervals to produce a 1ml of water per 100L instead of 10ml.

Nasa says that the most effective system is aeroponic systems run at on and off intervals. This may sound stupid but could you replicate those intervals by building up pressure and releasing that pressurized fog/air mix in intervals? Like puffing smoke rings.


The on/off intervals are simply the means for creating the correct environment for the roots to be suspended in :wink: The environment should provide a constant and perfect balance of water, nutrient and oxygen. Consider the mist environment created by misting 1/2 gallon every 30 mins or 1/4 gal every 15mins. Both deliver the same amount of water vs time but results in soaking wet non aeroponic roots. Now imagine misting 1/60th gal every minute or 1/120th gal every 30 seconds. The amount of water delivered in each misting isnt sufficient to wet anything but the roots suspended in this constant mist environment gradually become damp over time, like a tissue in a sauna :wink: Its hard to imagine 10 seconds more or less interval would make much of a difference, but, with this type of aeroponics its cumulative and over a period of hours it can equate to roots that are too wet or too dry. Spraying the roots directly with mist and/or having the roots hanging in air isnt the same but its a lot easier.
I run two main types of aeroponics, high pressure aero which offers a reasonable level of control, it consists of pump, accumulator, pressure reducer, solenoids nozzles and timer,
The second type is siphon fed air atomised aero, it consists of an air compressor, twin fluid air atomising nozzles, solenoids and timer. It offers much greater control over the mist delivery including droplet size.
The third type is pressure fed air atomised aero, it provides individual control of both the air and water.This type of nozzle can deliver individual mist pulses as short as 0.3 seconds.
My most advanced aero system wouldnt look out of place at NASA, they havent built anything similar that i`m aware of :wink:


0.3 seconds? Now am curious, I really want to see that system.

Thanks a lot for the clarification Atom, you have been very helpful. Am set on using external fogponics because it’s more scalable and customizable than complex aeroponic systems, but its limitations are becoming clear. I will make some experiments and keep you informed!


Looking at the numbers (disregarding droplet size and fog loss due to condensation) a single ultrasonic fogger running constantly should be ideal for an 1100L root chamber. For the later growth stages youd need to use 2 foggers. Finding enough space could be an issue but if you have room for 2 chambers, one twice the volume of the other, the results may provide some useful insights to take you forward. The most important thing is not to overfog which will usually show as wilting leaves but an upright stem. When underfogging the leaves and stem will wilt. I recommend using cuttings with a goodly amount of roots for your initial test subjects (aka victims) as a minimum amount of root area is required to intercept sufficient fog to keep the plant upright. Heres one of my ongoing aero experiments, 100L chamber with less than 50ml/hr providing the environment. This young pak choi (in transition) is doing ok for now but i`m sailing close to the wind with these settings so anything could happen, Thats what makes experiments so interesting :wink:


Those roots look amazing! Thanks for the tips btw. I am going to start growing soon. I want to try potato seedlings, a fogponic system should be good for that. I hope I can get some, if not I will try lettuce.

I want to make an experiment to test the nutrient (solute) concentration of the fog. I was thinking on fogging a chilled close container, condense the fog, weight the solution, evaporate the water and weight the solute. The problem is that much of the nutrients are liquid, would the experiment work or should I use solid nutrients?


Hi Juan
I dont think you would glean much useful information. The amount left on the walls of your container could be enough to skew the results by a lot. The nutrient concentration in individual droplets will be different to the condensate which will simply reflect the average of all the droplets (less losses) when recombined. I make my own nutrients and use reverse osmosis water so i nkow what is in the nutrient. I can compare the inlet and outlet ppm to get an idea of what adjustments to make and when. I know what the flow going in and with drain to waste i can measure the amount of run off, which, along with the visual inspection of the root wetness and structure gives an indication of what adjustments to make to the flow rate. One thing that becomes immediately clear in a controlled indoor environment is the roots dont need as much mist during the dark period,
Potatoes will need a good amount of chamber depth, i was hoping to experiment with potatoes myself this season and built a system especially for it but time got away as usual. The system consists of standard hpa components (pump, accumulator solenoid nozzles and timer) with 90L/24gal plastic dustbins (trashcans) for the chambers… buried in the ground to keep the temperature in check. I have two chambers for now but the system can handle upto eight.


I am doing the experiment any ways, but from what I can see so far you are right @atom. Here is a pic of my setup.

I had to change the setup several times because the system has been running for two days and there is a lot of fog but almost no water, so I have not been able to measure the ppm. I will run it for a couple more days and post results!


Hi Juan
Its good to see someone doing an experiment, regardless of the outcome :wink:


Thanks for the support. The experiment worked! I should have taken a control sample of the solution and use distilled water instead of tap water, but the results are pretty good still. Tap water had a ppm of 397, which I found really high. The solution’s initial ppm was 496. The test was run for two days until we had over 50ml of fogged solution. After the two days the ppm of the pre-fogged solution (taken from the reservoir bucket with the fogger, orange lid) was 508, and the ppm of the fogged solution (taken from the collection bucket, black lid) was 506.

I have a question @atom, I was thinking in putting the fogger on a timer, 4 minutes 15 seconds off, or something like that. But would it work better if the fogger turns on and off depending on humidity levels? Basically, the humidity sensor would control the fogger instead of a timer.


So to clarify: Fogponics works??? :scream:

My next question would be what happens at a PPM of more like 1400-2000. That being said, maybe you would never want that high because the droplets are so small?

If anyone out there who is reading this has ever successfully grown a full-size plant using fogponics I’d love to hear what PPM nutrient you used.

I like the idea of using a humidity sensor to control your on off. I’m not sure if it would work though, as it may read 100% constantly or be confused by condensation. I wonder though if any sort of laser distance sensor would work. Perhaps you could use that to measure the density of the fog and then based upon that turn the solenoid on or off. @webbhm