Invitation/project suggestion: An automated potted plant

ESPHome
#21

Well for me it’s good old Creality Ender 3 - the entry printer that never got replaced.

I’ve uploaded the files here: Flaura STLs - Google Drive

Print it with support, but on the “touching buildplate only”-setting and a support density of 15%.
I used 40% general infill, but I don’t think it matters that much.

In one of my prints the hose connector inside the pot (the outlet that goes to top of the soil) was printed as a super-thin shell that broke of opun touch. It was a slicer/printer error of some sort, not the model, but be sure to check it once the printer gets to it so you can abort if it should fail for you too.

#22

Will print the STLs tomorrow. I created a Arduino-Sketch to test the hardware. It works for me.
https://github.com/crabucate/Flaura

#23

I have to say, the whole point of having plants is to decide yourself when to water. Where’s the fun? If it’s a commercial veg growing thing then by all means automate it. I have been designing my own plant " remote watering" by wifi over esp. I press a switch to water the plants the next time the esp comes out of deep sleep it waters for the desired time ( this is being worked out). I see you mentioned about it not stopping watering. This happens when the water level of the pump is above the end of the watering tube. It’s a syphone effect. :grin:

#24

Hi Spiro,
Welcome to the party :slight_smile:
I think for many people, myself included, the point of having a plant is that they are nice to look at and that fresh herbs are a nice addition to the kitchen, not that the watering duty brings any particular joy.
No, the issue is that the transistor that drives the pump appears to remain on even when the gate is set to low. Not sure why. Maybe I just have to pull it down. The outlet of the pump is actually above the water level of the reservoir in the design, and also, haven’t made it to actually printing the thing. I just work with the electronics for now.

#25

Ok. I think I’ll try with the real lolin32 and a newly soldered print and see what happens. Maybe I just have a bad solder somwhere.

#26

I use a 5V relay to turn on the 5v pump . All powered by an 18650 cell and a small solar panel recharging the battery. Use a ESP12F as found them good with deep sleep and low power usage. Relay uses most power even if not used to turn on pump. Used the relay as the relay is what I had lying about. Will consider using a transistor. Is there any problems with voltage spike when turning off the pump. Do you need a snubber/flyback diode to protect the ESP board? The relay module has it built in.

plantbattery
plantbattery774×470 46.1 KB

It’s all for just a few potted plants outside that don’t respond as well to tap water. Even a dull day keeps the cell charged up.

#27

Yes, you will need a diode if you use a transistor to turn on the power. The upside is that you can also do pwm if you use a mosfet and they only use a fraction of the power that a relay does - but it proberbly isn’t needed in your case.

Those tiny boards are nice, but you run out of pins so quickly.

#28

Do you think you could make a github repo for this project. At the moment it is a mess. Or I can make one.

#29

Have you considered putting the battery outside the pot? Not as neat but could be much safer. It’s just the potential melt down of the pot if water gets into a fully charged lithium cell. The idea is to keep them inside your house. This could affect your home insurance. There are some nice youtube videos of 18650s burning.

#30

I actually think it would be more exposed like that as you have to refill the pot occasionally. It would also potentially take the full blow if the pot were to fall.

You have to expose the lithium inside the battery and get it in contact with plenty of water to get a fire going. It takes some brute force to expose the lithium of a battery, it’s highly unlikely to happen by accident.

Water could theoretically short the battery which would cause it to overheat (but generally fires don’t happen unless you short the battery for a very long time with very good conductors and are even rare then). It’s hard to imagine a scenario where there would be enough water to cause a short while not enough to cool the battery, and for enough time to not evaporate due to the heat created wherever the short is. All the electronic components in the design are elevated above the surface that pot stands on, so if there was a leak, the water would simply flow on onto the surface beneath and spill onto the floor.

1 Like
#31

Ok, you have thought of battery safety. Just thinking about the plants themselves. When watering generally you have to let the plant get a bit dry (will depend on plant type) and then give a good watering. Every plant you look at for care online says water well but allow good drainage. Roots tend to rot if left in a puddle. Usually they say just feel about 2cm below the surface to see if dry. I think that will be easy with some water sensors. What’s the drainage like and what’s the capacity of the tank? Is there possibility for some of the water draining back into the tank from the plant pot?

#32

Well the imagined logic of the system is to water a little, sleep a little, water again, and so on until a certain threshold is reached which will trigger a longer sleeping period. The roots should never be left in a puddle as you say. The aim is to get the plant perfectly moist by only adding a little water and let it soak in so a good measure of the moisture can be obtained.

If your particular plant wants it dry before water, you could simply add different thresholds for stopping and starting - should be fairly easy.

#33

Cool project, I was attempting something similar.

However, I had to shelf it because my captive sensor where all over the place with there reading. And I did not want to throw good money after bad money on any more sensors.

He has a great video on why the sensors are bad. And what to look for in a good sensor.

Many thanks for the links.

#34

There’s a link above for the sensor that I bought (based on the youtube walkthrough of sensors you mention). The ones I got works.

#35

I missed that link. but thanks.

I went with Capacitive soil moisture sensor module Corrosion Resistant wide voltage wire Analog Capacitive Soil Moisture Sensor V1.2|Sensors| - AliExpress .As the pictures look correct.
I purchase another 20. ( to go along with the 35 I have all ready) Wish me luck.

#36

I’ve made a git repository that contains my work on the code. It is nowhere near working condition, but there may be stuff in there that can be used. You can find it here:

1 Like
#37

So how did the print go?

#38

Are you putting the 3d files there too?

#39

Sorry. Forgot to push a button. The files are there now.

#40

Sorry, I didn’t print it. I don’t know what to do with a second pot. Maybe someone else can try?