Thread 2933821

Has anyone attempted this or something similar? What did it cost you and what were things you struggled with, did not expect or regretted doing.

https://youtu.be/Q9fjKeYOyqU?si=cHGZODLgmCTa8VKj

I havent built one for canabis only lettuce (had to take it down before i could do more) But its a basic hydroponics system so there are plenty of tutorials on yt.
I recomend haveing a reservoire and a pump to distribute the nutrient solution.

You will have to use a filter for the smell
Id reccomend just buying a grow tent/box whatever its called. The are like 40 bucks on amazon. Unless your going for bigger harvests.
There is a software for a pi (i forgor the name) that manages your nutrients and stuff its open source i think so ezpz
Why are you moveing your light and not just adjusting the brightness?

There is a hydroponics subreddit that is pretty usefull lots of weed growers on there (ikik eww reddit)

Good luck and keep us updated

>>2933900
>Why are you moving your light and not just adjusting the brightness?
By intuition and watching previous documentaries/youtube videos I would assume that keeping the light close within a few inches off the top of the plant would be more energy efficient than having the light on the roof and having it at full blast with tin foil lined walls.

From what ive seen from a few google searches the "grow kits" for sale start around 260 USD at the cheapest such as
https://www.amazon.com/Complete-Reflective-Hydroponics-Growing-Upgraded/dp/B0CKLC6N6T?th=1
And can go up to 1,000 - 3,000 USD for the nicer looking ones. From what ive seen theyre all the same; fan, bucket, goggles, heater, thermometer/humidity sensor and wall outlet timers.

They sell off the shelf components at a markup and most of them lack automation systems. Some like this
https://www.amazon.com/AC-Infinity-Integrated-Controls-Automate/dp/B09WJK8GVB?th=1
Have an app that you can control the light, turn on/off the fan/heater. But no automatic fertilizer feed system, water level regulation, or auto raising/lower of lights just brightness and color control. I would not trust this kind of app with my personal info, its obviously Chinese spyware.

lol that guy is my friend...he made that video and it looks impressive, but it actually requires more manual input and fucking around with than just doing it the normal way

>>2933900
>Why are you moveing your light and not just adjusting the brightness?
Keeping the light source a set distance above the plant as it grows. Too close and it cooks. Too far and you’re not hitting the optimal lumens/ft2 target. Growing pot is a hard science now. There are hardware/software bundles to do everything OP wants, with the possible exception of raising the light.

I could see using a series of electric eyes on an adjustable vertical bracket to control an automatic light elevator. Plant growth obscures a sensor, light raises until a fixed shade uncovers the next emitter/sensor pair. Or you could use a similar setup to control lights at different heights.

if weed is legal where you live this is a complete waste of money and you are retarded, just grow it in your back yard
if weed is illegal where you live the cops are already planning to do a SWAT raid and kill your dog your life is over

>>2934198
>Implying
You are such a wet towel

>>2934183
Could you elaborate
>>2934189
>I could see using a series of electric eyes on an adjustable vertical bracket to control an automatic light elevator. Plant growth obscures a sensor, light raises until a fixed shade uncovers the next emitter/sensor pair.
Yea thats what I was thinking. Using lasers or whisker sensors as feedback for the raising of the lights.

I really dont care about having the system connected to wifi/web server for remote control monitoring. But I would like it to be connected to a computer of sorts or it have a GUI to see all the daily/hourly data collected and be able to fine tune fertilizer dosage, light height/intensity, co2/fan control and theoretically automate those with parameters.


I have an old hp tower and a raspberry pi 5 with 8gb but I think both of those are overkill and maybe an Arduino would be more appropriate.

>>2935130
>I have an old hp tower and a raspberry pi 5 with 8gb but I think both of those are overkill and maybe an Arduino would be more appropriate.
https://autonomylogic.com/
Runs on all three things you mentioned. Looks like a PLC and uses modbus. Even modbus over tcp (Ethernet).
You'll run into problems with the sensors. Most of your sensors will either use voltages that will fry your pi or use I2C bus which is difficult to connect to openplc. There are work arounds, like buying 5v sensors from Arduino stores that specifically don't use I2C bus. Or learn c++ and write some code to connect the two. Personally, I don't have time for the second choice.
also:
https://www.home-assistant.io/
Runs on the tower or the pi. Easily connects to wireless stuff that connects to a hub (like amazon alexa or google assistant). If you want to teach yourself a little bit of YAML, you can connect to a PLC using modbus over tcp. It is already used for turning lights on and off, connecting to certain sensors, and oddly relevant things like raising or lowering blinds. It can generate graphs and you can make buttons you can push on your phone or tablet.

Both have a learning curve. Both start off as frustrating. Both are free.

I would run openplc on the pi and program the arduino to be a PLC slave (through the pi). I'd hook up 5v relays and stop switches to the arduino. I'd put HA on the tower and connect to wireless sensors for temp and humidity. The HA can be also used for data acquisition, graph generation, and remote control.

>>2935180
Ill look into those and do more research on my part, thanks.
>wireless
I intend for everything to be wired and the entire system isolated from my home network. I figure that if im looking at the current status and data collected I might as well be there in person to check up on other things. I dont plan to add cameras, to me its a gimmick if you cant tell what the plant needs or the computer cant tell what the plant needs just by using a webcam.

>>2934191
What about people who want to grow year round in a legal state that gets winters instead of just during growing season?

Well folks I have to say this has been absolutely awful. Ive spent the last 14 hours looking at youtube videos, hardware, conversing with chat gpt and this is what I have been able to sketch out. I knew absolutely nothing and the more I learned about what I was trying to do, how to do it and what I needed to do it, the more it pissed me off. And all ive done is look at at the hardware set up. I havent even looked at the software portion of this stuff, priced out the components or how im going to go about testing the setup before executing. If you have comments, questions and concerns please let me know.

The correct ground/data connections to the arduino/buck/stepper driver are not shown.

>>2935608
>the more it pissed me off
The dunning kruger journey is frustrating. You are learning multiple systems simultaneously. Each system is going to send you into the dunning kruger valley of despair. Like a sore muscle when weight training, frustration is a good sign. Good luck.

One way to simplify is to hold off on the stepper motor unless you're already familiar with them. The idea is simple (and indispensable when making a cnc machine) but the implementation adds considerable complexity.
Manual forward/reverse control with limit switches on each end would get a similar effect but you lose automation. If you get good with programming (I'd use ladder diagram) you could have a series of limit switches but program it to stop when a certain number of sensor switches are pressed. That gives you preset levels. Or you can use an encoder and have the same complexity as the stepper motor. Too many choices. I'd start with manual control.

As a side note, some (not all) opto isolators invert signal. Test for this when you get yours. Many ways to invert the signal back, but confusing if it's unexpected.

>>2935617
>One way to simplify is to hold off on the stepper motor unless you're already familiar with them. The idea is simple (and indispensable when making a cnc machine) but the implementation adds considerable complexity.
>Manual forward/reverse control with limit switches on each end would get a similar effect but you lose automation.
>limit switches
>preset levels
>manual control
Ive already considered these. I intend to automate it no matter how difficult it will be to learn and implement. That was the whole spark for this project, I will start by building a simple test version of the set up with a single or two sensors and then add more and more until I reach my goal. Ill look into the opto isolator.

Between my last post and now I was going to implement a 3d printed SCARA robot that would auto calibrate the ph sensor and then test the water that would be siphoned off into a test box, but I realized pic related would be easier and more robust.

This is what I have so far. I havent sketched out the plumbing + fertilizer/mixers/liquid containers yet. This is the very first time ive drawn a diagram, while I was drawing it I found it to be very messy and easy to get confused, maybe I should look up tutorials on youtube on how to draw up schematics, whats yalls advice.

>>2936223
Do the light level thing last: you have to check the plant fairly often, and a rope/ratchet assembly and ruler to check height is cheap.

in hydroponics,you might as well leave the air pump running 24hr, so it doesn't need be be controlled.

I have yet to find a a water level sensor that isn't crap for this kind of setup, I may have to build my own.

humidity and temp sensor will be one unit combined.

your "schematic" looked good enough, I could understand it.