Small Scale $10 automated watering

[Ca++]

Has anyone played with this?

https://www.aliexpress.com/item/1005004882851430.html

You can add 4 more sensor/pump sets to this board, for another $10

With pricing like this, you could probably buy twice. Using both at the same time. So if one failed, you wouldn't loose a plant.

My stumbling point is the fancy board. Is it realistic to be getting a cpu based controller for such low pricing. More importantly, where are the buttons? I guess we plug in a PC, so where is the software to study it's abilities.

It seems very good. I made stuff like this with metal probes, but they fail rapidly. This looks like it could work though. The relay is quite big, so pump choice isn't limited to this provided one.


It seems so good... but where is the software? I have no idea how to use this board, or how much I need to know.

 

[Creeperpark]

I don't trust those little devices to manage my watering management needs. If you get everything working and running automatically, post a photo so I can see what you are talking about. The less simple it is, the less I want to use it, but that doesn't mean it's not good. You keep us posted on those technical things Ca++ because you are smart and know a lot about all the new stuff. Thanks for the post friend.

 

[Ca++]

@xv25ddd
@ButterflyEffect
@unnamedmike

@Creeperpark Pics online show the board is an Arduino Revision 3. A quick search found 3 users here that use Arduino. Mike seems to have serious automation going on, with an interest in soil moisture as an up coming project (6 months ago)
I have put a shout out above, in the hope they pass by one day.

These capacitive soil moisture sensors are a few years old now. I actually thought I ordered some. I had a project centered around a fluid level controller. The type that uses floats to go from A to B when a tank gets low, and needs filling to the top. I used the A to B logic as dry to wet. Each float input, swapped for a moisture sensor. It worked, but nothing lasts in our root zones. So.. I guess it didn't work actually. Good try though

These 'capacitve soil moisture' probes are a big hit in search engines. Mostly connected the Arduino based control though. These PC boards are indeed, $2.50
Amazing really. A processor board costing less than the switches and relays I play with. The $10 kit really does have all that stuff in it. Though the pumps are a bit of a let down, as the wire isn't long enough for more than a buckets depth. While they also limit pot height to about 50cm, or they just won't get the water up there.
The relay boards are good. I have used quite a few, in situations where everything mattered.
The probes are stamped out of a big sheet it seems. The edges are sometimes prone to letting water in. It's wise to paint the edges with a sealant, to finish the job the factory should of done better.

The only thing I'm unsure of, is what good is a controller, if they are not showing us windows based software to set it up. I only speak English, with some use of binary programming, to get more channels on my CB radio. It's just not going to cut it

 

[GenghisKush]

@xv25ddd
@ButterflyEffect
@unnamedmike

@Creeperpark Pics online show the board is an Arduino Revision 3. A quick search found 3 users here that use Arduino. Mike seems to have serious automation going on, with an interest in soil moisture as an up coming project (6 months ago)
I have put a shout out above, in the hope they pass by one day.

These capacitive soil moisture sensors are a few years old now. I actually thought I ordered some. I had a project centered around a fluid level controller. The type that uses floats to go from A to B when a tank gets low, and needs filling to the top. I used the A to B logic as dry to wet. Each float input, swapped for a moisture sensor. It worked, but nothing lasts in our root zones. So.. I guess it didn't work actually. Good try though

These 'capacitve soil moisture' probes are a big hit in search engines. Mostly connected the Arduino based control though. These PC boards are indeed, $2.50
Amazing really. A processor board costing less than the switches and relays I play with. The $10 kit really does have all that stuff in it. Though the pumps are a bit of a let down, as the wire isn't long enough for more than a buckets depth. While they also limit pot height to about 50cm, or they just won't get the water up there.
The relay boards are good. I have used quite a few, in situations where everything mattered.
The probes are stamped out of a big sheet it seems. The edges are sometimes prone to letting water in. It's wise to paint the edges with a sealant, to finish the job the factory should of done better.

The only thing I'm unsure of, is what good is a controller, if they are not showing us windows based software to set it up. I only speak English, with some use of binary programming, to get more channels on my CB radio. It's just not going to cut it

the way to go for to program Arduino and similar microcontrollers on Windows is, imho, to use VSCode and the platformio SDK plugin.

i've used this with great success to program esp32 microcontrollers, which are very affordable.

I think I may have a couple of those capacitive moisture sensors, but haven't bothered with them because, well, I use blumats, which do this job without any digital computation. 100% analog.


these things (esp32 microcontrollers) are great for making small, low cost, networked (and/or timed) relays for turning on/off pumps or lights.

 

[Ca++]

Oh yes, Platformio's sdk plugin

The most loved IDE solution for Microsoft Visual Studio Code​

I will just go and run that through google translate. I won't be long.

 

[Ca++]

I think this kit, could benefit from a couple of VR's on the unused analogue inputs. One for dry, one for wet. Perhaps a switch to achieve wet at will. There are 5 inputs, so sensor, low, high, override.
That would be kit I could use. I know with actual programming knowledge, this kit could do many buckets. However, I think I prefer a kit per bucket, with analogue control. This way, I can run my buckets down to dry, and then turn the dry VR until it waters. Then, when it's wet enough, I can turn the wet VR until it turns off. JD. No need to plug in a PC ever.

This is software that could be freeware, or someone could redistribute the kits with PSU and code applied.

I literally don't know where to start, after the opening line I quoted. It tells me I can't read yet, so comprehending the book is years away. I know it could easily make an analogue kit though.

Yes.. that's it. I bet some of the bigger boards are the same but with more channels.

I need that 5th analogue input. I need a run-on timer. 0-5mins, to run the pump after wet is reached. I need run-off to occur after wet.

At power on, or after reset, it should wait for dry, not go straight to wet. We have the wet button if we want that. There is no need to prevent the run-on when using the wet button, as there is a board reset switch to get back to dry monitoring.

I can see the flow chart in my mind. I think the process control makes sense how I laid it out. A second look will refine it further. I can do all this with modules, but the cost and complexity offers nothing to the community. $10 though (plus psu) is within reason. I presume the sdk plugin is the software package that could be shared, once created. Leaving people like myself, just needing to wire it up, and box it.

 

[GenghisKush]

Oh yes, Platformio's sdk plugin



I will just go and run that through google translate. I won't be long.

OK sorry I was a bit sparse on details.
And the thing is an IDE, not (as I wrote) an SDK. No matter.

This is the install guide I followed.

PlatformIO IDE for VSCode — PlatformIO latest documentation

docs.platformio.org

You'll have to install Microsoft VSCode for this. It's free and widely used.

Visual Studio Code - Code Editing. Redefined

Visual Studio Code is a code editor redefined and optimized for building and debugging modern web and cloud applications. Visual Studio Code is free and available on your favorite platform - Linux, macOS, and Windows.

code.visualstudio.com

Hope this might be helpful, and not confusing.

 

[Ca++]

This is a phone

This is a record player

These are programmed in Ca++ language.

This is me reading your link

I will now exit this thread, wishing I had kept my mouth closed. Find another website, and practice with my python in seclusion.

 

[Ca++]

Don't be sad pal We all have our weakness. Mine is schooling in the 70s and 80s

This link is the project at hand https://www.makerguides.com/capacitive-soil-moisture-sensor-with-arduino/
Here it is again, but expands upon it, with cloud coverage https://www.makerguides.com/automatic-plant-watering-system-with-arduino-iot-cloud/
This time, a screen is added, to display current moisture https://www.instructables.com/Automatic-Plant-Watering-System-Using-Soil-Moistur/
If seeing is better than reading, here is a youtube
Down the side, there are many more youtubes.

It seems everyone of these, relies on the user needing programming knowledge. The idea of what wet and dry is, will change with substrate. Perhaps even with the idea of 'dry' changing, as we stay a bit wetter by mid bloom. The two extremes of dry and wet need analogue adjustment, or it's not going to be practical. Also, that run-on timer, as there is no way of getting any run-off, at all, in the projects I have found.


Grower meets programmer... A battle of whats needed, based on different experiences.

 

[GenghisKush]

I think the part that threw me was "what good is a controller, if they are not showing us windows based software to set it up?"
I sincerely think you would have a better experience developing for ~arduino aboard a unix machine, rather than ms windows. So much unnecessary byzantine bloated software just vanishes from the experience. you can do everything you need with a terminal and a text editor, as it was in the beginning.

 

[Ca++]

Is Windows bloated, or is unix something with it's bits cut off.

School yard humour is all I can add here. What I would of liked, if it had to be via computer, is something in Oxford English. Instead, we have code. The building blocks behind a user interface, with no interface a layman can use. It's not a finished product at this point. Though I feel sure that adding the analogue controls wouldn't be hard, for someone that understands this stuff.

I have no programming skills. Bring me a broken piece of electronics, and I will fix it. Programming though.. not a clue. We can buy this sort of kit with metal sensors for similar money. Giving a more logical feel to setup. I have a couple. Thus, using these better sensors is just a conversion board. The run-off function a timer board. It's not particularly hard.

I automated a grow once, but it was a shoe box sized project box, crammed with boards/relays/contactors. The box alone cost more than a PC based electronics set.

This 'should' of worked.

3.0C$ |Watering Plants Automatically|automatic Watering System For Indoor Plants - Soil Moisture Meter & Irrigation Controller

Smarter Shopping, Better Living! Aliexpress.com

www.aliexpress.com

3 gold coins, on your doorstep. Simple enough. Shove the spikes in. plug to board. Add 12v. Set min RH to start pumping, and RH to stop. Use relay to control 12v pump. All in, with a 12v adapter, about $7.50
They do work, but not even for 3 months. The probes are cheap, but I'm not looking fir cheap and constantly calibrating. I want reliable, but this simple.
What I need to do, is figure using the better probe. Then we have something that people can work with.

 

[Ca++]

This is how the old boy's might do it.

Under £20 having bought a 12v and 5v power supply. A 12v pump. Bit of pipe. A project box and hook up wire.

The sticks swing 0-3v as they get wet. Each gets a voltage controlled board, that flicks a relay at the moisture level you class as wet and dry. One relay is the 'empty' and one the 'full' as far as the fluid level control board is concerned. When it see's empty, it pumps until it see's full. Then additionally we have the run-on relay, so once wet is reached, there is extra pump run to achieve run-off.

That last timer is important. Your wet sensor shouldn't be set for 100% wet, as any sensor drift might leave it looking for 101% wet. So it would pump the tank dry. Instead, you aim for 90% wet, and then a few more seconds/minutes of run-on. This is missing from the Arduino projects I have seen. Which are incapable of run off.

 

[blondie]

Interesting thread. I cut my teeth years ago as a team member On a project programming a nuclear simulator for good old Uncle Sam. Never looked at VS but have many other languages. The Unix vs windows debate is almost as old as Bill Gates. It’s not worth the brain power.

Programming this would take an effort. How much is anyone guess without diving in. It would take some hours at least just to come up with an estimate. Myself I try to be a minimalist simple minded. As such this gadget seems superfluous. I have been using below on my current grow and it’s working well. Extremely simple and effective for less than 20$. No hours of programming needed, no spikes to replace. No point of failure. I can leave unattended for something like 4 days easily.

Self-Watering Fabric Pot Base, 4-Pack

Your source to create smart growing systems with components including advance grow tents, inline fans, clip-on fans, LED grow lights, and UIS™ controllers.

acinfinity.com

 

[Ca++]

Be interested to hear more about the wick system, and see it's results. It's not a common method. As I look at it, I picture my roots going straight into the reservoir, and the 1 gallon capacity is about 3 days water. Based on 4 plants per meter. I guess with small plants, water use is smaller, and you could go water only with a big pot. Big plants need feed though, which is another set of problems.

I think, tbh, I would fail with the wick system. Unable to manouver them all out for regular cleaning, I would run into manganese toxicity. It's those plants at the back, weaved through the net. They are not very easily accessed. Wicks of the past needed attention, as they changed colour, and roots interfered. Here I see holes the wicks drop through, which roots will squeeze through. I never saw a fit and forget system. Though the idea was popular, and plenty of mops met there fate in wick systems. They have remained a fringe idea.

I bet for a weekend away, they are very useful if you know how to dial them in. Better than standing your plants in the bath. I see this thread does keep looking at individual plants, but I'm keeping scalability in mind. The first thing I looked at with the wick bases, was how they plumb together. If you link them up, then they all go dry together. Watering frequency wouldn't be dependent on the thirstiest plant. You could also add a reservoir. Something like the airlock system of smaller humidifiers. So you can fill a decent sized bottle. It's only a bottle cap with a hole in it. So you might get to a week.


My needs obviously effect my decisions. I'm designing around what I can do, and need, and could perhaps be shared. One mainstay of my needs, is multiple plants, with a common feed supply. I do think about online injectors, but a single reservoir with a distribution system wins out. I don't mind if setting up takes hours, if it leads to easy maintenance. There are threads on here about filling tanks automatically. One such board is pictured above. Using a washing machine water valve (instead of the pump) you need not fill the bases at all. Just connect them up to a common control bucket. I use the board and pump set above, with the sensors swapped over. Common bucket. When the water gets high, it pumps until low. It's how my run-off is removed. It could just as easily fill your bases. Allowing extended breaks, if you trust them the wick system for extended periods?

 

[Ca++]

Anyone interested in level control, for filling or draining containers at predetermined levels, could have a read of this. It outlines the operation.

https://assets.omron.eu/downloads/datasheet/en/v3/f043_61f-gp-n8_conductive_level_controller_datasheet_en.pdf

That's an industrial control. It plugs into an octal base. The wiring diagrams presume you have one

You run these at mains voltages generally. So the pumps and valves you have, are fine. The electrodes should be marine stainless (316L) or better.
Though $200 items, you often see them for 10

1PCS Omron 61F-GP-N8 PLC Floatless Level Switch 220VAC New | eBay

Find many great new & used options and get the best deals for 1PCS Omron 61F-GP-N8 PLC Floatless Level Switch 220VAC New at the best online prices at eBay! Free shipping for many products!

www.ebay.com

 

[GenghisKush]

Interesting thread. I cut my teeth years ago as a team member On a project programming a nuclear simulator for good old Uncle Sam. Never looked at VS but have many other languages. The Unix vs windows debate is almost as old as Bill Gates. It’s not worth the brain power.

Programming this would take an effort. How much is anyone guess without diving in. It would take some hours at least just to come up with an estimate. Myself I try to be a minimalist simple minded. As such this gadget seems superfluous. I have been using below on my current grow and it’s working well. Extremely simple and effective for less than 20$. No hours of programming needed, no spikes to replace. No point of failure. I can leave unattended for something like 4 days easily.

Self-Watering Fabric Pot Base, 4-Pack

Your source to create smart growing systems with components including advance grow tents, inline fans, clip-on fans, LED grow lights, and UIS™ controllers.

acinfinity.com

Blondie those things look cool. I think if I weren't already using blumants, I'd use those. It sure is nice to be able to leave a grow unattended for days or weeks and come back to happy plants.

I 100% agree this is "hammer seeking nail" problem solving. Certainly not a waste of time or effort, but not likely a path that leads to a long-term sustainable endpoint. It is notable, I think, that the people with coding experience are recommending non-digital solutions to this problem.

So it would pump the tank dry. Instead, you aim for 90% wet, and then a few more seconds/minutes of run-on. This is missing from the Arduino projects I have seen. Which are incapable of run off.

Ca++ fwiw I almost never water my plants with any run-off. This only happens when I hand-water, which I do only occasionally to add inoculants to the media (coco).

 

[blondie]

Thanks for the exchange of knowledge. I truly enjoy meaningful respectful discussions. I had forgotten about wicks failing or needing replacement. I believe they are offered. I began using these mid March near the end of veg. They have been in use ever since. To date, I’ve only added water. Despite my brown thumb, the plants grew well. Any grow issues are grower error, not equipment. I have had cause twice to lift the fabric bag and move the plant. First I had a slow leak in one. Replaced immediately by the manufacturer and is not the point. When I lifted the bag, I fully anticipated roots to have grown through the bag and through holes into the base. I was pleasantly surprised this did NOT happen. All roots remained in the bag. The wicks were in good working order. I swapped out the base and set the bag back down. Zero issues. Second time I needed to flush a plant due to nute lockout ph issue. Same with this plant, No roots visible. Lifted the bag, set down gently. After some days I set back on the base. Wicks on her also appeared in good working order. So far so good. If you don’t add nutes into the base, I can’t see a need for cleaning during an entire grow cycle. For nutes , organics and if needed top dress or small volume of tea. Another member here does exactly this.

The benefit of individual bases is you are not forcing each plant to take an identical amount of water. I have two ka5h going side by side and they take different volume of water.

These are my experiences so far. I’m glad I bought these. I strive for less complications and less demands on time. This fits my needs.

Edit: I have no concerns leaving plants. This system has worked flawlessly to date. Water volume could be increased I suppose by some fiddling, if one desired.

 

[Ca++]

Interesting, and I thank you also, for the non-combative reply.
4 months run time, and still going, is long enough to call it functional.
This teddy was hung in a waterfall for about as long, and is officially fossilised

Your water hardness may vary. Along with flow rates. However, the wicks will have some sort of life expectancy. Though with RO, that might be a long time then.

It's a little concerning you would see no roots at all. Do you have them copper weave clothes in the bottom? have no played with them myself, but they stop roots dropping out of pots into reservoirs, where people often start to get rot. Maybe as a commercial product (not an old mop) your capillary gear has the weave?

@GenghisKush I believe a long term system is in fact what I'm looking at. I already have this up and running using a commercial solution, and here I'm ironing out the problems they didn't. Taking a product that should work, and getting it up to my standards, is, I would hope, a long term solution.

My tank already fills itself, and sorts out the pH and EC each day (actually, I'm mid move so it's boxed away) The pump is on a timer, like most drip systems. I'm just trying to add some drip on demand tech. Not initially for control, but for monitoring. I use water stressing in early bloom, then like to stop by week 4. However, they are drinking more each day around then, and so glancing around at some basic moisture sticks would be handy. I had one of them frogs that croaks when it gets dry once. Our salts just killed in in weeks.

I have known people do coco with no runoff. It was okay, which surprised me. It was better when they started to chuck the waste from the saucers though. Enough to make them get their pots up off the floor, and build a system. Often 'kiss' is just lazy, and leads to a lesser outcome. Tech in farming is getting really good now. Well beyond the scope of anything I'm ever likely to use. Things like some run-off for your coco are really worthwhile though. It's not a 'kiss' matter. Coco is constantly decomposing, releasing sodium and potassium. More that your plants are happy with, or the coco's cec can manage favourably. Our plants will eat sodium instead of potassium, and grow softer, more mold prone material. The cec won't be able to take on the calcium and magnesium it wants to, due to the sites being heavily occupied by sodium and potassium. Everything is out of balance. Getting worse each feed. NPK gets eaten quickly atthe ratio's a plant wants. Many other things are just present in a high enough number, for the plant to find some. At each feed, the NPK (though not K in coco) needs putting back, but many of the others are still abundant. Over time, feeding them again and again, leads to toxicity problems. We don't see signs with a 10% over feed, but we do see it in decreased results. That's where AI eyes up our plants better than we do. Cannabis testing has shown the losses of excess feed and when we actually notice. A 20% decline is quite possible, with just mild signs. Coco of all things, really does need the excesses washed out, for a notable improvement. This ion balancing thing, can't be done without it.
A friends Sage really did surprise me though. I told him exactly how long it had been down, which surprised him. It wasn't bad though. It just got better when I clued him in.

 

[Ca++]

I'm revising my method. In typical watering situations, the pot becomes dry, and a predetermined amount of water is added. It might be a jug. It might be 2 mins on a drip timer. However, the data points, are dry and volume. Until now, I have been looking at dry and wet. I don't really monitor for wet on a daily basis. Certainly capillary just puts in a bit. Like the carrots do. I think sip is similar. Autopots measure in a bit. Drip guys might run for as long as they know it takes to fill a pot their size from dry. Plus 10% longer.

This means, all I really want it the plastic sensor. 50 cents.
Then I want a board that puts that sensor voltage on a screen, and that's a moisture meter built.
If that same screen board, could have a voltage programed in, as which it flicks a switch. Well that's $4
Then if that switch told a digital timer to go, we are looking good to wire a pump to it. Say a timer that can be set to the second. That's $2.50
Plus the 12v and 5v power supplies. $1 pump. It's a two boards project.
https://www.aliexpress.com/item/1005005506836637.html <meter
https://www.aliexpress.com/item/1005004968877145.html <timer
Use any pump you like really. However, the board has 12v present. Which is a safe voltage. An old router psu might give you 12v, and a phone charger the 5v. I would likely make the 5v for the stick, from the 12v. It's a $1 dc/dc convertor

I think I have it back to a $10 kit. Only this time, with screens that give real time figures, and programming the Chinese can probably explain to us.

I feel I have accomplished something within reason now. Though I'm still trying to trim it down and make it more user friendly, I'm not sure how much better this can get.

 

[kingape]

Anyone tried automating the mixing of nutrients through perilstatic pumps?

LEDGardener had a good series going on, but he stopped.

Please share any resources