Ideas for Automatic garden watering system project?

[rfengg]

Folks,

I am looking for ideas and suggestions to implement an automatic garden watering system in our garden.
I am aware that there may be commercial, off the shelf systems that may suit my requirements but I would like to take this on as a personal project from scratch.

My requirement is that I should be able to automatically turn on the water for the drip irrigation system for 10 mins only every hour.
I am thinking of getting a  solenoid water valve to control the system such that the water supply for the drip irrigation system  is ON for 10 mins and OFF for 50 mins every hour but do not know how to setup the electronics to drive this?

Would something like Arduino to set this up be the best option?

[admiralk]

Some time ago I started a similar project, for a misting system, using a Nextion display. I think it got too complicated for the way that works. I recently started rethinking the project using regular AVR chips. Arduino would probably work fine depending on the complexity of the interface. If you just want to plug it in and not worry about changing the settings, Arduino would probably be the simplest approach. If you care at all about the accuracy of when it switches, you will need to use a RTC also.

At its simplest, it is just a blinky program; on 10 minutes / off 50 minutes. How complicated you want to get after that will determine what is best to use. 

[wilfred]

My first thought was to have a comparator monitor the charging of a capacitor over an hour. Switch on when the voltage is zero or at some threshold and off when the voltage passes the next threshold. A 60minute sawtooth. A pot or two could make adjustments easy.

I also thought about a small pump to charge an elevated reservoir (ie a bucket)  from which water would be gravity fed. I have thought about this as a solution for watering pot plants whilst absent. A small scale issue for a family member.  I even considered loops of tubing with drip feed holes at the bottom of each loop and one loop or two for larger pots. About 12 pots in total. Each loop would contain the required quantity of water and the level monitored by a small bead in each loop. The loops of tubing would form the elevated reservoir. This would highlight the drip hole getting blocked.

There is an air conditioner on the balcony with the pots and the pump was a way of incorporating the outflow of water from it.

[floobydust]

So far it's just a timer you need to connect to your solenoids.
I'm not a fan of simple watering systems because they are not smart enough:
If it's raining, you probably don't want to water.
If the weather is hot and sunny, you need more water. If it's cold and cloudy, you need less.
If it rained solid for a day or two, probably don't need water for a day or two.

A good quality soil moisture sensor would help.

I would use an ESP32 in the Arduino ecosystem, so you have remote (Web) access, if you are the smarts deciding to turn on/off, like if you are away on holidays etc.
Otherwise, it's possible to grab local weather conditions on the Internet and decide automatically in your software if you choose to make it smart.

[ledtester]

Would something like Arduino to set this up be the best option?

Yes - just google "arduino watering" for lots of ideas and examples.

Most of the cheap "moisture sensors" have problems of one form or another. Here's a video which reviews a bunch of them:

[soldar]

My requirement is that I should be able to automatically turn on the water for the drip irrigation system for 10 mins only every hour.

You want the valve on for 10 mins and off for 50 mins continuously 24/7? That's strange. What kind of plants are we dealing with?

At any rate, doing a 1/6 duty cycle is trivial with a counter like a 7492. I guess it does not have to be very precise so you could use a 555 to generate the base clock signal.

I am dealing with the same issue myself. I have a big yard and I am tired of watering by hand in the summer but the first issue I need to deal with is designing the drip hose system and I have no idea where to start. Maybe I will start a thread in the general chat section. Once I have the hydraulic part the electronic control system should be easy.

In my last system, years ago, I was using 220 V valves taken from a washer. Now I plan to use low voltage valves as I have more awareness of the safety issues.

And I have a very small system for a couple of pots made with a very low voltage aquarium pump.  It just takes water from a container and pumps it into the tube over the pots. Only works for a very small scale thing but cost me nothing to build.

[Wimberleytech]

I have a system that waters my tomato plants once per day.  It consists of
1) 12 volt battery
2) solenoid valve (https://ebay.to/2OJxJjM)
3) Timer (https://ebay.to/2Tcqvn0)
4) Solar cell
5) Diode

I am knee deep in arduinos and everything else, but I was not looking for a project...just wanted to make sure the tomatoes were watered.

The down side of this solution is that the solenoid draws lots of current while on.  Not good.  For me, though, I water once per day for one minute.  Not a problem.

Irrigation systems use latching relays.  Power is drawn only when the state of the relay is changed.  That is what you need.

[soldar]

I have a system that waters my tomato plants once per day.  It consists of

>> 1) 12 volt battery
I use a small adapter/PSU which I believe is simpler. 

>> 2) solenoid valve (https://ebay.to/2OJxJjM)
That works if you have some pressure. Those valves will not work with extremely low pressure which is why I am using a little pump in my tiny system.

>> 3) Timer (https://ebay.to/2Tcqvn0)
I am using something similar.

4) Solar cell
What does this do? I just have mine set to water at midnight.

5) Diode
??

Can you explain a bit more?

>> the solenoid draws lots of current while on. 
It is not difficult to activate the solenoid with higher voltage/current and then reduce the current to a lower holding level. Still, I do not think this is much power.

>> Irrigation systems use latching relays.  Power is drawn only when the state of the relay is changed.  That is what you need.
I am not sure this makes a big difference. A relay does not consume that much.

[Wimberleytech]

>> 2) solenoid valve (https://ebay.to/2OJxJjM)
That works if you have some pressure. Those valves will not work with extremely low pressure which is why I am using a little pump in my tiny system.

Correct.  I have a high-pressure source

4) Solar cell
What does this do? I just have mine set to water at midnight.

5) Diode
??

Solar cell and diode charge the battery when the sun is out

>> the solenoid draws lots of current while on. 
It is not difficult to activate the solenoid with higher voltage/current and then reduce the current to a lower holding level. Still, I do not think this is much power.

About 1/2 amps.  That is a lot.

>> Irrigation systems use latching relays.  Power is drawn only when the state of the relay is changed.  That is what you need.
I am not sure this makes a big difference. A relay does not consume that much.

It does make a difference

[ogden]

You don't manufacture light bulbs. You buy them. The same here. - Just buy timer and spend your time building something more exciting. There are mechanical < 10$ timers that run off grid frequency, yet they may not match your 10 minute/hour requirement exactly. You may end-up with 15 minutes every 1.5 hours or something else that requires calculator. Then there are electronic timers, those may have other limitations like maximum number of recurring events you can set-up (usually 10), but they are much more reliable/time_stable compared to mechanical timers.

https://www.amazon.com/Best-Sellers-Home-Improvement-Plug-Timers/zgbs/hi/6291363011

[Chriss]

Hi!
Ok. So far so good.
The fundament of the project should be first setup.
And I would develop some board with uC like the AVR family or any other with what you familiar. For me is some AVR type uC ok.

I would start to develop the most simply parts like the timing with an RTC module integrated.

This is the first step what hase to tick life into the project.

When the syatem starts to make thinks happen then I would polish some functions and add new features like a soil temperature and humidity sensor to have info about the soil characteristics. Then I would integrate a sort of rain and sun sensor. If it is raining I probably don't need to add extra water. But with a sun sensor I could collect sunny/rainy days etc. After that I would integrate a mem card to save some setups and data to it just to have some statistics etc... and the major idea is for me to build up a system what I can later at any time modify without to much soldering or hardware hacking/modding...

I know maybe my idea sounds too robust but trust me, it isn't... :-)
Just start to have fun and at the end we all will have a huge result. :-)

I hope I helped a bit to realize your project.

My best regards.

Sent from my SM-J500F using Tapatalk

[JeanF]

Hello,

a few weeks ago I wanted to build a very simple system to water my plants while I'm away. It's very basic but it's a good opportunity to share ideas in this thread. I've let it run for some time before my holidays to have enough time to check it was working well and to spot any leaks.

The main requirements were
- off-grid water supply. (in other words, some sort of bucket). I’m renting the apartment and I wasn’t too enthusiastic about letting homemade pipework under pressure while I’m away
- it shall be very simple and quick to build as I had other stuff to do before leaving

My dad also has two bay trees that would need watering, so for these I used a large 80 liter plastic trash can, a small aquarium pump (Newa MaxiJet MJ750) and a few meters of clear PVC tubing. These pumps can move a fair bit of water at no load, but they have a very small maximum pressure (1.65 meters of water, so around 0.16 bar)

At this pressure, I wasn’t sure the drip emitters that could be found locally would work, and there was not much time left for testing with other emitters (and I wasn’t sure they would be delivered in time before I left) so I thought the small diameter tubing itself would regulate the flow, so I left the end of the lines open.

The timer is a simple mains powered digital timer that I bought for around 7 euros, I’m sure there are many local equivalents. I can program 20 events, each with {day, start time, stop time}, and “day” can be any single day, or each day, or M W F, or T T S, or MTWTF, or SS
https://images-na.ssl-images-amazon.com/images/I/81A4HYyybjL._SY355_.jpg
It’s quite capable for the price, saves the hassle of DIYing one as Ogden said. Only downside is, its resolution is 1 minute, so one can’t water the plant for say 30s, could be an issue for very small pots if the pump is too strong.

That worked well, so while I was at it I built a second one for a window box in my own place. The pots stay outside and I didn’t want to drill holes through 25cm of hard concrete so I smuggled the pipe inside a vent hole at the top of the window. This is higher than 1.65m so I had to use another pump (a Laing DDC-1T that I already had laying around) which has a greater max pressure (around 3mH2O, if I remember correctly). The downside of this pump is that it can’t be used underwater, so the reservoir can’t be a simple bucket, it has to have an outlet fitting. I used a jerrican from Pressol, model number 21167, that I already had. Also it runs on 12V so I had to use a dedicated PSU. All in all this second system is less pretty than the 1st one, but works just as well.
I even built a 3rd one for 2 more small pots, I’ll spare you the details, it’s just like the 1st one (same pump, same timer, only a smaller bucket). On this last one, the two pots are inside a plastic tray which sits on top of the bucket, and has a hole drilled at the bottom, so any excess water falls back into the bucket. There is a crude filter below the hole (one half of a stainless mesh tea infuser) which will hopefully prevent the pump from clogging with too much dirt.

You may see these 3 in action here  (1min, no sound) :

Obviously these were built quickly, and have some drawbacks:
-flow is heavily influenced by tubing length and height. Pressurized systems with drip emitters are way better in that regard (particularly with pressure compensated drippers)
-it doesn’t scale up well, as you need one pump and one timer for each plant (or group of plant that need the same amount of water). A more complex system with a microcontroller, a single pump and several solenoid valves would be better. I’ll may give it a try in the future...)

Two interesting links I found:
https://aces.nmsu.edu/pubs/research/agmech_eng/RR773.pdf
https://medium.com/@ezheidtmann/low-pressure-drip-irrigation-a-homemade-investigation-364285428417


Thank you all, they are some interesting ideas in this thread

[emece67]

.

[soldar]

If you are watering more than just very few plants then the hydraulic design is important and possible more complicated than the electronic control.

Rather than the thin "spaghetti" tube I used the regular, black irrigation tube, about 16 mm diameter. That means much lower pressure drop. I had some old lengths and used it. I just drilled some holes in it and it works well.

If you have a longer run and pressure drop is a concern then you can start the run at higher altitude and run the tube downhill to maintain pressure.


I have a big project that I have been putting off for years which means I have wasted many hours watering by hand which I could have saved if I had repaired the existing irrigation system.

The previous tenants of this house left on bad terms and destroyed everything they could including the irrigation system. (That's the way things are done in Spain where impunity runs rampant.)

The main tubes are buried but I can see little drip tubes here and there. There is a register box or well and I see five ball valves so it seems they had five zones but it seems they were all turned on at the same time. I guess I could try to pump water into each one and see where water came out.

On the other hand, it may be simpler to just design and build a brand new system.

The electronic control part is the least of it.