wireless battery powered garden irrigation advice

I am designing an irrigation and monitoring system for my gardens, and want to have a wireless setup so I do not have to run lots of wire.

I was thinking of an xbee system for the wireless, and some sensors (moisture, temperature, etc.).

The remote ardunio/xbee/sensor units would be battery powered. They would not have to communicate very often - maybe a couple times a day to send current temp, moisture, etc. so the batteries I would think should last a long time if everything can go in to a low power mode...

This would all feed back to a PC with an xbee so the PC can monitor the data.

The key piece that I am having trouble with is the irrigation. would like to have these arduinos open a valve to water the garden. So the idea is the PC would send a signal over xbee to the arduino, then the arduino tells the valve to open.

I am having a hard time finding a reasonably priced valve that will run off battery power. This is standard household water / pressure.

What is frustrating is that I have a 'Nelson' brand digital water timer that runs off 2 "AA" batteries (3v) and inside it has what I believe is called a 'latching valve'. I think the idea is that power is applied, the valve rotates or latches or whatever and water flows and continues without any power. Then the unit applies power again to turn it off until it rotates/latches...

These timers cost from $25-40, so I would think I should be able to find this (or a similar) valve somewhere for that cost or hopefully less. I have spent HOURS trying to find something that meets the cost and low power requirements...

If anyone can help with the valves, or if you have any thoughts or see any issues in my idea, I would greatly appreciate it!

By the time you find the valve, and a housing to hold it, wire it up, and hook up hose ends, you’re going to be back at that same starting price. You might want to start with that, and just remove the timer section.

So I started to do some searching to see how cheap I could get a timer, and came across this:

It looks like it plugs in to a battery powered timer, and someone commented that it lasted at least a year for them… A couple complaints on durability, but I think I can deal with that.

So maybe I will take a chance and buy this is open it and see what it will take to make it operate…?

did you know the standard rainbird style of valves use 24VAC? That's basically 2 12VDC batteries in series and then switched on/off at ~60Hz. But I think they also pull something like 1/2 and amp or so of current.

So, before you get all excited about doing wireless, you might want to look at the power requirements of the valves based on how long you need the valves on. My guess is that even a couple of motorcycle batteries are only going to last a couple of weeks without some kind of charging system like solar. Otherwise, you're running 110VAC out there and you might just as well run an RS485 or ethernet wire out there.

You know the rainbird style of valve is robust and is less likely to fail in the ON position. Keep that in mind when thinking about what valves to use.

I've been thinking of hacking an Arduino onto a remote 4 channel irrigation system we have. I've got 2 old car batteries connected to an inverter which powers the 24VAC controller and valves. Got a 15W solar panel charging the batteries and a timer turning the inverter ON/OFF so it's not running all the time. In other words, I've had this stuff on the mind for a while. I figure it'll be easier to hack the Arduino onto the controller than do the circuit so Arduino directly controls the valves.


Yes, good point about the power requirements, that is why I am looking at the valves that are in the consumer 'sprinkler timers' I mentioned - I have one of these, and it runs off 2 AA's (3v) and mine has lasted 2 seasons so far off one set. I water almost every day (during the season)...

that's impressive if you've run it every day for 2 years off of 2 AA batteries. Still, I wonder what the life expectancy is on those valves. I've got 8 valves at the house that have been going atleast ever other day for about 12 years. I had to with failed diaphrams when the pressure regulator failed and the failure was stuck ON. In SoCAL with the water shortage, it's going to get expensive if a valve fails and it's not caught ASAP.

It makes me wonder why all the valves are not made like those 3v ones you've used.

@Dougl - I'm a noob about to try and do something very similar to your setup, but I have a couple of questions...

My first question is: What wattage solar panel should I buy? Do you have any ideas about how the actual wattage produced during a 24 hr day relates to the max output listed on the panel? This all depends on how many watt hours I expect to use on a daily basis... I'll try and check out the solenoid data sheet to figure out how they work when I get home from work...

A related question is: Suppose I wanted to measure power use in watts. Would I need both a multimeter (for voltage) and a clamp meter (for amperage)?

Another related question about 12 V DC (battery) and 24 V AC or DC (out to solenoid) is here if you have some thoughts about that... http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1269302167/15#16

All thoughts/opinions welcome...

@SSzretter I'm in the same boat. Right now I'm putting together a garden data collector, and will later try and have it do something useful based on the data (eg water the garden)...

So far, this is what I have sketched out:

  • Basic design is a tupperware on a ~18" piece of 1" PVC which sticks into the ground
  • Coming out of the PVC (this is buried into the ground) are temperature and moisture sensors at different depths
  • Above ground is a temperature sensor, a light sensor and possibly an old anenometer I have from my parent's old weather station

Soil temperature is interesting for several reasons (it tells me very speficially when it is time to plant seeds of different varieties to maximize germination rates). I saw two main approaches to this, one digital, one analog. The digital approach is to use something like the quad thermocouple approach http://ryanjmclaughlin.com/shop/quad-thermocouple-interface-12/ while the analog approach would be to use an AD595.

A cheaper digital tack (from what I can tell) is to build leads on to multiple DS18S20's. I think the range and resolution are not as good as the much more expensive thermocouple approach, but I think it's still fine for my purposes. http://www.arduino.cc/playground/Learning/OneWire states that

For getting started, especially if your chip is within 20 feet of your Arduino, the parasitic option is probably fine.

Hopefully if I run copper core ethernet cable down the PVC pipe and stick out the protected sensors at different depths, I should be OK, and only have one ethernet cable for all my temperature probes. http://para.maxim-ic.com/search.mvp?fam=temp_sens&374=1-Wire&374=1-Wire%20Non-Network&hs=1 Somewhere I saw a post about waterproofing the sensors by enclosing them in copper tubing...

Somewhat more important is soil moisture, and I was planning on using the recommendations of http://www.cheapvegetablegardener.com/?s=sensor&x=0&y=0 either the purchased http://www.cheapvegetablegardener.com/2009/08/how-to-use-vegetronix-soil-moisture.html or diy varieties http://www.cheapvegetablegardener.com/2009/11/how-to-make-cheap-soil-moisture-sensor-2.html

After that (I'm a statistician, so I'm trying to design some data that will have good explanatory power for garden yield, time to maturity, etc) I'm interested in estimating the lumen-hours of light during the day using a CdS, humidity and a couple other things... I'm not sure yet how best to measure humidity. The only option I found so far was: http://www.sparkfun.com/commerce/product_info.php?products_id=8227 and I don't know how it works (eg What part can be exposed out of the weatherproof enclosure the arduino will be in- somehow I'll have a lead that connects to sht15 + something? Any ideas for outdoor use?

I'd like to put it all together with an Xbee or wifi that sends the data to my server which then throws it into a database used for model fitting/calibration.

Does this sound like your project?

Did you ever get anywhere with that cheap valve? I have been using 4 of those for 2 years, and they are outlasting my pro Toro valves, even with the extremely hard water I have.

I go through 1 set of batteries per year, so I was hoping to build an arduino based controller which can control these valves, and interface the controller using xbee. It's a simple concept, but surprised no one else has tried this :(

I'm looking to do something similar to what has been discussed. I have some property where there is no power but there is a shallow well. I'd like to have a garden setup and automatically irrigated.

I'd like the Arduino to do the following: - Sense daylight to control time of irrigation - Sense battery voltage to ensure enough power for pump - Sense panel output to make sure batteries are getting charged. - Turn on the pump (obviously)

So far the I'm thinking of the following pieces" - Single solar panel (75W) - 2 sealed 12V PbCa batteries - Solar battery charge controller - 12V utility pump (Harbor Freight #9576) - Voltage regulator for Arduino power

I don't think in my case I'll need a valve, but instead just a relay to turn on and off the pump. I'm wondering if I could even just run the pump directly from the panels without pulling from the batteries. Then all I'd need is a small battery/charge controller to power the Arduino.

Has anyone played with the Arduino and relays large enough to run a decent pump?

I'm wondering if I could even just run the pump directly from the panels without pulling from the batteries.

Keep in mind that the "75W" rating on the panel probably means something like "4A at 18V in full sun" (approximately). That's not going to set your pump on fire, since it's designed for a vehicle-type system that can run at 15V or so, but it may shorten its life a bit.

Other downsides: you must use a 75W or larger panel, and you must have enough sunlight whenever you run the pump. Using the batteries, you could use a smaller panel, since it would probably have at least a few hours of charge time for every hour of operation (but maybe not: it depends on the amount of time per day you irrigate). And the batteries could pick up the slack on cloudy days.