Battery for Irrigation control system.

Hey guys, this is my first proper arduino project, well have only completed a hello world tutorial for the lcd screen, but I am not a tech novice.

I am looking at building an Irrigation control system using an arduino for my mum's allotment, but I am a bit confused when it comes to choosing a battery for the project an wondered if anyone here would like to share their 2 cents worth with me.

The hardware I am planning on using is 1x Arduino Mega 1x 4 ch 5v relay board 2x 12v 3w aquarium pump 1x 16x2 LCD screen 4x 12v 0.6w N/C solenoid valves 4x moisture sensors 1x water level sensor 1x temperature sensor

I would like as much life out of the battery between charges, and i know that the way i code the thing will have a big part to play in how much life i get, but would like to know what people think about the options i think i have.

The one I think would be the best for me Is 12v Deep Cycle Battery I've seen these used in things like UPS and home alarm systems, just dont know if it suitable for my project.

My next thought is 12v Li-Ion Battery Pack again, dont know if this is feasable but I was thinking of getting 2 of these and run them in parallel to get 9600mAh

Or does anybody have any other ideas.....

Im going to keep the program as slim as possible, checking the moisture of the soil of each of the 4 sensors once a day, maybe twice during the summer. One pump will fill a holding tank when the water level is low, and the other will be feeding the irrigation system, so neither will be run for more than 1-2mins each day. And the LCD screen will only be on when either a button is pressed to activate it, and then it will time out, or when the sensors are being checked.

Thanks in advance Iain

A deep cycle 12V lead acid battery with a solar panel to keep it topped up?

Remember an average of 100mA vs 500mA is 1 day vs 5 days of running....

Use the sleep() function to maybe save some juice.

Ensure the moisture sensors are always off when not being read (they corrode fast when a voltage is across them). This will also save a lot of juice.

Can you get a reading of the device's current draw? This will help determine your best option(s).

The solar panel would definitely be a suggestion, but wouldn't know where to start with the recharging part so may have to wait till the summer for that, she goes away in march and wants just the basic for when she goes away.

I can't get a reading of current draw yet as I havent started putting it together yet, planning on getting started on it this week as work goes quiet after the holidays and I still have to purchase the solenoid valves and a few other bits.

I was hoping to use the sleep() function to keep everything off and possibly my routine would wake up once a day about 10am, check all the sensors, water plants depending on moisture level, check water reservoir and fill if required, check temperature and if over a certain temp set it to repeat at say 6pm, if not sleep till next day. I also want to have a physical button that would wake the system up anytime and run the routine.

I'm not sure of the difference between Li-Ion and LiPo cells, but LiPos do not not like being discharged to a too-low voltage.

For long life Lead-Acid batteries should be fullly charged at (say) 7 day intervals and they, also, should be be discharged to a too-low voltage. A full charge will probably require about 8 or 12 hours.

Apart from that Lead-Acid batteries are pretty much bullet proof. Of course, they are very heavy if you have to take them away to recharge them. If you have to do that it might be wise to have two batteries and alternate between them.

There is a lot of info on the Battery University website.

...R

We can assume the solenoids use 0W when off.

Look at the datasheets for any sensors you are leaving on the power rails and find the "quiescent current" or if that is not mentioned, the actual current they use.

Not tempted to build your own arduino to get rid of the inefficient USB and power reg?

An ATMEGA 328p could go for a very long time with the extras taken off...

Boring maths bit:

Assuming 5A (25W) power draw for 5 minutes a day. Assuming a 100mA quiescent current draw (all off) for the 24hours-5minutes.

ON consumption = 5/12 Ah = about 500mAh OFF consumption = 0.1*24 = about 2.4Ah.

Your big problem as you can see, is the leeching quiescent current.

Now, using a custom design with only the 328p IC and not USB and bad power regulator in sleep mode...say 10mA?

240mAh in OFF and 500mAh while ON = 0.75Ah per day. Double for "battery self dishcarge" = 1.5Ah per day required.

Use a trickle charging solar panel (say a 50W):

50W * 5% efficiency (winter?) = 2.5W Current from panel = 2.5/12 = about 0.5Ah 8 hour day = 4Ah of charge.

You may get away with even less. A 20W panel would give maybe the 2Ah you need a day.

MAGNATEC BZY93-C15 Zener Single Diode - A 20W 15V zener diode to place in parallel across the panel (will regulate the output voltage to 15V maximum). A 5A Schottky Diode in series to the battery from the panel after the zener in parallel will stop back current flow in to the panel from the battery over night.

Get someone to check my idea of the keeping the panel and battery happy...There will be a 0.3-4V drop along the Schottky diode, so a 15V zener will mean 14.6-7V regulated going to the battery.

Thanks for the replies guys, I'm not sure of the differences in batteries at all if im honest, I've just heard some aren't safe if left to charge for too long and with it being for my mum, and knowing how forgetful she can be, i dont want it setting her house on fire.

The solar charging thing sparked my mind and went looking and found This Instructable and wondered if this is the sort of thing i need with something like This Solar Panel???

I'm not scared to get my hands dirty and build my own Arduino at all, soldering skills arent to shabby, but time constraints i wouldnt mind the simplicity of using an Arduino board. Maybe if/when we come to adapt/expand the system.

IainStott: I'm not scared to get my hands dirty and build my own Arduino at all, soldering skills arent to shabby, but time constraints i wouldnt mind the simplicity of using an Arduino board. Maybe if/when we come to adapt/expand the system.

You will be expanding the system by another £40-50 on a larger solar panel and a much larger battery. Making your own arduino is very quick AND easy!

It is a big saving...in money and specs.

10W panel. Winter assuming 5% efficient = 0.5W

P=I/V

I = P/V = 0.5W / 12V = About 0.05Amps.

8 hours of "light" = 8 * 0.05 = 0.4Ah of charge on a winter day.

We were looking for more like 1.5Ah charged a day. I would say a 20W-40W panel is more what you need. Of course, you could come over and manually charge the battery with a power pack or the like as well...

I get about 100-200mAh from my 100W panel in winter at the moment on overcast days. With the 8 hours of light at that level, I get the 1Ah I need. It tops up a 100Ah lead acid battery with a solar charge controller.

Of course, on a sunny winter day, it produces much more (maybe 0.5A for 8-9 hours a day) giving me a good 4Ah.

So you may get away with a 10W panel, a 10W zener diode and a 10W Schottky Diode.

do you have access to city water ? where is the water you want to pump up to the reservoir ? why can't you use that directly ? not sure how much water you expect to pump in 1-2 minutes. a cup or two ?

if you pump up to a reservoir, you can just open a valve to drain to the plants. no need to pump down to the plants.

solenoids are a bad choice as they consume power constantly. a servo could open a valve, then wait, then close the valve. only using power when moving.

I would offer that you should check your soil moisture 4 times a day, then you would have more information from which to decide how much to water.

if you have city water, under pressure, then you can just use that and forget the pumps.

as for power, start with pump selection. you can calculate watts then figure running them 4 times more often than you think so as to build in some elbow room. add the operation of your solenoids, or servos. then figure what a fully running Arduino would draw, with sensors running all the time.

that is your absolute worst case. but you will have a sense of what is the biggest use of power and where you can save the most.

I would offer that you can use solar panels to pump during the peak light hours and not worry about batteries. if your reservoir is sized for 5 days, you can survive for a week after that. plants can survive droughts for some time.

as for pumps, the higher the lift, the more power needed. that is the direct correlation to power.

dave-in-nj: do you have access to city water ? where is the water you want to pump up to the reservoir ? why can't you use that directly ? not sure how much water you expect to pump in 1-2 minutes. a cup or two ?

if you pump up to a reservoir, you can just open a valve to drain to the plants. no need to pump down to the plants.

solenoids are a bad choice as they consume power constantly. a servo could open a valve, then wait, then close the valve. only using power when moving.

I would offer that you should check your soil moisture 4 times a day, then you would have more information from which to decide how much to water.

if you have city water, under pressure, then you can just use that and forget the pumps.

as for power, start with pump selection. you can calculate watts then figure running them 4 times more often than you think so as to build in some elbow room. add the operation of your solenoids, or servos. then figure what a fully running Arduino would draw, with sensors running all the time.

that is your absolute worst case. but you will have a sense of what is the biggest use of power and where you can save the most.

I would offer that you can use solar panels to pump during the peak light hours and not worry about batteries. if your reservoir is sized for 5 days, you can survive for a week after that. plants can survive droughts for some time.

as for pumps, the higher the lift, the more power needed. that is the direct correlation to power.

I thought these were solenoid valves. Ie. Closed via a spring and open when power applied?

The moisture sensors should be kept off until being read. That should be a non issue.

The 6W of pump, I am assuming they have a water butt but also some plants at a height (strawberry towers? Hanging Baskets etc?). Otherwise, making a few foot high stand for the water butt and gravity feed via a valve WOULD be a better option true.

A servo must have power at all times to have a set position.

Solar panels do not always give their rated current, especially on overcast days. A 20W panel may give you 1W of power on an overcast day...so nothing would work except on sunny days.

Cheers for all this info guys, its opening my eyes a lot :o

I don't think the winter months are going to be too much of a concern as the system is only being used for seedlings, so it will only be used really in the spring/summer. If she uses it for plant storage over winter we may end up looking at a mains supply.

We are hoping to use water collected in water butts, as im unsure whether gravity alone would be enough to move the water. The height difference is going to be 1 meter elevation max but saying that she may want to add hanging baskets at some point so maybe 2m elevation. Plus she wants to use misting heads on the irrigation pipes which will need a bit of pressure behind them. We have decided to use a reservoir because she may want to mix in plant food, im not too sure.

And after looking at it, with a 10liter drum i was planning on using with the 200LH pump it would be running for 20 mins.

So I've purchased a 12v 100Ah battery, a 20w solar panel and am in the process of building the box with all the electrics in it. I'm starting to look at how to build the power and charging circuit at the moment but I was wondering if anybody had some links to one they have used before?? Charging and solar power are completely new to me.

Cheers in advance Iain

With a 20w solar panel = 1.67 amps at 12v I think you would get away with just wiring the panel direct to the 100Ah battery - with a diode to prevent the battery back-feeding through the solar panel.

You will need something to warn you if the battery is not being recharged sufficiently.

...R

Cheers Robin2, I was planning on doing something to do with checking battery level I just haven’t thought that far in advance yet… My initial idea was to have the arduino check the battery level and display on the LCD screen, I just don’t know how complicated that will be. For the time being while I’m testing and tweaking I will just be checking the battery level with a multimeter.

Cheers
Iain

I think you need to sort out the hydraulic side first.

3w aquarium pump ??

IS that air ?

If its a circulating pump its highly unlikely to give you any lift , certainly not enough for a misting head.

1 M head is plenty for a drip system but solenoid valves are not normally much use at this pressure.

Servo valve , no. A rotary valve power to close and power to open is best, only uses power when operating, this also uses much less power overall than using solenoid valves.

If you can find a multi way spool valve that would save on x separate zone valves.

As an example one i built for similar use including 6 max misting heads required a 50 W pump.

EDIT

Pumps with this performance generally require a 24V supply.

One warning about allotments, batteries get nicked.

Cheers Boardburner. The way im planning on setting it up is the solenoid valves would be opened before the pump is run to spray the plants, and then the pump would be switched off before the solenoid closes. Would this not solve the issue of operating pressure. The misting heads are variable flow drip/misting heads, the misting capability isn't that much of a necessity, but I hoped they would distribute the water more evenly than just a drip system. Would you think something like this would do the trick.

Cheers for the thoughts about security, but her allotment is nicely secure in a nice quiet area.

Cheers Iain

looks ideal.

Using hozelock cone sprayer heads it should drive 3 or so .

One further thought if you put the arduino into sleep mode you can get quiescent current below 1 mA, you would need to change the on board regulator for something like this though.

http://www.onsemi.com/pub_link/Collateral/NCP4682-D.PDF

Not sure about the mega, you may have to disable other bits.

I used a nano.

Johnny010: So you may get away with a 10W panel, a 10W zener diode and a 10W Schottky Diode.

A similar project with very similar power requirements.

Solar garage door opener.

I got away with 2 x 7AH batteries and 2 x 10 w solar panels year round.

Biggest bugbear was the 7 mA drawn by the remote receiver which would have required doubling everything at least, but i got around it.