Automated water level monitor/ irrigation system
 

It's been a while since I've posted so I thought I'd share some of my experience designing and building (well prototyping at this stage anyway) an automated level monitoring/irrigation system.

I recently purchased 3 acres of land next to a creek in an area that has a decent amount of rainfall (about 1.1 m annually). I want to build a shed on the block initially and vegetate certain areas with endemic native plants as it backs onto a creek line and would've been a riparian zone in the past which was cleared over the last century for pasture.

I won't be drawing any water from the creek. One, because these days its legislated against and two, the last thing I want to do to an already degraded bank is take out what little there is. If you knew how intermittent the creeks in Oz can get you'd understand what I mean, not to mention the utter paranoia the medias invoked in the population about water usage combined with legislating out everyone who wants to use it anyway.

Suffice to say I won't be drawing from mains either, there is none.

The solution was to source two large rainwater tanks each 24,000 L or 6500 gal, 13,000 gal total. At first I actually thought this was overkill, that I couldn't fill this volume compared to my shed area. But then I calculated it and realised I could fill it three times over in a yearly period. Problem solved. Price was right at about AU$4,500, too easy.

I like Donald Rumsfileds saying about there's things we know we know, and theres things we know we don't etc, so my next issue is the fact that I can't be there all the time, but know how much water certain young natives need and don't have the luxury of mains power...

So I set about designing a solar system that runs on about 20 to 40 W of power and is gravity fed - no pumping costs, there's a 3 m slope on the property to the lowest point.

What I wanted this to power were a few motorised valves that I've built in the past, a handful of water sensors/conductivity meters, one of which senses tank level, the others as inground sensors and a simple voltage regulator to charge a bank of 2 70 AHr 12 V deep cycle batteries and a 7 AHr sealed lead acid.

To top it off I'll be building it around a simple HD44780 2 line x 16 character LCD module for tank volume, battery Voltage and watering regime display.

My next issue was generating a sine voltage because of electrolysis at the probe, again there is no mains to use, besides 50 Hz is next to useless for this application. A simple Wein bridge oscilator was the solution using a TO72 quad opamp connected to the microcontroller via an 8 channel analog multiplexer, which will use four of the micros pins - I only have 16 pins in total, five A/D inputs, two of which are for the batttery/PV voltage. I could buy a chip with more pins and about 4K of memory, but it would be overkill, I'll only need 2 K of code space and could even at a pinch, make it 1 K, so I'm sticking with a 16F819 micro.

So far I've got the circuit diagram roughly in my head, built the oscillator, ordered/received the chips and am writing the code to interface the LCD and the micro using a couple of pins with a 74LS615 shift register to shuffle out the data

Thats pretty much it, in total about 4 chips costing less than $15 in total and a handfull of passive components. The most expensive part was the 20W PV panels which I picked up for $200 each and the high torque DC motors to turn the valves, though they're only $20 each.

This system will monitor how much water is in the tank, how much rainfall over the past few months, the battery charge state and set up a watering regime depending on soil moisture. I'll never need to be there and can come back in six months to some decent growth, thats the plan...:emotions16:

Re: Automated water level monitor/ irrigation system
 

Dang -

My system is to buy food stamps

From the disemployed for 20 cents on the dollar,

Then trade them in at full price for fuel for the genny

Which pumps the well.

Needless to say I am being extremely cynical here............

scyth

Re: Automated water level monitor/ irrigation system
 

Wow -

A poor attempt at zen poet

Pray tell what drives this thinly veiled cynicism

Masking bombastic buffoonery lampoonery

Actually I don't even care

:111: :cry1:

aeon

Re: Automated water level monitor/ irrigation system
 

Interesting system. Good to see you posting again.

Re: Automated water level monitor/ irrigation system
 

can we see some pics?

Re: Automated water level monitor/ irrigation system
 

I will get some pics as I go along, at the moment though its still on a breadboard, (I'm back at work and have had trouble finding the time) but I'll be happy to help out anyone who wants to build a similar system, might even supply limited assembler code for the PIC if need be, or help someone to learn and access the peripheral features like the timer and the ADC.

Quick note about the components: the shift register I used for the LCD was a 74LS164 (not 615, my mistake) and the op amp was a quad JFET bipolar TL074, not the dual TL072 as I initially stated.

Also, I've decided to add a wireless link between the main unit and the conductivity sensors. I sourced some low cost AM transmitters and receivers for about $20 in total, they have a range of about 50m, but at a push I may be able to get them to transmit over that to 60m depending.

They are Quasar QAM-TX1 and QAM-TX2. I decided to forgo the encoder/decoder chip set and will use the PIC to handle all that stuff.
The data format will be Manchester encoding to keep it as simple as possible, besides bandwidth isn't an issue nor is speed, they will just transmit the ADC results for conductivity from one micro to another at the right time...yeah, the right time...

...this was difficult to figure out because I want the main PIC (receiver) to go into sleep mode when it is not performing any functions to save on power, so I set up a routine to wake roughly every 7.5 minutes (goes evenly into an hour) but how is the transmitter supposed to synchronize with the receiver if it is going into sleep mode?

I decided that because the watering will happen at night when evaporation is at a minimum that this would be the time to transmit the data. So I'll synchronize the two PICs with an LDR (light dependent resistor) which will both change at roughly the same time as the sun sets, after which I'll set one to transmit for a 10 minute period and the other to skip the sleep function until the data has been received, I can then guarantee a window where both micros will be awake at the same time to send and receive data, this will only use up two more pins on the main PIC, and I have enough.

It might sound complicated but it is actually a rather simple system, most of it focuses on setting the assembler routines to perform as required.

We talk a lot about self sufficiency on GIM and here is one way to set up a watering system for food based production that

1) requires no pump, hence no excessive power requirements other than about 20W, easily achieved with a small PV, and

2) Is fully automated

3) can be controlled from a central location, say within a house without the need to dig up the ground and place wires everywhere.

4) Is low cost, comes in at around AU$350

5) Can be maintained by the operator because of the simple design, and easily upgraded or fixed at minimal cost.

As far as I'm concerned we missed the boat in regard to computing power, we have these huge 64 bit juggernauts that can feed images 24 hrs a day and yet a simple 8-bit micro that costs less than $4 and a 20W solar panel ($200) can feed your family for along time with just a little bit of knowledge and some planing.

Next time I'll start a thread to discuss my ideas for an automated aquaponics system which uses 10% the water of regular horticulture, and how to build an oxygen sensor from a BIC biro pen and some conduit.

:wavey: