Let's Make Robots!

Moisture sensor.(Just another DIY)

Moisture sensor for your garden...so it measures moisture...

Headphone plug turned moisture sensor.

For a little while now I've been looking at how to do a diy moisture sensor that I could use for my garden project. There are a few designs, but there are drawbacks to each. For instance, one design uses galvanized nails or wire that you stick into the soil. Problem with these is that they are coated in a chemical, and of course will eventually corrode. Another design uses uses two wires encased in gypsum(plaster of paris). As the gypsum absorbs moisture the resistance drops. This is great but even at full saturation this may not vary by much. See this guys site for more info. http://gardenbot.org/howTo/soilMoisture/ So in comes my design. Since the sensor requires 2 metal probes or surfaces, I decided to try a headphone jack as I acquired several that a friend was getting rid of. Since I just started the testing of this little sensor I'm not sure how fast the probe may corrode, but I am able to get readings on fully saturated soil, partial saturation and of course completely desiccated. The circuit is based on this sensor with some slight resistor modification to cut down the current used. http://seeedstudio.com/wiki/Grove_-_Moisture_Sensor

Thanks to Patrick for pointing out the sensor above.

Voltage meausrements currently go from 0 to 1.65v when fully saturated(soil). Direct contact outputs around 1.85v and using fingers over the contacts gives around 0.4-0.5v and when giving it a good squeeze, around 0.8v.  The resistors I used were R1 = 8.5k and R2 1k. This drops the curent from around 14ma to around 1ma. I'll set this up to do constant reading and send that data via xbee to  a base station setup. I'm still considering how best to use this setup. A simple approach would be to just use a quad comparator, and when the level goes below a certain point, it'll trigger a signal to the proc that it needs water or something. I could also just keep constant measurements throughout the day....not sure just yet.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.


A shot of the probe after I added some water to the plant. Previuosly it was showing a reading of around 1.4v

The battery I'm using is a 3v lithium watch cell.

Let me go plug my headphones into my plant and see what it has to say. :P
I like the headphone jack idea. 

Redefining acoustic sensor technology since....now! :D


The slightly modified version of the circuit from sead studio....I just added a mosfet switch that allows me to enable the circuit when I need to get the adc reading.


The next part of this is to integrate it with a uProc adc input and save the data.This could be done simply with a picaxe and saving the data to the built in memory space, or even an arduino as there is a lot of code to interface with mem chips. I decided to pave new ground and continue on with the MSP430(to which I could not find any working examples). For the past month(off and on throughout actually), I've been pulling my hair on a MSP430 based setup that  appearing not to work, but in the recent days I have achieved success(write to and read from with correct data results). So with a bit more testing I'll have a working prototype that allows me to save all data to the mem chip and output to a computer.

Have you tried putting the headphone probe in a glass of water?  I've tried several moisture sensing circuits only to have the probes turn dark and frizzle after a minute in water.

Here's the only circuit I've found that doesn't corrode:


Built the circuit to run on 5VDC a few summers ago to control an irrigation solenoid -- basically it doesn't allow over watering.  It sometimes rains in Arizona.

Actually I did put in in water as a saturation test but noticed that the voltage was actually less than the max voltage I was getting while in the soil. I need to do some long term testing to see what the corrosion factor is and which types of plugs would work better(like the gold ones). Thanks for the link! I'll take a look at that sensor, it looks be an interesting switching circuit....as a side, I was considernger a modified hbrige with alternating controlled via 555 timer. Thats just a thought at the moment....