Hints for outdoor soil moisture sensor

Other Hardware Home Automation
1

Hi.

I want to measure the soil moisture in a raised plant-bed. My problem ist, the "standard" capacitive soil moisture sensor is not outdoor capable, and also measues the moisture at the surface only. I'd like to measure the moisture at ~50cm depth.

does anyone has recommendations / experience?
The only sensors I have found are:
Pino-Tech SoilWatch 10 (quite expensive with over 20€),
DFRobot SEN0308 (expensive, and cable a bit too short)

Both sensors I would have to bury completely - not sure if they would make this over time.

Greetings,
arnschi

2

Have you considered making your own? They are not that hard to make. Simple thought, two metal probes with shrink tubing on the outside allowing it to be insulated until the tip is where you want it. Stainless Steel should work OK. Also be sure to use an AC or polarity switching circuit on the sensor.

1 Like
3

As @gilshultz says you need to use an alternating polarity and "inert" material for your probes to prevent electrolytic corrosion.
I made one that uses that technique and it works really well. You may need to fiddle with the dimensions of the probes (ie exposed area and spacing)
Its desribed in detail here

You could also consider using a four point probe system

2 Likes
4

Two welding rods with insulation down to the 50cm mark, connected to any cheap soil moisture tester, so you get an appropriate signal.

5

Two Carbon arc electrodes.

image

6

I covered my capacitive soil sensors electronics and connector with silicon and buried them. The MCU is in a nearby by box ran by solar. I run a few sensors per MCU.

7

How did you calibrate it? Or do you just use it for a qualitative indication?

8

After soldering in a new wire length, I hook the sensor up to a ESP32. Once hooked up I hold the sensor in the air, wipe off with a paper towel to get the dry reading of the sensor. Then I dump the sensor into a bowl of water to get the 100% wet reading. Using the 2 readings on could then use the map() function or some other math thingy to create a % of moisture scale for the sensor.

1 Like
9

Its not simple to get an accurate result.

You can calibrate by volume or by weight. (VWC or GWC)
However by weight you need to know the density of the media.
Do you consider saturated medium as 100%?

1 Like
10

Indeed - hence the question! :stuck_out_tongue_winking_eye:

In most cases, a qualitative result is probably sufficient

11

So what's wrong with @Idahowalker's solution?

12

Which has been working for +3.5 years.

14

Well as usual it depends on what information you need and why. A few LEDs would give you a qualitative result - too dry, ok, too wet.

1 Like
15

* silicone

Which seems to be all that Pino-Tech and DFRobot have done with their "waterproof" sensors (mentioned in the OP):

image
image782×271 31.8 KB

image

16

Out of interest, in what orientation do you bury them - horizontal? vertical?

If horizontal, is that parallel to the ground surface, or perpendicular?

If parallel, does it make any difference which way up?

17

I have buried my soil sensors in my raised beds about 4 inches below the top of the soil so that the sensor is closer to the root system.

To determine which way is up I use the Right Hand Rule. https://mathworld.wolfram.com/Right-HandRule.html

18

But do you bury them vertically:
image

or horizontally, like this:
image

or horizontally, like this:

image
image990×135 5.05 KB

or horizontally, like this:

image
image954×144 4.41 KB

19

like this

20

There is nothing wrong with measuring % of water content within any medium. Even if it is a slow release medium, which can be measured as change in %water over time.

As Idahowalker has pointed out calibration is easy enough for % water, all you really need is a soil sample to find out the relative change over time with natural drainage to get a calibration curve.

A simple method would be to fill a large plant pot with the soil, saturate it, then take data samples at regular intervals, not forgetting to keep the pot in a place where you know the temperature. You can expand on this for different levels given you have hydraulic conductivity data to consider, but that would be for more professional use.

Alternatively you could just rely on a linear change between 0 and 100% for a very rough idea of water content.

My approach for the garden, stick moisture gauge into soil and let the watering system, water when it is too dry. :slight_smile:

Not forgetting that roots need air too, so letting the soil dry out a bit is good for the plant, as long as it is not left without water for too long. Read up on how hdroponic systems work for the air water cycle.

21

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.