PLEASE HELP!!!!!! Moisture sensor plan not working - or is it me :-(

Can anybody PLEASE help, I am in the UK and eager to grow melons next spring. I need to know when the roots of the tub need water so decided to make a moisture sensor. I found a ‘plan’ online, bought about £20 of components (I need two sensors) and have spent all afternoon TRYING to get it to work. One was connected via a circuit board and the other without. Neither work, the LED comes on on both versions but does not go out when the terminal probes are put in water, as they would when the plant roots are moist.

i enclose the original internet plan and a (poor) picture of my ‘‘air connection’ version....should this plan work? And if so, why isn’t it?

thanks in anticipation

sort by: active | newest | oldest
1-10 of 16Next »
ggom2016 days ago

This circuitry should work as described with a 100 ohms resistor in serial of led circuit.
When the probes connections are shorted together, led should go off.
(With E and C reversed, the 2222 should not light the led.)

You should soon see that working.

When that will work then the problem will be to determinate the necessary size of the probes.too small probes and the led will not go off, to long or too wide it will stay off too long.
Square inches and how deep is a matter.
I did construct a few of thoses things in the past, and the problem that came
very quickly is oxydation and corrosion of copper witch changes the results.
I did solve my problem using alternative current probing and some gold plated
pieces of circuit board as sensor.The type where the printed board is gold plated to make a direct insertion connector (HE901model).
(My circuitry is too complex, it uses a few very low power Integrated circuits and would not be cheaper than buying a brand new commercial one nowadays.)
If no gold plated available, you might try hight quality stainless steel .
It begins with a pair of knives, or spoons and then if needed, marine grade stainless parts are the best grade.(screws, bolts or anything)

TN1946 (author)  ggom2016 days ago

thank you, I have been busy for the last couple of days but hopefully tomorrow I will have another try - once again, thNk you all for tKing the time and trouble to help me. It is appreciated.

ggom20 TN194614 days ago


Okay you got it working !

For more light,You might try 82 Ohms instead of 100 ohms.

Or 3 Batteries witch makes 4.5v instead of 3 v.

Or a different type of led , Some make more light than other.

When satisfied with light,

Then experiment with the probes.



TN1946 (author)  ggom2014 days ago

the main problem is that a moisture sensor needs the default state to be LED off then the LED should come on when no moisture is shorting the probes (plant needs watering). And this cor unit is doing the opposite.

TN1946 (author)  ggom2014 days ago

Hi, Thank you GGOM20 and Iceng for helping me!

Please see attached picture, in this circuit the LED is on (pretty
dimly) by default and when the red probe leads are shorted, as they would be by
moisture in the plant pot, the LED goes off.

The problem is that a moisture sensor needs the default state to be LED
off (otherwise battery will be flat) when the probes are shorted by sufficient
moisture (plant doesn’t need water) and then the LED should come on when no
moisture is shorting the probes (plant needs watering).

Am I nearly there or shall I ‘bin it’?

Thanks again

iceng19 days ago

I assume the wires going into the soil are bare wires unlike your picture which shows only insulated red wires...

This circuit is a water conductance detector... As water dissolves ionic materials in the ground the resistance between the wires decreases and eventually draws current away from the NPN transistor base which turns off the LED...

The circuit is deficient and needs that 100 ohm resistor in series with the LED.. Because you might draw too much current through the LED and damage both LED and NPN...

BTW the distance you separate those wires makes a big difference in how wet or dry you detect... And local water or filtered water can affect it too... Stay with one water source...

TN1946 (author)  iceng19 days ago

Hi iceng, thanks for helping, the ‘circuit drawing’ which I am following has the LED turned OFF until the moisture at the bottom of the container gets too dry - then the LED and a buzzer will come ON. I intend to put two 18ins long plastic terminal posts inside the header box, then run the red terminal wires inside these plastic tubes and then solder the ends of the leads onto galvanised roofing felt nails at the end, push the nails back up the end of the terminal rods and seal. So that the flat galvanised head are the sensor tips.

I have found an error in the drawing though - in the coloured drawing that transistor the NPN 2N2222 is shown as being, from left to right, looking at flat face, CEB....but in actual fact the tech info below says it is EBC(???)

iceng TN194618 days ago

Exactly, in free air the LED is always ON, then when you touch the probes
together the LED turns off.. I have submitted two images of the TO-92 NPN Transistor pinouts which agree ! This is a good time to point out the amount of Fake Information that is available out there.. Your pic was not detailed enough to assign curvature of the TO-92...
You need solid non-oxidizing nail shaped probes held at a fixed distance apart.
A zinc coating will be removed by acidic soil in an hour...
Copper comes to mind if a heavy gold plated probe won't do...
An Effective probe is how much area is exposed to moisture and where that moisture is important which is the plant root ball..
Suffice it to say that soil measuring surface resistance is not helpful
when moisture content in root water is your intention..

TN1946 (author)  iceng18 days ago

thanks again iceng, I will find copper tips.

Just to moist soil I want LED off and when dries out I need LED (and buzzer on). Do you think the coloured picture circuit which I posted is viable, because it iis not working. Thanks again

iceng TN194618 days ago

Add that 100 ohm resistor on either side of the LED then touch the leeds together see if it lights..

If it does not light, the LED or NPN cold be backward or damaged.

1-10 of 16Next »