Jump to content
Electronics-Lab.com Community

Question on the Plants Watering Watcher circuit


Recommended Posts

To All,
What did I get myself into?:

"Hey, what's that thing? For the plants? Cool, can you make me some?"

Do you know how many times I have heard that in the past few days? This dumb project is becoming very popular with my friends. ;D ;D

A few minor bugs: ::)
1) When carrying the project with the probes bare, the LED is fairly bright. But occasionally, when the probes are inserted into very dry soil that happens to have a resistance that is close to the setting of the trimpot and its series resistor, the LED turns off. It stays off until the resistance of the soil changes a fair amount (or if you remove it and re-insert it).
This bug is caused when the resistances are equal, because that is when the signal at probe P2 is cancelled. When the probes are bare, the signal at P2 ramps to logic high and low levels. But if you suddenly cancel that signal at a moment when P2 is low, then P2 never gets high enough to trigger its gate's high Schmitt threshold voltage (1.5V max. with a 2V supply), so the gate stays off.
This bug rarely happens, maybe once in 20 tries. I can't easily fix it.
2) When the resistance of the soil equals the resistance of the trimpot and its series resistor, the LED is bright. But when the resistance is much higher or the probes are bare, the LED dims a bit.
This bug can probably be fixed by reducing the value of the capacitor at P2. Then the occurance of bug #1 above happening will be even rarer.
3) The LED really does dim a lot when the battery runs-down to 2V. The LED's current drops 77 percent because the IC gets tired with only a 2V supply. An added LED driver transistor will reduce the dimming to only 33 percent, which is barely noticable.

Link to comment
Share on other sites

  • Replies 98
  • Created
  • Last Reply

Top Posters In This Topic

Hi Guys,
After making a bunch of this project but modified my way, I have realised that it is as boring as watching grass grow. It also eats the two AAA alcaline batteries (AA are too big and heavy) in about 3 weeks.

So I have decided to spice it up a notch, save battery power and give a low battery indication:
1) Instead of the LED being on continuously, make it flash at 1Hz. It will still show the dimming effect, where wet soil is very dim then gets brighter and brighter as the soil dries.
2) Make the duty-cycle of the flashing low, so that the LED is lighted for only about 50ms-100ms each second. This will allow the batteries to last 10 to 20 times as long.
3) Make the duty-cycle change when the batteries are low. New batteries will make a 5/95 to 10/90 duty-cycle and low batteries about 60/40.

It is easy to do all these things by using a spare NAND gate as the flash oscillator, addind 2 transistors as a NOR gate and LED driver, adding a resistor and diode for the duty-cycle and allowing the voltage-drop across the diode to control the duty-cycle change at low battery voltage. It fits on a Veroboard the same size as the dual AAA cell holder. The revised circuit is here:


Link to comment
Share on other sites

Hi guys,
My "flashing LED" revision that I made to this project works perfectly. ;D ;D
Since the LED is powered for such a short time, 2 AAA batteries will last a long time. The flashes are bright when the soil is dry, attracting your attention before the thirsty plant wilts. It even flashes through wilted leaves that might hang in front!
I added a resistor across the probes to reduce the modified project's soil-resistance range, and to fix the problem that happened without it, when sometimes the LED stayed off if the probes were suddenly given a soil resistance that equalled R2 (cancelling the signal at P2).
I'll post this flasher as a new project.

Link to comment
Share on other sites

As a soil moisture indicator, two 1000mA/hr (to 1.0V) alcaline AAA batteries should last a couple of years! The PWM of this circuit is logarithmic, so when the resistance of the soil is half-way to being very dry, the power to the LED is not also half, but actually only 1/10th, which looks like about half-brightness. Most of the time the resistance of the soil is much less than that, therefore the average current is very low. After all, this circuit is CMOS.
With new batteries and very dry soil, my circuit pumps 35mA through the LED, so it is very bright, but with a (flashing) low duty-cycle to extend battery life. The original circuit put only 1mA through the LED, which was very dim, and that was at its brightest.

Link to comment
Share on other sites

  • 4 weeks later...

hihi, sorry for asking u a favourite, :), can you send it to me, to my mailbox, [email protected], 'coz i'm scare that the admistrator of the page late to post ur new projects, this wednesday i need to pass up the projects with reports, so, i need it asap. But anyway, if u felt tat is not very conveniece for you to do that, just forget about it. ;D

Hope u will send it to me.....


Link to comment
Share on other sites

  • 2 months later...

Hi Olihou,
Welcome to our forum and thanks for the compliment.
My circuits have been blinking (I let my plants dry-out before watering again) for over 3 months with the original AAA battery cells and the 3.0V has dropped to about 2.7V. I'll replace the cells when they reach 2.0V.
The LEDs are probably down to near half-current now, so the next 0.3V of battery voltage drop will take 6 months. When dry, the LEDs seem to blink just as bright as with new batteries since the eye's response to brightness is logarithmic and therefore half the current through an LED looks almost the same.
The plants with the circuit grow better than those without, so it helps them. Maybe the 2KHz through the soil stimulates their roots. My ordinary tinned copper-wire probes have not corroded.
If you use the really bright wide-angle LEDs that I recommend, you will like it very much.

Link to comment
Share on other sites

Dear Audioguru

I have succeeded in building the unit according to your design. It is working exactly as you have described. I am completely satisfied with the circuit, except that the brightness of LED does not seem to give a clear indication of the level of dryness. The LED I have used is of ultrabright type, the brightness of which changes very little when the soil resistance (simulated by a trim-pot) varies from 1 kohm to 100 kohm and to infinity. :-[

I have hence realized the beauty of the design you have proposed im May:

"My new idea is to have a flashing LED indicate soil dryness. When the soil is wet then the LED is off. As the soil dries, the LED begins flashing slowly. As the soil dries even more then the flashing-rate increases. When the soil is very dry then the flashing-rate will be so fast that the LED will appear to be on steadily."

Are you developing a circuit like this, or will you? :D

Thanks and best regards,


Link to comment
Share on other sites

Hi Oli,
I am glad that your project works well.
Is your soil different from mine? I have designed the circuit so that very wet soil that is only 47 ohms (with about 50mm of the probes in the soil and spaced 0.7" apart) turns off the LED. Very dry soil that is 100K or more makes the LED flash very brightly.

One of my plants was fairly dry yesterday and its circuit was flashing fairly brightly. About 1/2 hour after watering it with a weak fertilizer solution (ordinary water works the same), the LED turned off. The soil was damp, not soaked. This morning the circuit is flashing dimly and the soil is just right, not damp nor dry.

I tried resistors just now:
47 ohms = LED is OFF
100 ohms = extremely dim flashing, seen only in a very dark room
1K ohms = very dim flashing that is clearly visible in a brightly lit room
10K ohms = flashing that is not dim nor bright
100K ohms = almost full brightness, very bright flashing.

Different LEDs vary in the way that they shine:
1) Some are linear only at very low current, up to about 2mA, then "saturate" so that 4mA gives only 1.8 times more light than 2mA, 8mA gives only 3.0 times more light and 16mA gives only 4 times as much light instead of 8 times like the current. Yours are probably like that.
2) My LEDs have a linear brightness increase with increasing current so 4mA produces exactly 2 times as much light as 2mA, 8mA is 4 times and 16mA is 8 times. The max (new battery) peak LED current in my circuit is about 39mA and the LED's voltage is 2.15V. With a fairly dead (2.0V) battery, the max peak LED current is 8mA and the LED's voltage is 1.8V. Of course the 2KHz PWM pulses at the 2Hz flash rate have a very short duration so that LEDs with a max continuous current rating of only 30mA are not stressed.
3) Other LEDs get more efficient with increasing current. 20mA produces 3 times as much light as 10mA, etc.

Can't you purchase or get samples of the LEDs that I recommended?

I was dreaming when I thought of changing the flash-rate with changing resistance. It would be fairly easy to make it flash very slowly when the resistance is high (dry soil) and flash quickly when the resistance is low (wet soil), but that operation is backwards and it would be too complicated to reverse it.

Link to comment
Share on other sites

Dear Audioguru

Thanks for your prompt reply. I will check the performance of my LED. It is one with unknown part number, just picked up from the shelf.

I agree that as you said "It would be fairly easy to make it flash very slowly when the resistance is high (dry soil) and flash quickly when the resistance is low (wet soil), but that operation is backwards and it would be too complicated to reverse it.", or in academic terms, it is easy to build a "resistance-to-delay" circuit but not a "conductance-to- delay" circuit".

Having said that I am still dreaming for a not too completed circuit for the latter -- am I too much a perfectionist?

Nevertheless I would think that the present circuit is already good enough for the practical use. After all one's main concern is when to water the plant, and a rough idea of the soil dryness is just sufficient.

Thank you. May I come back to you after further experimenting.


Link to comment
Share on other sites

Hi Oli,
In addition to using LEDs that have linear brighness vs current, make sure that they are wide-angle so that they can be seen when not pointing directly at you. Many "ultrabright" LEDs have a very narrow beam.

Another problem with having a reverse resistance-to-flash-rate circuit would be its high average current consumption when the soil is dry. Do you want a circuit with big and heavy D cells? (A potted plant toppler.) Or requiring replacement of fairly expensive AAA or larger AA cells every few days?
My existing circuit draws very little current, even when flashing brightly, so that is the way it will stay.

Link to comment
Share on other sites

Dear Audioguru

Thanks for your comments on the LED. Mine is really one with a beam of very narrow angle and I will look for a replacement.

I think that current consumption might not be a problem for the resistance-to-flash-rate option, as we could control by design a low flash rate to reduce the average current. For example, from 1 pulse per 5 second for very wet to 2 pulse per second for very dry (this is a range of 10 to 1).

Do you agree?



Link to comment
Share on other sites

Hi Oli,
I'm not talking about us electronic genius's, but the average (stupid?) person probably won't be able to comprehend a flash rate change of only 10 to 1. Besides, a flash rate of 5 seconds is so slow that you could walk right by and not see it between flashes.

I designed a communications system for the government. (Lots of stupids there!) I had one type of warning signal make a low frequency beep slowly, and another warning signal make a high frequency beep quickly. They complained that there was no difference! ;D

Link to comment
Share on other sites

Dear Audioguru

I have obtained the following measurement results:

Ohm between probes On Pulse of LED Battery Current
---------------------------- -------------------- -------------------
47 Nil on scope 0.6 mA
100 0.3 uS 0.6 mA
1 K 1.5 uS 0.7 mA
10 K 14 uS 0.8 mA
100 K 100 uS 1.0 mA
Opne circuit 600 uS 2.0 mA

Note: WIth R1=560K and C1=.002 uF, the operation frequency is around 800 Hz, i.e. T=1.25 mS

R3=12 Mohm, R4=100K, giving a pulse of 30 mS at 1 pulse/2 seconds

My observation is that for

47 ohm and below, no visible light from LED
100, a little bit light
1 kohm, easily visible
10 kohm, brighter
100 kohm, even brighter
open circuit, very bright

My feeling is that if, in actual use in soil, the soil resistance falls in similar range as above, then we can tell from the brightness the degree of dryness quite confidently. However, if the range is only from 100 kohm and above, i.e. the LED is seen quite brigt to very bright all the time, then it would be a bit difficult to tell the degree of dryness. Of course this will very much depend on the separation distance of the probes, the sizes and shapes of the probes, the chemical composition of the soil ......

This should be my next steps of experimenting with this wonderful toy !

Comments are welcome !

Thanks and best regards,


Link to comment
Share on other sites

I added a resistor across the probes to reduce the modified project's soil-resistance range, and to fix the problem that happened without it, when sometimes the LED stayed off if the probes were suddenly given a soil resistance that equalled R2 (cancelling the signal at P2).

Hi Oli,
Your measured pulse widths and battery current are a little more than twice what I measured, which is due to your change of the oscillator's frequency from my 2KHz to your 800Hz.
Your pulse width keeps increasing with an inter-probes resistance more than 100K, which makes me believe that you did not install R8 (47K between the probes). My quote above shows a problem this will cause, but your timing has to be just right for it to happen (a resistance of 100K between the probes exactly at the moment that Probe 1 goes low). Without R8, my circuit had that problem occur once or twice if I randomly inserted the probes into 100K soil ten times.
Another problem happened without R8. As the soil's resistance gradually rose to 100K, the LED's brightness kept increasing. At around 100K, the brightness suddenly dropped, then didn't rise any more as the soil's resistance increased to open circuit.
Adding R8 fixes both problems and reduces the battery current (and therefore LED brightness) with a very high resistance between the probes.
Link to comment
Share on other sites

Dear Audioguru

Thanks for your comments.

I hope you do not mind that I have not follow your design faithfully. I have just used parts readily available. Hence the electrical performance is different from what you have built. :( However functionally I trust it is as good as yours. :)

As a matter of fact, I do have R8 in the circuit. I have used 56 kohm instead of 47 kohm.

I do not have time yet to do what I intend to do: to study and hence to design the probes so that the equivalent soil resistance would fit the range of the circuit best, and so that the change of brightness of LED would reflect closely the condition of soil.

I will report to you and all interested in due course.

There is one more thing I intend to do: following this "watering watcher" project, I would like to build an automatic "watering" device. In fact this is what I have dreamed of before the very fortunate encounter with you guys at this wonderful fourm ! :D

Best regards


Link to comment
Share on other sites

Hi Oli,
I have been thinking that my circuit probably won't work if the plants are watered with distilled water, and the use of organic instead of chemical fertiliser. I will try it with synthetic soil too.
I was amazed at the low resistance of my plain and activated-charcoal-filtered (no difference) tap water.

You have a good idea to make an automatic watering system. The outdoor ones that I have seen operate only when it is raining! ;D

Link to comment
Share on other sites

Hi Audioguru

In my place I am given to know that commercial "distilled water" is not trully fully distilled -- there are still minerals such NaCl as ions which makes water conductive. Even if it is 100% distilled and free of minerals, there are still likely minerals in the soil.

So there is little chance that out circuits would not work -- unless you do use 100% distilled water AND the synthetic soil you have mentioned (What actually is it? I do not know if there is anything like this in my place?) ;D


Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Create New...