Office plant monitoring

[NetRap]

This sort of sensor is not good, because the electrolyticeffect.
You always should use soil sensors, which are based on the capacitive effect.

Like this one:
https://forum.mysensors.org/topic/4474/capacitive-soilsensor-for-measurement-the-humidity-of-flowers

Best regards,
NetRap

[Fat Fly]

My sensors is not good yes. I do not understand what is soil moisture % really. :( My pepper was overwatering. Very wet and moisture was 80% only.

[mfalkvidd]

@NetRap there are downsides but "no good" is incorrect. I have used these sensors successfully for almost a year.

[NetRap]

The problem is, that your design is based on conductivity.
This means, that the electrodes will be destroyed over time
and the produced metal salts are in your flower,garden,...

I don't will have metal salts in my tomatoes !!! ;-)

[mfalkvidd]

@NetRap I doubt that one second of 3V electric power every hour (0.03% duty cycle) will produce measurable amount of salt but yes, for food production capacitive measurements are probably better. The topic of this thread is "office plant monitoring" though and I hope you don't have to live off your office plants :)

[m26872]

"Metal salts" doesn't sound very dangerous unless you're more specific. I prefer some NaCl on my tomatoes.

How about some gold- or silver plated electrodes if you're of the worrying kind?

[Fat Fly]

I want to try electrodes from stainless steel.

[Lars65]

Normaly when you use eletricity and water, you need to sacrify something. For an example, when you cool a boat engine with water, you normaly use zinkanodes. I guess this is also the same when we want to messure humidity for an plant.
So, if you should use copper anodes for messuring, you could paint them with a zink color.
What it effects the resistance, I don't know.

[Fat Fly]

Maybe in scketch needed change measure time? Maybe need use function round ?

[LastSamurai]

@mfalkvidd Short question concerning battery life: I am using a custom board with an atmega328 and an nrf powered directly by a coin cell. When measuring temp/hum with something like an SHT21 the voltage drops are minimal (less than 100mV in more than a month). Using your sketch from post 25 here and connecting the fork to pins A1 and 2 (changed that in the sketch) I get a much higher power consumption. Any ideas why?
I guess the measuring takes too long or something? It also reports lots of 1% changes to my domoticz server... which is kind of strange. Shouldn't the threshold take care of that? Perhaps I have to use a higher value here...

[mfalkvidd]

@LastSamurai are you using the same sketch on both nodes?

The sketch in post 25 sends the moisture value every 10 minutes. There are no thresholds and no comparison to previous value.

Do the 1% changes look like a square wave? If so, the reason is probably that the pullup resistors differ from pin to min, even on the same atmega. A newer version of the sketch takes care of that, but I am still using the old version in my oldest node.

As we discussed around post 79, my node is dropping very slowly. It is currently at 3.117 V after approximately 9 months on the same 2xAA. The node started at 3.187V so that is a drop of 70mV, less than 10mV per month.

[Fat Fly]

Good morning people.
Can somebody help add to this moisture sensing sketch relay ?

[cattoo]

So does this work with 2.0?

[mfalkvidd]

@cattoo yes if it is converted.

[cattoo]

What needs to be added/edited so its converted?

[mfalkvidd]

@cattoo that is described in the instructions I linked to :)

[cattoo]

Its driving me crazy! Could you not convert this mfalkvidd! Its better to have your sketch on the front instead of that one there is there now.

[mfalkvidd]

@cattoo I won't convert it in the near future. I have too many projects that are more fun and more useful. My current setup works great, upgrading everything to 2.0 would mean a lot of work and no gain.

[Martin Tellblom]

After my other failed attempt I needed to try this.
I have a 5V Mini using A4 and A5 and a fork YL-69 and I get these result:

Direction: 1
Moist: 171
Direction: 0
Moist: 23
Direction: 1
Moist: 172
Direction: 0
Moist: 0
Direction: 1
Moist: 164
Direction: 0
Moist: 0
Direction: 1
Moist: 156
Direction: 0
Moist: 0
Direction: 1
Moist: 144
Direction: 0
Moist: 0

This is not the percentage value , its this value I print: moistureLevel = (1023 - analogRead(SENSOR_ANALOG_PINS[direction]));

Any Ideas?

[mfalkvidd]

@Martin-Tellblom could you describe what an YL-69 is?

How much time do you let the measurement settle after switching directions?
Which of the code versions are you using? There have been many revisions in this thread and they all work slightly different.