Hi, I was going to order some soil moisture sensors.
I am looking for one with the following features:
wireless WiFi/BLE/Zigbee
weather resistant (for instance IP65)
no base station & no cloud required
in addition strudy an discreet
Much to my disappointment I didn’t find any such model. Most of the sensors on the market look like indoor toys that wouldn’t even withstand irrigation water.
The only sturdy looking sensors I found are the Ecowitt WH0291 and the Gardena one, but both require a base station.
Do you know of any good solution? What do you use outdoor?
I’m using the Thirdreality Sensors about 2 months now. They work well & have proven water resistant so far. Measurements generally look plausible, it’s quite sensitive and I’m using it to control some plant irrigation.
Only issue that I noticed:
The measurements are varying by a lot depending on the temperature.
Two days ago the temperature dropped from 28°C to 21° within 2 hours. In this time the humidity measured changed from 64% to 68% without any watering. I solved this with some calculated temperature compensation. Now Zigbee2mqtt informed me about a firmware update for this sensor. Hope it doesn’t change too much…
I tested other alternatives before:
Two cheap capacitive Aliexpress sensors didn’t look warmer resistant and even under a roof, they worked only for about a week and decided to measure 100% all the time then - no matter how dry it is.
Another one from Aliexpress using 3 pins & conductivity (search for Haozee Zigbee soil sensor) is rugged & water resistant but measurements are useless: A 3 hour rain event of 7.3mm made it increasing the measured humidity from 35% to 39% which continued to be measured for the next 12 hours.
As far as I understand they are waterproof.
Mi flora with BLE moisture, pH and nutrition values.
I haven’t got them yet but a review says they work with HA.
The question is, what that sensor really reports. Not likely VWC, since 65% would be over saturated soil. If it’s trying to mimic relative humidity that could have temperature relation (like air RH has).
From what I know of plants, they can take any amount of water as long as they get enough time to dry out before they get a soaking again. Stick your finger in area where probe is. Of it feels dry then go with that reading.
Temperature under the soil if probe pushed down far enough shouldn’t change much from day to day but will over weeks espicially if measured just before sun has a chance to hit the soil.
This I plan to check myself.
That sensor doesn’t even measure the temperature of soil…
My point was not what it measures, but what it reports. What is the %-number it outputs…
It’s not VWC, so what it is…?
It’s probably not any scientific value, just some relative number that can have temperature relationship or not…
I was thinking of building something myself that was based on capitance and soil temp with an esp32. The OP has just sparked my interest.
I tend to know when to water my plants outside that are in pots when I am away from home. I currently use my cameras to see if it’s been raining and if plants are wilting. It tends to rain here a bit more, except when I go on holiday. Then there is a drought. I run a script to water my plants. Again it’s the gardener in me that wants to decide when to water.
That’s good approach. When you do it by yourself you decide what relationships you have. You can program your setup to output 666 for perfect nice humid soil for growing your carrots…
It’s not Zigbee and they don’t seem to mention the IP6x rating, but I’ve had 2 of these buried 6" deep in vegetable gardens for 2 years now with no issues. Connected to an ESP32 to send the data to HA via WiFI.
One thing to remember with all of these sensors, they are not scientific grade. In a smart home context they are just a guide or a trend. I use the Giex brand sensors. Per garden bed have a minimum of 3 sensors to get a mean value across the differences in the garden bed. A (mean) across all sensors as opposed to (average).
I’m using the sensor mentioned for potted plants (tomatoes, peppers etc.) under on my veranda. It has a transparent roof & is open to two sides. Even if the sensor seems to report the ambient temperature above the soil, it is enough for me to compensate the fluctuation in humidity measurements. I assume that the temperature in the pot might be around the same with a little delay.
Regarding the reported values: I don’t care much - as long as the variation helps me to prevent the plants from drying dead.
I’ve just put the sensor in the pot of the tomato that I assumed to be the thirstiest and adjusted irrigation in a way that the soil always feels moist, but no excess water is being collected in the coaster of the pot. The other plants receive a relative amount of water compared to the measured pot. Since I use a strange Gardena Pump & distribution system that originally is designed to run 1 pumping cycle / 24h, but runs a cycle whenever it’s disconnected from power and being connected again, it leads to many small cycles throughout a day. Since the plants grow fast, the number of cycles is increasing throughout the summer.
For me that’s good enough & I enjoy plenty of tomatoes, peppers & stuff.
And that is the only correct approach. The value sensor outputs is meaningful only by experimenting…
My point was that since the sensor tries to mimic some humidity %, temperature relation is not necessarily an error.
I’ve been using those ones for years and regularly splash water all over them when watering plants… still going strong. They do have a reasonable looking seal on the battery compartment which I guess has been good enough for water pouring on it.
