For those testing the upcoming sensors:

I propose that we change the settings a little so that the top end ( wet end ) is a little more sensitive. At the moment the sensors generally hit higher soil moisture percentages than you would expect if you put a finger in the soil, and for really moist soil they hit 100% quite quickly.

Does this align with your experiences?

Changing the settings would show a drop in percentage on your charts from when the sensor updates firmware.

Morning Chris, with a good watering mine usually goes from mid to late 40 to 55. I just drowned the plant and went to 82. Arguable this should be 100 but not far away.

Hi Chris, i would say yes. My reading only goes up to about 80 even when it’s freshly watered and soaking wet with filled water tray.

Thanks both. So much is dependent on how much physical soil there is in proximity to the sensor face, which relates to both the density of the soil ( eg clay vs loose compost ) and the amount it's packed down around the sensor. We're seeing if we can get further "reach" from the face of the sensor to reduce this effect too.

Probably agree now that I’ve given it a big soaking today

Great stuff, thanks Gavin. What I find interesting is how long soil takes to dry out below the surface. The sensors indicate that there's more moisture down there than you would expect, indicating that watering less would be OK. How does that fit with your experiences so far?

To accelerate drying, you could put the sensor near a larger thirsty plant. Can we send you another one for comparison testing?

I have a couple of stand-alone soil moisture sensors that also have a 'blade' that sticks into the compost. For those I had to callibrate them by soaking the pot so that it was saturated ( then press a button I think - it was a while ago ). Clearly that was not a good idea with a plant in the pot, but if one filled a suitable sized pot with the sort of compost in use at that point in time then calibration seemed to work. It might solve the problem of trying to make a sensor that can cope with all compost types and make sense for both the desert plant grower and the carniverous plant enthusiast. If one changed the make/type of compost in the future then one would just re-calibrate. That would only deal with desicated and saturated - I am not sure that it would be practical to ask people to also do a calibration reading with 'ideal'. But perhaps I am underestimating your customers?
Ease of use at the same time as real flexibility and high level of function is the holy grail of all systems, but not always easy to achieve.

Hi Chris - that's interesting. What are you doing with the soil sensor data? Using it for alerts, or to trigger other watering?

Not a lot I'm afraid. They are not capable of downloading the data and they are in a tiny glasshouse which is watered by hand and does not have any capillary matting trays. There are three of them wih a base station that displays the readings. The plants have different water needs, so I just monitor the moisture content of the compost to help decide whether I need to check that group of pots manually to see if watering is needed. The base station does not do alerts and does not cycle between the sensors, so it is not the most useful of devices. I have to pick it up and press buttons in order to check the readings. They also give soil temperature, which is handy in the winter.