OK, so here is the problem with this design: the output voltage changes as a function of the parameter hFE of the on board transistor given a fixed impedance at the probe (hFE spec 70-300 but I've measured as high as 470) AND hFE is temperature dependent. What does this mean: The probes are not interchangeable given fixed software defined "moisture" thresholds; If you were to calibrate each probe at room temperature and you take a measurement at 100F it will read much "moister" than one taken at 72F. So if you squint your eyes and wave your hands you will see that the output voltage is proportional to hFE and inversely proportional to the probe impedance - simple fix: add a P-FET to short the probes in order calibrate before each measurement (this takes the temperature out of the equation given the next part), move the probe resistor to the base to give operating headroom for control of the P-FET and increase the probe resistor to a value of 10s to 100s of Kohms (small changes in the probe resistance + probe resistor result in a relatively constant linear change in output voltage over a large variation in hFE).

I'm adding 5 of these on one arduino, and planned on powering them from a digital pin. I am assuming that too much current for a single arduino digital pin. I am not super at electronics, so I was wondering if anyone had advice on an example circuit to power the vcc pin on all 5? I would normally go for a transistor as a switch, but given the vcc pin just goes to one probe pin, I am a bit uncertain how to wire that up, as usually a transistor as a switch powers the vcc to a device (such as an led) then a gnd comes out of the device and into the collector of a npn transistor, but that doesnt seem to work here... any suggestions?

Hi guys. I'm using this sensor. Do you know how I can calculate the percentage of humidity?

I have been using this sensor for about 3 months. The a/d converter is 1024 counts for full scale (3.3V) and I calibrated by taking a sample of my garden soil, weighing it dry then adding water to saturation. As the soil dried, I weighed the container and recorded the adc counts and the weight. I found that the sensor output curve flattens at low water values and high water values but is pretty linear in the range 800 - 880 counts (2.6 - 2.8 ).

In the Garden, the probe is powered only a few microseconds once each minute so I have not observed any corrosion issues.

From observation, it appears that the sensor output is also dependent on temperature (lower voltage for cool temperature; higher for warm). Consequently my garden generally gets watered near dawn when the temperature is lowest.

Has anyone else noticed this dependency? or have any temperature data?

It seems that on this product and many others the sale price has gone away? Was it supposed to last until the end of today?

Is it possible to use this with 3.3V outputs? I'm attempting to provide power with the digital i/o pins on a Particle Photon.

Check out my instructable : http://www.instructables.com/id/Thirsty-Flamingo/

The corrosion problem really is a thing. Not only does it kill your sensors in a pretty short time, but in my home garden the resulting electrolysis (I think that's what it's called) resulted in soil that was toxic to my plants. Nothing grew in the containers that I had the sensors in. The containers without sensors grew great. As has been suggested by the datasheet and other people on here, not constantly powering the sensor was critical for my success. My system was Arduino based, so all I did was power the sensor through a transistor that I enabled with a digital IO pin from my Arduino. The function that read that sensor powered the sensor on, read the value and then powered it back off. I think the sensor was powered on for less than 20mS over the course of an hour.

I'm not really that smart, and haven't played with RTCs yet (although I keep buying them...they look great in my part bin!) so I simply incremented a counter at the beginning of the MSensorRead() function. When that counter hit 50,000 (I think it was) I read the sensor, and ignored it the rest of the time.

Quick video I did showing the sensor reading/displaying: http://www.dagobah-system.com/node/31 watch closely next to the moisture reading value.

Kinda sad that SF went the resistive route. Capacitive (though a bit more expensive), has none of the corrosion issues of one like this. I built a similar sensor out of galvanized nails, and it rusted to nothing in about 3 months.

I understand ENIG is more expensive but the extended reliability is worth it, especially since no one will notice the "DO NOT POWER CONSTANTLY" annotation in the schematic.

I'm trying to characterize these sensors so I set up a little experiment with soil from my garden. I took a soil sample and dried it in the oven (before my girlfriend woke up and found out what I was doing in the kitchen). I let it bake at ~150F for an hour occasionally mixing, this seem to produce very dry soil. I made a Mariotte's bottle out of a water bottle to provide a constant yet very slow way to add water to the soil. Here is a quick picture of the soil in the bottom half of a Gatorade bottle with the constant flow water bottle on top. The issue I'm having is that the soil water content doesn't seem to correspond linearly with sensor output voltage. At some point it jumps up to ~4.7 V and stays there. Her is my recorded data of sensor output voltage vs time and the approximate moisture content of the soil vs time. Are there any soils experts out there that could chyme in, I'm quite happy with the little sensors I'm just trying to get the voltage to water content relationship established before deploying them outside.

Wouldnt a device like this make the soil sour through eletrolysis of the hydrogen? Ive thought of building many devices like this myself but i dont think its good for the plant. also if there is any copper, its very bad aswell. A simple copper nail can kill an entire tree ive heard.

I too am having trouble confirming that the output is as it should be. With the probe NOT inserted in soil, I get a reading of about 8. In dry-ish soil it was about 3300, and then in saturated soil it went DOWN to about 3200. Also as I slowly added water every so often, I wasn't get a nice consistent change in the reading. Can anyone comment on this? FYI i'm going into a Photo board. thanks

My electronics students noticed a sentence in the description, as of 2015-10-25: "The two large exposed pads function as probes for the sensor, together acting as a variable resistor." Sparkfun might instead write "The two large exposed pads function as probes for the sensor, together acting as legs on a variable resistor." The variable resistor itself, in this case, is the variably moist soil between the probes. Perhaps a completely different sentence would be appropriate.

Is there a max voltage with which I can safely power these? Working on a project that uses 12V external power and I'd like to rig up a few of these in parallel from the same power source.

nice hobby sensor ... probably not ideal for large commercial applications or sensors that will stay in the ground for many years ... but for 5 bucks you probably cant beat it ... most of the commercial ag sensors run in the hundreds (although I suppose you get better accuracy for that extra money, and support, and less corrosion, or replacements in the event of corrosion) Decagon sensors are very nice ... but cost alot more... (this looks like it was maybe based off of Decagons 10HS sensor)

Depending on the dimensions, some plaster of paris and a popsicle tray might make perfect sensorsicles. Not sure how much that helps with corrosion though.

I suppose one could use this with a Particle Photon and be able to get notification on my phone when the soil needs more moisture. Yes?

This sensor is just screaming "Connect me to an XBee!!!" - and I will if I can come up with a decent power supply for the XBee and sensor(s).

Could we get a little more information on "encasing in gypsum" to extend sensor life? I think more than a few of your customers are interested in this for controlling drip/sprinkler irrigation schemes for larger gardens and the info would be helpful.

Are there any data suggesting how long this board will actually survive in soil? I'm considering using some of these to help control the sprinklers in my home automation system. I don't want to go to all the trouble of running wires, etc, if it's only going to last a year... cheers, Doug

What is the danger of constant power? Corrosion on the probes? Looking at the schematic it's just biasing a 3904. I don't see the problem.