Hi!
I have a bunch of soil moisture probes that are attached to a device from a now defunct company. I can't find any information on them and don't see any data sheets or instruction manuals for this device. The form factor of the sensor looks very similar to a Decagon Devices EC-5 analog soil moisture probe. I have one such probe that works just fine using the ADC pins. When I wire up this probe the same way, nothing. I connected a multimeter and did find that the voltage goes high for about 1 second then goes low. I wasn't sure if this was just an initial pulse to begin digital communication so I bought a Seeed Nano DSO oscilliscope to track voltage over time. I connected everything up and was able to visualize the wave form. I have included it in the pictures. Immersion of the probe in water appears to have no effect voltage nor resistance doesn't change. The waveform looks exactly the same. I've included the pictures of the device and the sensor. Is there a signalling method that could be going on that I am not able to visualize with my oscilliscope? I have many of these sensors and it would be great if I could utilize them because the Decagon EC-5 is about 100.00 per sensor.
How about the frequency of the waveform?
Some capacitive sensors signal their measurements frequency encoded.
If the waveforms look the same PWM is ruled out.
It's not PWM. Maybe waveform is not the right word. If you look at the screen The whole sweep of that screen is about 12 seconds. Each peak is 1 second long. It has three wires. I connected the signal leads to red and white in this case. Since then I connected a lead to the non insulated shield wire and compared red to shield and then white to shield. One had a higher voltage than the other White/Shield reads a peak voltage of 2.8V the Red/Shield is 1.7V. Each connection has the same duty cycle as Red/White
Without seeing the waveform of the other extreme I can only guess, not even rule out PWM.
melicha8:
It has three wires. I connected the signal leads to red and white in this case. Since then I connected a lead to the non insulated shield wire and compared red to shield and then white to shield. One had a higher voltage than the other White/Shield reads a peak voltage of 2.8V the Red/Shield is 1.7V. Each connection has the same duty cycle as Red/White
That does not make sense to me and your condensed text block makes it even hard to read__*__.
With three wires I would suspect one being the signal and the other two the power supply.
Your picture does not show the wires of the sensor clearly.
Where did you get the sensor?
__*__compare the above to
Without seeing the waveform of the other extreme I can only guess, not even rule out PWM. -long-quote- That does not make sense to me and your condensed text block makes it even hard to read. With three wires I would suspect one being the signal and the other two the power supply. Your picture does not show the wires of the sensor clearly. Where did you get the sensor?
The device pictured is called an AquaMiser. The sensor's silkscreen print says ProMeter moisture sensor and I presume the company that built the AquaMiser is called FertileEarth.com. The product was designed for residential irrigation systems. It distrupts programmed irrigations from an irrigation controller(up to 4 valves) based on the sensor value and a user defined cutoff.
The shield wire is bare wire. In my experience I've never seen a bare wire used for anything other than shield/ground. Sorry the images aren't clearer. The forum wouldn't let me upload a larger image as an attachment
Mods feel free to move this thread to the device hacking subforum. It is probably the more appropriate forum for this.
Maybe the company went out of business because the sensors are worthless.
More clues may appear if you open up the sensor housing and have a close look at the circuit board - especially markings on ICs can be very enlightening. Waveform is quite unusual as in it looks like it's driven high, then pulled low (it's rising much faster than falling, and the falling edge looks like a cap discharge profile).
Do get waveforms at different moisture levels - my best guess is also that it's either frequency or pulse length that has the info you want.
Update. I destroyed one the the sensors to inspect the PCB on the sensor. The silkscreen was still on it with identifiers. It was indeed a decagon product. It was an Echo 5 Moisture Sensor. The sensor was an older design of the current sensor. I was able to get the pinout on the silkscreen as well. Once I tinned the leads and put it in the breadboard the sensor responded perfectly. Fired up on some free sensors!
