Calibrating Soil Moisture Sensors for Irrigation
Вставка
- Опубліковано 6 сер 2024
- Capacitive soil moistures sensors are inexpensive and fairly easy to program for determining "wet" and "dry" soil conditions. However, arbitrary sensor thresholds are not adequate for determining irrigation start times or permanent wilting points in response to soils losing water to evaporation over time.
This video summarizes an experiment to see if capacitive soil moisture sensor response may be fine tuned by carefully measuring volumetric water content of a silty clay over time and comparing results to published recommendations.
The spreadsheet referenced in this video is posted here : docs.google.com/spreadsheets/... . This contains additional references and worksheets that might of interest to tinkerers - feel free to double check my calculations or use it to run your soil experiment ;)
Here's the paper that I studied in detail-- see Table 1 sourced to Ratliff et al. (1983); Hanson et al. (2000) used for deriving that multicolored chart for predicting VWC thresholds:
1. "Understanding Soil Water Content and Thresholds for Irrigation Management" extension.okstate.edu/fact-sh...
These are the papers for Table 1 referenced in my primary source:
2. "Field Limits of Soil Water Availability as Related to Laboratory Measured Properties"
www.ars.usda.gov/ARSUserFiles...
(My sincere apologies to ARS and Ratliff for the mispronunciation in the video.)
3. "Monitoring soil moisture helps refine irrigation management"
calag.ucanr.edu/Archive/?arti...
Here is another excellent reference from the UA Cooperative Extension
4. "Methods of Measuring for Irrigation Scheduling"
itc.tamu.edu/files/2018/05/az...
5. The multicolored chart in the video is sourced to a commercial irrigation sensor website, but I lost the link for proper attribution :( It sources its data to Ratliff which matches the numbers in Table 1 in my primary reference, so the attribution should really go to Ratliff and Hanson.
Note that it's been about 25 years since I took a soil science class. To be clear, this video is meant to share my exploration and lab notes on a topic of interest to me personally. Any corrections or suggestions for improvement are most welcome via the comments section. - Наука та технологія
Outstanding video. Thanks for sharing your explorations in calibrating DIY capacitive soil moisture sensors. We are long time users of Teros 12 sensors and consider them the gold standard for capacitive (indirect) volumetric soil water content measurement.
Thanks and yes- the Teros sensor is what my colleagues at the University of Arizona are using. I wish I could get my hands on one to explore- a bit too pricey though for a backyard citizen scientists. I still think there’s a lot we might determine from temperature as an affordable analog : ua-cam.com/video/4yfhJRE4qDU/v-deo.htmlsi=IPuJ9EaXba2qnj67
@@ModestMaker Dielectric permittivity is a decreasing function of temperature. Generally for subsurface probes we don't take this into account as diurnal soil temperature does not fluctuate to the same degree as air temperature. To estimate available water capacity including the permanent wilting point, the soil water characteristics calculator may be helpful to you www.ars.usda.gov/research/software/download/?softwareid=492&modecode=80-42-05-10
Estimates of percent sand, silt, clay, bulk density, organic matter, and salinity for your soil are input values. Bulk density is easy to measure, percent sand, silt, clay can be obtained from USDA-NRCS Web Soil Survey for your soil type (preferred) or estimated via column sedimentation. Since you're located in an arid region salinity (saturated extract) may be important. Check with Web Soil Survey or measure this with an inexpensive meter. And good luck with your adventures in capacitive soil moisture sensors.
Thanks! An ARS calculator for wilting point- I love it. Agreed- not too much temperature variance at depth, but definitely a diurnal signature a few inches down. io.adafruit.com/biod101/dashboards/soil-experiment-ambient-shade. Maybe the first few inches can tell me something, although I agree rooting depth is the key. It’s fun to measure all this stuff anyway. ua-cam.com/video/1fKoxlAXaAU/v-deo.htmlsi=pQZ5eV6Q4TPdMJfG
Nice Video and a lot of informations. Thanks for your work
Glad you liked it!
Excellent video, thank you for the time to produce the experiment, document it, and put up a video of your results. I am trying to come up with a sensor system to help indoor watering, specifically in the winter. I have done a ton of reading, youtubing etc, and so far, not found something that produces consistent results over longer periods of time. Resistive sensors deteriorate too quickly, capacitive sensors have promise, but the trick is to find a high quality one. There may be other solutions, but so far, I have not found any.
For what it's worth, I've been working with the University of Arizona on an irrigation control system that uses Teros 12 sensors (metergroup.com/products/teros-12/).
Working with a student, we managed to get these to interface with an ESP-32-S3 for controlling a suite of irrigation control valves (solenoids) at a research site. The Teros 12 sensor supports the SD1-12 communication protocol for which there is an Arduino Library.
Since the code and project are not my own, it wouldn't be appropriate for me to share that on my channel but I will confirm it's doable. I have reached out to Meter Group asking if they would donate a sensor so I can produce a formal tutorial on how to do this for the benefit of others, but they declined. The issue is the sensor is costs about $258 which is something I can't afford, but it is an alternative you might explore.
@@ModestMaker the Teros 12 sensor is highly unlikely to pass the better half approval test. But I will check into it. I was planning on building at least 4 sensors, but more likely 8, as we have winter storage in 4 different location. We have a large collection over wintering. I am fine with a sampling technique in each area. Thinking a large pot and medium pot sample would get me close enough for my objectives. I want each one to be wireless, so I can remote monitor them. As a retired professional coder, software will not be an issue. I am new to esp32s. I have done a little bit of micro PLC work, so building the hardware should not be an issue. I love the work you have done. Very nicely executed !
Thanks. My experience has been that the capacitive sensors do not hold up to the elements very well over time. Here is the full playlist : ua-cam.com/play/PLqJ5k4cakypy1E4J_tvEiD0kpFK1tGY3D.html&si=r4vya3sAPACIluSV
Excellent! In the "maker" world we sometimes call a build successful if it even survives long enough to log some data. Suppose it does -- now, about that data?
I clicked on this video looking for 'pro tips" about a specific device and its application. But in teaching that, @Modest Maker is also demonstrating the thoughtful approach of a skilled integrator. Prototypers of all kinds on whatever platform, watch: this is how one coaxes a component to reveal capabilities beyond what its pricetag or (non-)documentation might suggest.
Yes please, OP @Modest Maker, I will enjoy getting your test data when you upload it!
Of course - updating description now with link to spreadsheet. I removed some confusing notes from preliminary tests that weren't relevant, but you'll find other calculations and references in the spreadsheet that might be useful: docs.google.com/spreadsheets/d/1Jl9fN1wNGhqj5SkljHGKHVYtNJAVK36e/edit?usp=share_link&ouid=100450587141122348038&rtpof=true&sd=true
@@ModestMaker Thanks for the spreadsheet, although, it's a link that requests access, so you may get a few emails asking for access to the document... Like from me! 😄
@@ModestMaker It's just lovely. thank you!
@@Aragorn450 Thanks Charlie - fixed!
@@RonnieTheSim Appreciate that - thank you!
Thanks for the detailed investigation!
I noticed you placed the sensor in the middle of the tube, while the holes are in the side, so they get dry faster than the central core.
Actually, when you took out the probe, it was still wet.. which could interfere with the reading value.
May improve the experiment with other configuration? eg: no lateral holes but allow evaporation from the top only?
Since the tube is small, it may be evaporating faster on the sides and it receive more sun
Thank You. Yes- I understand the potential for error but based on other experiments, the challenges are similar in field conditions independent of being in a column. I am not clear as to the reason- it perhaps may have something to do with soil compacting after wetting. I just find they are not reliable at least for my tested soil type. See 10:40 for field installation.
Great video! I've got a couple of these sensors myself (for irrigation monitoring for my home garden) and this is good to know but unfortunate. I'm wondering if there's an issue of adhesion and compressed soil.
1) Water sticks onto the coating of the sensors, so it may be that if there's any water present, it'll be on the
2) Clayey/silty soils are pretty compressible. Inserting the sensor into soil will create a dense region around the sensor which may reduce water permeability away from the sensor itself.
Possibly. This is why I qualified my conclusions for “silty clay soils”, although I have witnessed the same behavior in premium potting soil (no measurements taken though). Thanks for those observations and suggestions 👍.
Interesting experiment. Was the soil loose or compacted?
The soil was initially broken up but settled when I saturated the columns. The full playlist with subsequent experiments is here : ua-cam.com/play/PLqJ5k4cakypy1E4J_tvEiD0kpFK1tGY3D.html&si=CV6QYCToKY1i2yJY
how did you get the values and how you measure soil weight and dry soil volume? please
I weighed everything with a scale. There’s a link to a spreadsheet with calculations in the description of the video.
trying to use capacitive soil moisture sensors for irrigation of tomato plants on my balcony. Re soldered timer chips for TLC 555 (because NE 555 chips). using Arduino NANO 33 IoT with set // Set ADC to use 12 bits, analogReadResolution(12);, that is giving me resolution from DryValue1 = 3480; to about WetValue1 = 1910; (dry soil to almost dripping wet saturation). i probably need to look more at soil type (bulk density), field density, irrigation point and Willting Point.
overal great video, I have to learn a bit more. It is not as easy as to set "two point calibration" and hope for the best :( :D
I agree on the two point calibration challenge. My concern is the sensor response is non linear as evidenced by relative insensitivity to water loss. I don’t think increasing resolution will help since (at least for my soil type) the signal itself is insensitive, but please share if you come up with any interesting results or further mods. Thanks for sharing!
@@ModestMaker yop, i have resolution from about 1000 (fully saturated soil) to 3550 (air), using simple kalman filter to at least somehow smoother spikes, mapped and constrained values to percentage (0 to 100 percent). but i will need to do full calibration for each sensor (for "medium" that i am using) determining field capacity and wilting point for coco coir + perlite. oh why have i gone through this rabbit hole ? :(
I know- right? I am working on a radiation shield for temp but as soon as I think I have it figured out, something else creeps up 🙄. The reward will be to finally get it all right but it’s a long and winding road though.
@@ModestMaker yop, i have set up second monitor for one particular plant. Two sensors in one pot (15cm apart) protected from water (resoldered timinig chip, glue heat shrink on electronic part pcb and epoxy on sides of pcb not to soak moisture from sides) just to determine wilting point by observation (to look at plant with eyes, when it will start to suffer from not enough of water) roots are in fairly constant temperature (20-21 °C) plant have not been watered for about 4 days and moisture percentage is slowly but steadily raising on one sensor (dont know why) plant is still pretty much thriving ...
@@J0stikEyes are the best!
is it possible that your sensor in this setup have a faulty reading because of the bad sensor? I have come across this video: ua-cam.com/video/IGP38bz-K48/v-deo.htmlsi=Af19SZJ_XlXwAuC0 , Did your sensor use ne555, doesn't have voltage regulator or missing output resistor?
I checked all that here: ua-cam.com/video/lb4HdcpYWCQ/v-deo.htmlsi=_xzaAGwenIYx4eBV . All sensors went through thorough quality control before selection.
coding ?
Buying, Programming, and Testing Capacitance Soil Moisture Sensors
ua-cam.com/video/lb4HdcpYWCQ/v-deo.html
Exactly, all this hobby soil moisture sensors are crap. We tried a lot of them. The capacitive once are just so bad. Soil type is a factor, but changing the sensor a bit, a mouse or worm will give you different readings.
Yes- in addition, two of the three I have installed in my formal setup just started reporting weird data after 6 weeks installed in the field. Temperature shows some promise- I will post some new results shortly, suffice it to say I haven’t had much luck with capacitance.