Moisture level
computer guy
Posts: 1,113
Hi everyone,
Does anyone know how I can measure soil moisture level with the propeller?
Thank you
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Check out my robot using the propeller RECONAUTOR
If you like my avatar then check this out Propeller Domed Sticker
Does anyone know how I can measure soil moisture level with the propeller?
Thank you
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Check out my robot using the propeller RECONAUTOR
If you like my avatar then check this out Propeller Domed Sticker
Comments
There may be simpler ways, like measuring the resistance of a couple of wires stuck into a buried sponge perhaps?
Regards, David
Thank you will try to build an SMX module as described.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Check out my robot using the propeller RECONAUTOR
If you like my avatar then check this out Propeller Domed Sticker
Turns out that the moisture sensor "Watermark sensor" is $100.00
Any other ideas?
Thank you
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Check out my robot using the propeller RECONAUTOR
If you like my avatar then check this out Propeller Domed Sticker
Let me know if you find a link and price to these gypsum blocks mentioned on the emesys.com site.
David
I found a site that has them however there is no method of purchasing one.
www.environdata.com.au/Product/Sensors/Gypsum_Block_Soil_Moisture_Sensor.html
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Check out my robot using the propeller RECONAUTOR
If you like my avatar then check this out Propeller Domed Sticker
Note: an extra 1m in cable length costs an extra $19
www.tain.com.au/sensors_and_accessories.html
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Check out my robot using the propeller RECONAUTOR
If you like my avatar then check this out Propeller Domed Sticker
This site: www.mea.com.au has lots of soil measurement stuff, it even has a page (Link) with a gypsum block FAQ.
Nice find on the $16 gypsum block. Now, how to use them...
David
Had lots of useful information.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Check out my robot using the propeller RECONAUTOR
If you like my avatar then check this out Propeller Domed Sticker
I've been kicking around the idea of putting an Earth Box kind of thing (see www.earthbox.com) on my roof and automating it. Maybe I can grow some tomatoes that don't taste like sawdust.
Home grown tomatoes are always nicer.
Mmmmm Home grown.
Look forward to their response.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Check out my robot using the propeller RECONAUTOR
If you like my avatar then check this out Propeller Domed Sticker
I'm not so sure however if there is something in the dry-wall material that you wouldn't want leeching into the soil, but these days materials used in building interiors has to be pretty safe, at least in the developed World.
No reply today from Tain, after all it is Saturday here and Victoria is hours later than Jakarta time.
David
Mold a sensor out of Plaster (or Plaster of Paris, same-thing), which becomes gypsum when it hardens? According to Wikipedia (I know, I know):
When the dry plaster powder is mixed with water, it re-forms into gypsum. Plaster is used as a building material similar to mortar or cement. Like those materials plaster starts as a dry powder that is mixed with water to form a paste which liberates heat and then hardens.
So Voila, instant gypsum-block sensor material.
Hmmm....
Edit, Oh yes, it also mentions on the gypsum Wikipedia page that gypsum dissolves over time with water. So one wonders how long these sensors may last.
Hmmm.... again.
Now back to the Topic...
Found the Watermark granular gypsum sensors for $32 USD each in unit quantity along with sensor-driver boards in board-only ($4.25), kit ($19.50), or pre-built ($27.50 with conformal coating $? option). Maybe the $100 price for the Watermark you found Computer Guy includes a sensor driver? Link at:
www.hobby-boards.com/catalog/product_info.php?products_id=1547.
It seems the Watermark granular gypsum sensors have internal compensation for commonly found salinity levels, and are not affected by freezing. Plus they have some form of composite composition that delays water dissolution. Something a home-made gypsum block may not be able to handle as easily. But a home-made sensor out of Plaster may be fun to do, and refine anyway.
From the MEA FAQ, it seems these Watermark sensors will dissolve eventually, 1-5 years is cited, perhaps 10 years. Not clear.
I'm not comfortable with this, so I did a quick search for a dielectric sensor solution, hopefully using just an etched PCB. Found this...
Here's are links to a Boston Univerisity wireless soil monitoring system that uses a PCB capacitive sensor (maybe the best way to go?) Picture and schematics are included for the project in the paper along with sensor calibration examples. I would conformally coat the PCB sensor to prevent any soil-metal galvanic/chemical interaction. A simpler and perhaps more robust approach would to build the dielectric sensor PCB (or something similar, not necessarily a PCB) then dip it and the attaching wires into a latex compound, like that you can buy at hardware stores to coat the handles of screwdrivers, etc. But I'm not sure of the dielectric properties of the latex coating especially over a broad temperature range, experimentation needed. But if the latex dielectric effect isn't a big factor, then you might be able to ignore any temperature effects for indoor or greenhouse applications:
Base link (scroll down to "Soil moisture monitoring of individual plants."):
http://hulk.bu.edu/projects/ecomon/summary.html
Link to acutal 4.5MByte document in .doc format:
http://hulk.bu.edu/pubs/papers/2007/TR-05-02-2007.doc
I can envision something where the Propeller can not only generate the AC excitation of the dielectric sensor, but for a small installation do it for many sensors. Does Propeller bring something special to this application, well yes, nice timers, 32 bits, and certainly a fast high-level development environment; and no, lack of internal A/D+mux etc. Fast-easy prototypiing is key though for me. If Propeller didn't exist, I'd certainly prototype this in a BASICStamp first, but the Prop does exist and is the future IMHO. Certainly Propeller will shine if you need to process many-many sensors in a "production" end-product. Hmm.... scalability.
This is fun. I've always wanted to educate myself on this soil moisture sensing topic, this thread has been a catalyst. Thanks Computer Guy for giving me a kick in the pants!
A last comment: It occured to me during this post that we're talking about a subject with a lot of "key words" that may be scraped and parsed by certain authorities searching for those growing "illicit" stuff (damn, more key words). Let it be known: I'm in Jakarta at the moment, they Hang you for said activities here, this is Not may aim. I was just thinking about trying to get some tasty tomatoes from a roof-top EarthBox - Really.
Remember, on the 'Net, IMO privacy is dead.
Regards,
David
Nice find on the $32 Watermark sensor.
The one I found was exactly the same thing except for $100.
The design of the sensor done by Boston University is nice, don't know how long the contacts will last though.
Glad to help.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Check out my robot using the propeller RECONAUTOR
If you like my avatar then check this out Propeller Domed Sticker
If you want a complete introduction to soil moisture measurement, be sure to check out the SOWACS site. It is a clearinghouse of information, and there is a mail server (rather quiet lately), but the mailing list archives are there for search. There are many ways to do measure soil moisture and, many tradeoffs of price and performance. Most of them are far more expensive than the Watermark. The $32 list price is about right for the Watermark sensor.
A gypsum block can certainly be a DIY project. Embed a couple of SS screws in a block of plaster of paris, using a polyethylene film can as a mold. The Watermark includes a pellet of gypsum up at one end near its stainless steel electrodes. The bulk of the sensor is filled with an extremely fine sand mix/aggregate, and that is contained inside a permeable membrane and stainless steel perforated shell. The purpose of the gypsum is to act chemically as a buffer against the pH and ionic content of the soil. So the reading depends on moisture alone, not so much on those other factors as it would if you simply put the electrodes in the native soil.
The Propeller would be a good bet for a capacitance type probe. The best ones operate at high frequencies, 40 mHz and above, so that they can probe the complex imdedance of the soil and in theory separate soil conductivity (salinity) from volumetric water content. For example, the ECH2O from Decagon Devices.
One difficulty in soil moisture is that the soil tends to pull away from the sensor, or a coating forms on the sensor, and the resulting gap spoils the readings.
Sometimes the goal of measuring soil moisture is not the volumetric water content, but the osmotic pressure, which is the work that a plant has to do to extract the water. Clay soils hang onto water much more tightly than sandy soils, so the plants have to do more work to extract it. A tensiometer is a device that measures osmotic pressure directly. Most of the devices that use capacitance and resistance do so very indirectly and have to be calibrated for each soil type. But once that is done they are cheaper and more rugged.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com
In the winter you have to buy Campari... I get mine at Walmart.
Et al...
I am absolutely not an expert on this subject, but I don't see why you couldn't:
1) wrap a Sensirion sensor in a semi-permeable membrane 2) put the unit into a probe with small holes in the probe wall to let the moisture in and 3) put the probe inside of a metal sheath, which you could retract to expose the holes in the probe when you get to the right depth.
The measurements would only correlate with the absolute water content... but that is probably what some of the other alternatives do too... with the Sensirion you could temperature correct your numbers ... which some of the other methods probably can't[noparse]:)[/noparse]
To calibrate your measurements you could microwave a soil sample to dehydrate it, weigh it and then slowly add water. As Tracy has said, and as you probably already know... soil composition adds some variables for you to consider but they could be handled in a similar fashion. What you are interested in is the amount of water per cubic yard, which is actually available to your plants. A clay mixture and a silty soil might give you the same "soil humidity" reading when there was much less water actually available to the crops in the clay mix than in the silty soil.
If you dealt with enough different kinds of soil, you'd eventually end up with enough data to create a nomogram relating your readings to soil composition.
$30 in parts and the object is already in the exchange[noparse]:)[/noparse]
Rich
In the BU paper and more-so in my post on the paper, the key is to coat the sensors with an inert material, conformally coat the PCB, or as I suggest coat the board in latex. That way the contacts on the sensors have indefinite life. They never come in contact with the soil or water. The sensor is capacitive dielectric.
Hi Tracy...
You raise some good points. I think the dielectric sensor in the BU paper measures relative mosture content, not absolute, which I think (correct me if I'm wrong), an absolute measurement will coincide with the osmotic pressure. Hmmm...
A pressure tensiometer tube measures osmotic pressure. Does the gypsum-based Watermark sensor measure osmotic pressure too?
I think a relative measurement may be ok for simple applications, but if you disturb the soil around the sensor, the sensor's baseline may need recalibration.
I've bounced-off the Decagon Devices site and tried to find pricing on the Echo type sensors. Nope... What do these things cost? Do they deterorate over time like gypsum blocks?
I agree, the SOWACS site is the Web pivot-point for this subject. I found a lot of the links I've cited in this thread through SOWACS, should've cited it in the beginning.
Hi Rich...
Which Sensiron sensor are you referring to? You said, "...$30 in parts and the object is already in the exchange". What object are you referring to. Is the object computational a look-up table or other data translation utility. Traditionally a Nomogram is a graphical data translation thing, like a Smith chart or log-log paper.
Hmmm...
By now it's obvious this soil moisture sensing thing is a pretty complex topic. I'm tempted to just stick a couple of stainless steel spoons in the soil with the ends sticking out, drill a hole in the spoon end-and bolt a connector to it, then black Epoxy the connection (black to prevent UV deterioration) with the spoons and measure either resistance or capacitance with a wide sample "window", e.g., needs water or not.
Of course, there's another way to avoid this whole soil-moisture sensing issue: Grow in a closed hydroponic system. No soil required! Not to say sensing other things isn't necessary, but the soil moisture content sensing goes away.
Rgds, David
Sorry about that.
http://www.parallax.com/Store/Sensors/TemperatureHumidity/tabid/174/CategoryID/49/List/0/Level/a/ProductID/94/Default.aspx?SortField=ProductName,ProductName
I like the way your brain works[noparse]:)[/noparse] but I think computer guy is looking at commercial farming appications (I hope[noparse]:)[/noparse].
The Prop is actually fast enough to support NMR measurements... in theory you could tune yourself to hydrogen and then map out the water content at various depths (and identify the various states of water) on the fly. In application, it would look very much like sonar data or ground penetrating radar. Not real difficult to do... either the Swiss or the Germans will probably do it first[noparse]:)[/noparse]
Thanks for the clarification.
After studying this subject in-detail, it seems not a lot has changed in decades. I once consulted for the design of an orchard sensor system to gather environmental data to spot Apple-Scab likely locations that needed spot-specific prevention rather than (costly) blanket spraying. Apple orchards are often on hilly terrain, so there are "micro-climates". I don't understand the biology/ecology of Apple-Scab, others were dealing with this. I dealt with the sensors, interfacing, micro-controllers and wireless interconnection (mesh, modified AX.25). Got pretty far with it, can't say much more. This was 30 years ago.
Certainly this soil moisture sensor technology is stagnant IMHO. I think we'll see this addressed at the MicroElectroMechanical Systems (MEMS) sensor level soon and eventually perhaps at the nano-technology level. But the MEMS technology development IMHO is staggering over frivolous Patents and attendant law suits, thanks mostly to the corrupt and/or inept US Patent and Trade Office (USPTO), fed by the Trial Lawyers of America (TLE). (OOps, more trigger key-words, "they're" really going to be watching this thread now).
One reason for the stagnant development (again in my opinion) of cost effective agriculture automation is the huge Socialist-based farm subsidies that exist in the EU and USA. More and more huge corporate farms are reaping the benefits of these subsidies, which are really not based on true growing conditions on a season by season basis. Yes I agree farmers need government subsidies as insurance against bad seasons, but the current models are blanket pay-outs, vote-buying and the result of lobbying if you will. Let's not even get into how these subsidies suppress third-World farmers.
So, we've come full-circle in this thread. How to measure soil moisture, various affordable methods (didn't delve into Neutron monitoring etc.), and the state of the technology today. I finish with my rant about missing MEMS-like technology and how subsidies thwar technology development.
I think I'll let this thread go in terms of politics. The Forum is about Propeller after-all, not about World/govt policies and how they affect technology development, I'm getting off-topic. But there are some things that need to be said.
Regards, David
You can let it go... but I have one more thing to say. Whenever governments impose regulations, which lead to a loss of heterogeneity in the market place, they place our entire economies at risk. This is not just foolish, it is insane... and it isn't just agriculture it is everywhere.
Sorry for the OT here, but I think we are on "friendly ground (with unknown water content)."
Rich
I haven't read the Boston U. paper yet, but I've downloaded it from your link. I think that capacitance sensors generally measure something closer to volumetric water content, rather than by weight or by pressure. They respond to the dielectric constant of the water that is in closest proximity to the probe, and insofar as that water has full degrees of freedom for polarization, it will still have that dielectric constant of around 80 that swamps all other materials. The measurement is relative, not absolute, because it has to be calibrated with a constant of proportionality for each soil type. I don't know enough to say how much the dielectric constant is affected by solutes and by binding by capillary forces in the soil matrx.
The gypsum block (or Watermark) measure something closer to suction pressure needed to extract water from the soil. I was fast and loose with the terminology. The main pressure component is usually called "matric" or "matrix" potential and is due to the capillary forces in all the tiny intersices of the soil matrix. Clay is composed of thin layers that swell with water and hold it tightly due to capillary and chemical (van der Waals?) forces, but sand can only hold water between the grains. Osmotic pressure refers to the solutes, i.e., if you have pure water separated by a water permeable membrane from a salt solution, the osmotic pressure refers to the pressure that would balance the diffusion of water through membrane. The third pressure component is gravity. A tensiometer works better at toward the wet end, and can even be used to measure the depth of a water table in saturated soil. But tensiometers do not work well in dryer soil, dryer than -70 kPa. Gypsum blocks and Watermark aggregate sensors do not work well in wet soils, but are okay in the range of 10 to 200 kPa.
Making the measurement at higher frequencies makes it possible to extract more information. That is done in TDR (time domain reflectometry) and other techniques that measure both attenuation and propagation velocity to get at both conductivity and moisture content. The Propeller could be a real good candidate for that approach.
That's interesting that you worked on an apple scab project. That is where I got my start with this field too, working with people in the Entomology Dept. We got a grant from the "appropriate technology" office, and that developed into a data logger based on Trash 80 model 100s that sat in Applie Orchards and phoned in data to the university mainframe every night, at a whopping 300 baud. The Tandy ran a Coddling Moth degree-day model that had been developed at Cornell.
Soil moisture measurement is squishy science, and the research mostly takes place in University and Ag Extension offices, and in efforts by individual companies to validate (and promote) their own products. They all work but all come with caveats! As you pointed out, there are soooo many microclimates . Moisture measured at a couple of depths at a few locations will not necessarily be representative of an entire area. There is also remote sensing via infrared, which can spot surface moisture over wide areas from the air. I won't go into my take on the politics. But I do think there are openings for new approaches.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com
I really love Ag people... My first science project was trying to make guinea pigs scorbutic... turns out you can't. Then there was an MRI project to look at rhinitis in hog snouts... on pigs grown in a negative ion environment... by a world class physical chemist. You should have seen the look on my technician's face when I told her that we were going to do MRI on pigs[noparse]:)[/noparse]
I really like this problem... hope this thread stays around for a while.
rjo_, Guinea pigs with scurvy - now that's off topic (LOL).
David
http://www.circuitcellar.com/fi2003/F2002.htm
solar powered with a rf transceiver too!
Post Edited (Fred Hawkins) : 5/26/2008 12:00:13 AM GMT