-
Long Range RF modules
08/23/2015 at 18:58 • 0 commentsFor agricultural use, long range wireless connectivity is a hard requirement. The Vinduino-R remote sensor board can use the ESP8266 wifi module for short distance, as well as affordable APC modules made by Shenzhen-based company Appcon, for longer range.
You need 2 Appcon modules for a radio link. One at the host, and one at the remote node side.
It's a simple set up: serial data in = serial data out. Multi-node networks can be achieved by timed transmission slots, this is more energy efficient than a mesh network where all stations are continuously operating.Replacing the default antenna with a "real" antenna gives an instantaneous range increase. Even a quarter wave length wire doubles the range. The popular APC220 operates at 433 MHz, which is not allowed as license-free ISM (Industry, Scientific, Medical) operating frequency in the USA. In Europe and other countries it's OK tho use this frequency, though.
Alternative Appcon modules that use the allowed 915 MHz USA ISM band are APC320 and the more powerful APC340, which also uses secure spread spectrum. I have no information about FCC approval for these modules.
The APC220 can work with 5V supply voltage, APC320 and 340 require 3.3V. The Vinduino-R board has a header that is pin compatible with all APC modules, and has a LDO and power selection strap to set the correct voltage.
Below is an overview of available modules with their high level specifications.
-
Soil Percolation Test
08/17/2015 at 03:04 • 2 commentsLast weekend before the Hackaday Prize entry closes. And what a weekend it was!!
With these extreme temperatures, reaching 110F (43C), each grape vine consumes 4 gallons per day.
Irrigation is set to replenish the daily use of the vines in the coming week.I tested soil percolation speed, and general functions of the system, by applying a generous amount of water and log the time for the sensor resistance to respond.
Interestingly, it takes 5 hours for the irrigation water to reach the sensor at 2 feet deep, and it takes nearly 24 hours for the water to fully saturate the soil at 2 feet depth. Of course it will taken even longer for the water to percolate farther down to the sensors at 4' and 5'.Results may vary, depending on the type of soil, and installation depth of the sensors.
Because of this, you cannot use a single sensor in the active root system to act as an on/off controller for irrigation. By the time the water reaches the sensor, too much water would have been irrigated.
Will keep tracking the soil response to irrigation. This is important data for building an automated irrigation system.
Also made the required 4 minute video for the "Best product" contest category.
Good luck to all participating hackers! -
Prototypes, harvest, & hair curler sensors
08/11/2015 at 04:27 • 0 commentsIt's been a busy week. We are getting closer to grape harvest and preparations are in full swing. This weekend I managed to finish the 3 Vinduino handheld reader prototypes for the Hackaday Best product contest. Three colorful products are now on their way to the eagerly awaiting Hackaday judges.
Meanwhile in the vineyard, I still see low moisture readings from the higher (2' and 4') sensors. Irrigation during daylight shows a possible cause (see picture). I installed an additional dripper to irrigate directly in the sensor zone, removing another dripper to make sure my readings stay "honest".
With just a few weeks before harvest I am not experimenting with irrigation schedules anymore, no changes until the grapes are brought in and processed. Picture from this weekend, all is looking well. No sign of water stress, although we saved 33% water last month compared to last year.
Spanish Maker Paco Canton shared his design of a soil moisture sensor with me last year. Using hair curlers and 3-D printed components he made an interesting variant of the gypsum sensor.
http://unpuntilloalambre.blogspot.com.es/2014/01/gypsum-block-for-soil-moisture-sensor.htmlMatt Upson is on the same track: http://ivyleavedtoadflax.github.io//arduino/SylvaSense-2/
-
August 3 Update
08/04/2015 at 02:58 • 0 commentsLast year we used 441 HCF (hundred cubic feet = 748 gallon) irrigation water in July.
This year we used 295 HCF in the same period. That's an encouraging 33% less, and a reduction of more than 100,000 gallon.Yesterday, I installed the first Internet sensor node in the vineyard, and now can receive live soil moisture data from the vineyard!
Later than I wanted, unfortunately I had issues with unreliable WiFi connections. Exchanging the Arduino and ESP8266 WiFi module did not help. In the end I solved by using another Vinduino-R board that has been operating stable for several months.The Vinduino node connects to a WiFi hotspot, that connects to the Internet via the 4G mobile phone network.
The data is sent to the ThingSpeak.com web site, providing a nice graphical presentation of data trends.Live data at the following link: https://thingspeak.com/channels/37248
First data coming in already provides new insights. The deepest sensor shows a high water content. The sensors in the active root zone are significantly dryer. The amount of water is sufficient, just the vertical distribution is not optimal.
That is not how I envisioned it. Nevertheless very useful information.
Further "tuning" of the irrigation may be required, with potential for further water saving. Also the position of the irrigation drippers may be too far from the sensors to correctly monitor vertical transition of the water through the soil. Sufficient food for thought :-)To monitor the internal "health"of the system, there is a DHT-11 sensor that reports the temperature and relative humidity level inside the enclosure. Li ion batteries are not supposed to be charged at temperatures over 65C. There is no protection built in the charger, so I keep an eye on the temperature, just in case.
When the enclosure leaks during rain or vineyard spraying operations, I would be able to detect the humidity going up and take care of it before the electronics get irreversibly damaged.Battery voltage is a useful datapoint for a solar operated node. With the Arduino you can measure the supply voltage without any external components needed!! Voltage varies between 3.9 and 4.2 V, so the battery charger and solar cell seem to be working OK.
Below picture shows the installed Vinduino node, also showing the 3 connected gypsum soil moisture sensors.
-
YouTube Update
07/29/2015 at 15:51 • 0 commentsCello metal band Apocalyptica gave me special permission to use their music in my Hackaday 2 minute YouTube video.
Now how cool is that?Enjoy!!
Also finished the electronics for 3 "Hackaday Best Product" prototypes of the handheld reader.
And, yes, they all work :-) -
Installing Sensors
07/13/2015 at 00:15 • 0 commentsThis week, I installed extra gypsum sensors., and got a 4G WiFi hotspot for Internet access in the vineyard.
Connecting to the ESP8266 WiFi module works fine.
Tried DSL last year, did not work as the latency and slow speed was not usable., even for IoT.Pictures below show the assembly using 3/4" PVC pipe. The gypsum sensors fit tightly in a 3/4" pipe coupler.
Recommended sensor depth for grape vines is 24", 48" and 60". For other crops, different depth recommendations may apply.
I cut pipes to the required lengths, leaving 1 foot above ground, and marking ground level with tape.
Below shows the sensors installed at 2, 4, and 5 feet depth. Recommended is to soak them before installation, and have plenty of mud around the sensor to ensure good soil contact. Solder the wire ends to avoid oxidation.
In case you are interested in the vineyard status: The vines are doing well under the new irrigation regime.
It will take 2 weeks for moisture levels to settle, only after that period the sensors provide meaningful data.
Some of the grape bunches started to show a change of color, "Veraison" in grape grower terms. -
Handheld Reader Assembly
07/06/2015 at 02:25 • 0 commentsA friend grape grower uses my Vinduino soil moisture meter in his vineyard now, so I needed to make another .
The electronics assembly is the easy part, mechanical finishing of the enclosure is a job out of my comfort zone.The enclosure I chose for this project is the VM-series from Polycase. They also offer an optional protective rubber boot, that gives additional ruggedness if you have a need for that.
Back to the assembly. I used paint for a "confidence " stamp on the side where the cut out for the display window needs to be. I use a caliper to scratch lines for the accurate dimensions . For the window I use Lexan acrylic, which does not shatter and -according to the manufacturer - is 250 times stronger than glass. All properties that are highly desirable for farm use.
See below photo collage for assembly steps.
And here is the finished result!
-
Target: Saving 25%
07/04/2015 at 02:41 • 5 commentsLast year, under professional vineyard management, the annual water consumption of the vineyard was 2325 HCF.
HCF stands for Hundred Cubic Foot, equal to 748 gallons.
This years goal is to save 25%, that makes the target consumption 1750 HCF.After starting timer-based irrigation in Spring, it took some trial and error to get the system properly set up. Now it is dialed in, and the effect is clear from the June month reduction. From this data, 25% less water use looks a feasible goal.
However, Summer just started and we are looking forward to having quite some dry days with temperatures well over 100F.
The challenge will be to meet the target without impacting the crop yield.The 2014 curve (green, below graph) follows a typical shape, reduced irrigation is starting well before harvest.
The vineyard location does not have Internet. It's a rural area and choices are limited.
I tried a DSL connection, but latency was so bad that it was practically unusable.My project for this weekend is to try and set up an Android phone as WiFi hotspot.
Took me two drives to the phone store to get a suitable data-only plan.
I have it "kind of" working now. However, the hotspot tethering feature stops after 30 minutes or so.Suggestions welcome :-)
-
Eagle files added to Github
07/02/2015 at 00:40 • 0 commentsToday I added the Eagle files for Vinduino-H and Vinduino-R to Github. See Vinduino-R schematic below.
The circuit supports up to 3 soil sensors. there are 6 2N7000 FETs used to isolate the sensors from each other.
Without these FETs there is considerable measurement error because of soil resistance between the sensors.The Arduino Pro Mini is most of the time in deep sleep state to conserve power. The RTC alarm fires an interrupt that wakes up the system, takes measurements and sends the data, then goes back to sleep again. This allows tiny solar panels to keep the system alive.
To further conserve power, pin 13 enables a load switch that controls the WiFi module via a 3.3V regulator.For connectivity, use ether APCxxx or ESP8266, both together does not work. Remove the programming cable (FTDI) for operation.
Errata:
R9 shown as 3k3, needs to be 10k.
ESP8266 pin 7 and 5 must be wired to 3.3V supply
RTC module supply strap need stop be connected to either back battery or Vbat -
First project progress update
06/28/2015 at 19:19 • 0 commentsComparing two DIY gypsum sensor with a commercial Watermark SS200 sensor. With the Vinduino-R board, connected to the Thingspeak web site, I collected measurement data -every 4 minutes- for 3 weeks. I used soil taken from my vineyard (decomposed granite).
For the Watermark sensor, there is a lot of calibration data available to convert measured resistance to soil water potential.
Below is the comparison result. The Watermark shows more sensitivity in wet soil range. Both sensors show a temperature dependency. For watermark sensors, temperature compensation equation are well documented.
In summary: both sensors types look very usable for irrigation control.In the vineyard we saw the first grape berries changing color, "Veraison" in grape grower terms.
The vineyard is looking well, even under the current reduced irrigation conditions.