Poll

What major features would you want in an OpenAg Sensor Electronics Application?

Cheap/Accurate Soil Sensor Network
1 (14.3%)
Light / Lux / Temperature / Humidity Data
2 (28.6%)
Irrigation Control
2 (28.6%)
SmartPhone Control / Website Data Graphs
1 (14.3%)
Other Things
1 (14.3%)

Total Members Voted: 4

Author Topic: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech  (Read 1946 times)

Andrew Barney

  • Full Member
  • *****
  • Posts: 528
  • Karma: 43
  • Northern Colorado, Semi-Arid Climate, USA
    • Pea Breeding, Watermelon x Citron-melon, Purple Foliage Corn, Wild Tomatoes
    • View Profile
    • My blog
    • Email
  • Koppen zone: Dfc / Dfb
  • Hardiness Zone: 5b
So, I'm running an early poll about something I'm planning on working on in the near future. I found a really interesting professor on campus who is running an unusual class and has an unusual lab. What caught my eye was that they had a poster and he has two graduate students working on what they are calling as of now an OpenAgSensor electronics device. At it's heart is an arduino-like Adafruit Feather chip that is capable of being a cheap and open source Internet of Things device (IOT). Basically that means that it can connect through bluetooth, wifi, or mesh radio to create something very useful to Farmers and Gardeners professional and amateur alike. Since it's aim is to be Open Source it should be very hackable, easy to set up, and relatively cheap to implement compared to standard commercial devices currently on the market.

So my precocious "jumping-the-gun" poll aims at getting an early look into if such a device would interest you, what kind of things would you really want it to be capable of, and what features would be killer or make-it-or-break-it items.

I can conceivably see this as being useful to 3 kinds of people, though there could be more. 1. The average home gardener, 2. Hydroponics / Greenhouse control and monitoring, and 3. The average large-scale farmer.

Some ideas i came up with are these:

  • Cheap / "Precise Enough" Soil Moisture Sensor Network
  • Soil Temp sensors
  • Light / Lux Sensors
  • Temperature Data
  • Sprinkler / Irrigation Control
  • SmartPhone Control (probably bluetooth or wifi)
  • Web control and/or graph data monitoring
  • Humidity Sensors
  • Buzzer for sound alerts
  • Leds for light alerts
  • Camera?
  • Light Control
  • Vent Control
  • General Stats such as overall water use, etc.
  • LCD Screen for monitoring / control without a computer or cell phone.

In addition i would like to know what kind of connections you would need in your circumstances. Is WiFi readily available? Would Bluetooth to Smartphone be the preferred method? Remote farm conditions that would either need long-range mesh radio or Cell Phone Capable units in field? Battery powered and possibly solar charging vs electricity available.

If something like soil sensors are of interest how many would you want to use? 1-5? Over 100?

Ideally i'm thinking i will either get access to the prototypes already made or/and i will design and create my own based on input and improvements i think i can see should be done. I should be capable of having some PCBs made and assembling some early prototypes. Anyone interested in Alpha / Beta / Protype testing would probably be welcome to a unit to test and provide feedback. Those with high Technology skills might be most helpful in the early stages, but even moderate tech skills might be fine. Obviously the long-term goal is to hopefully make it accessible and easy to use for those for anyone and everyone.

Thanks
-Andrew

spacecase0

  • Full Member
  • *****
  • Posts: 67
  • Karma: 4
    • View Profile
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #1 on: 2019-12-18, 11:48:40 AM »
I have been looking for an irrigation solution for a while now,
trying to set up trees on 13 acres of land. elevation changes are over 200 foot, with one plot of land being not physically connected to the others.
wireless internet to cover the land needs many routers. many in places where there is no electricity.

the water I have clogs almost any filter and the irrigation emitters/valves.
so I end up hand watering everything (open end of PVC pipe and ball valves don't clog for many years.
due to elevation changes and water supply, some places I can only water 5 trees on one valve

so I have been looking for some automated system that can run on small solar
something that is somewhat affordable.
best I can do now seems to be about $240 a tree to install for a reliable system, and that is never going to happen.
the small solar timers might work (I have a neighbor that has some), but they do mess up, and the loss rate there would not be acceptable.

I thought about building my own solution, it is not like the valves and other parts are that pricy, but I am not really any good at programming.

also, none of the wireless hardware I have seen uses each component as a repeater, so that as long as they are close enough, you can keep expanding the hardware as far away as you would like.

I don't know how common my issues are, but I did run across one farm company that is moving into the hills for the lower priced land, and they were having the same issue I was having.

Andrew Barney

  • Full Member
  • *****
  • Posts: 528
  • Karma: 43
  • Northern Colorado, Semi-Arid Climate, USA
    • Pea Breeding, Watermelon x Citron-melon, Purple Foliage Corn, Wild Tomatoes
    • View Profile
    • My blog
    • Email
  • Koppen zone: Dfc / Dfb
  • Hardiness Zone: 5b
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #2 on: 2019-12-18, 12:38:58 PM »
I have been looking for an irrigation solution for a while now,
trying to set up trees on 13 acres of land. elevation changes are over 200 foot, with one plot of land being not physically connected to the others.
wireless internet to cover the land needs many routers. many in places where there is no electricity.


so I have been looking for some automated system that can run on small solar
something that is somewhat affordable.
best I can do now seems to be about $240 a tree to install for a reliable system, and that is never going to happen.
the small solar timers might work (I have a neighbor that has some), but they do mess up, and the loss rate there would not be acceptable.

also, none of the wireless hardware I have seen uses each component as a repeater, so that as long as they are close enough, you can keep expanding the hardware as far away as you would like.



Awesome! This is great feedback!!

Im not the best at programming myself,  but I know that the Arduino and CircuitPython capable boards like the Adafruit Feathers are really opening up things to the DIY and hobby community. Something with the Adafruit Feather being the brain of this whole board.

While im still learning about all this technology, and I expect to learn more this spring,  I know that there are feather or feather- compatible boards that either have long range radio or mesh capabilities. I don't know if you can get both, but I bet there either is or perhaps the long range radio would be sufficient that mesh communication would be unnecessary.

Here are two products that might fit the bill, but I know there are lots of other options. I think it would be awesome if this ended up being the best solution for you. I personally have a lot of interest in working on this project overall.

https://www.adafruit.com/product/3078

Quote
Our initial tests with default library settings: over 1.2mi/2Km line-of-sight with wire quarter-wave antennas. (With setting tweaking and directional antennas, 20Km is possible).

https://www.adafruit.com/product/3997
« Last Edit: 2019-12-18, 12:41:00 PM by Andrew Barney »

spacecase0

  • Full Member
  • *****
  • Posts: 67
  • Karma: 4
    • View Profile
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #3 on: 2019-12-18, 08:18:44 PM »
I use radio where I live every day,
often use a radio that is 0.1W out on the 430MHz band (I am licensed for it),
with a large external antenna it talks to every point of land I have (and beyond)
the hand held radios I use have 1W out (or more), the antennas they have are not that great...
at 0.1W and 1/4 wave default antennas, they would have no hope of talking to each other around here.

so it looks like I can get what you linked to to work (as they do say 0.1W max),
to bad they are not higher power... would make things easier
the battery requirements and power usage are perfect

they even sell latching water valves that run on 12V
https://www.adafruit.com/product/997
https://www.adafruit.com/product/996

if it matters, I worked doing electrical engineering for a few years,
prefer analog and CMOS chips, I was away from that field when things like the Arduino came out
I would have no issues designing things like the solar power supply system and interfaces
looks like the hardware is easy now
suppose the software can be figured out with enough effort
I am all in with this project.

Andrew Barney

  • Full Member
  • *****
  • Posts: 528
  • Karma: 43
  • Northern Colorado, Semi-Arid Climate, USA
    • Pea Breeding, Watermelon x Citron-melon, Purple Foliage Corn, Wild Tomatoes
    • View Profile
    • My blog
    • Email
  • Koppen zone: Dfc / Dfb
  • Hardiness Zone: 5b
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #4 on: 2019-12-19, 10:32:45 AM »
Well, I'm still learning about a lot about the technical details myself, and I'm far from an electrical engineer, so you almost certainly know a lot more than I do in a lot of these areas.

What I do know is that some of the Arduino compatible radios have connectors for external antennas. There are even some modules capable of communicating with cell phone towers. I don't know exactly how much power all of the external ones can transmit or how far, as I think they are all different and directional antenna design may also affect that greatly. But yeah, as you know the lower the frequency generally the better the range. It may also be possible to implement your current radios into such a circuit, but I don't know. Most of these microcontrollers are capable of serial communication, I2C (very popular now), UART, GPIO, ADCs, and DACs.

There seem to be modules of these that do not have built in radios, so an external radio would be needed anyway.

Here is one that looks nice without a radio, includes the newer USB-C, and a quick connector for those I2C modules for plug and play compatibility. https://learn.adafruit.com/adafruit-stm32f405-feather-express

Here are two with external antennas designs, though they may or may not be what you would need. Just an example of more designs.


https://www.adafruit.com/product/4075

https://www.sparkfun.com/products/15435

And more stuff...

https://www.sparkfun.com/products/15131

https://www.sparkfun.com/products/15069

https://www.sparkfun.com/products/12823

https://www.adafruit.com/product/3998

The xbee modules with the external antennas might work as they say they can transmit up to 2mi, and I think they can be configured for mesh networking.

Andrew Barney

  • Full Member
  • *****
  • Posts: 528
  • Karma: 43
  • Northern Colorado, Semi-Arid Climate, USA
    • Pea Breeding, Watermelon x Citron-melon, Purple Foliage Corn, Wild Tomatoes
    • View Profile
    • My blog
    • Email
  • Koppen zone: Dfc / Dfb
  • Hardiness Zone: 5b
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #5 on: 2019-12-19, 10:42:23 AM »
As far as programming goes, it sounds like this new micropython stuff might be easier to program than Arduino code, which is basically c, but I currently know nothing about it.

As far as Arduino code goes, even me with terrible coding skills can usually hack something together due to the friendly nature of the Arduino IDE environment and tons of sample code and pre written libraries. These days if there is a common piece of hardware someone has already written Arduino code for it and that makes it easy to copy and paste it into your own code or copy it completely.

These newer microcontrollers already run at 3.3vDC but additionally they might also have low power modes to save even more battery power.

spacecase0

  • Full Member
  • *****
  • Posts: 67
  • Karma: 4
    • View Profile
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #6 on: 2019-12-19, 08:57:49 PM »
a few thoughts on ideas and links,
OLED screens fail with time and heat. that limitation should be kept in mind for any designs
external antennas are easy to design, I will help all you want with that part.
and you don't need a connector on the board to use one, with a knife and soldering iron you can modify almost anything.
things like 3.3V in is not an issue even with a 12V solar power system, there are plenty of easy to get low priced DC to DC converters that don't loose much power in the conversion.
for the price I am seeing, radios built in look like a good deal and the way I would want to go, but if you are running water, running cat5 cable is not really that much harder. the power requirement is low enough they could also carry power. hard copper wires do have advantages...

good to know that the code should not be to hard.
I learned a bit of c, it is complex, but at least it made sense to me even though it seemed to hurt me doing to much of it.

something else that occurred to me about this idea is that it should be way better than standard systems I have seen on the market,
you can have something auto watered, and if something fails, it can tell you that you are out of water or something like that.
people do loose plants over automated systems failing, will be fantastic to have one that can check the soil moisture and tell you if something is not working.
a gopher detector would also be handy.

they say that over half of software errors are in the design phase before any code is written.
for a few years I had a job where the only errors I had to deal with were ones I had made, then I changed jobs and only dealt with the errors of others...
will be nice to get to logic the entire thing out before the code starts so that there are no fundamental errors.

is there some website or forum devoted to this project ?
is this thread this project's home ?
if not, should we start one ?

edit:
we also need ultrasonic detectors so that the plants can tell us when they need watered.
« Last Edit: 2019-12-19, 08:59:44 PM by spacecase0 »

Andrew Barney

  • Full Member
  • *****
  • Posts: 528
  • Karma: 43
  • Northern Colorado, Semi-Arid Climate, USA
    • Pea Breeding, Watermelon x Citron-melon, Purple Foliage Corn, Wild Tomatoes
    • View Profile
    • My blog
    • Email
  • Koppen zone: Dfc / Dfb
  • Hardiness Zone: 5b
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #7 on: 2019-12-20, 10:12:32 AM »

is there some website or forum devoted to this project ?
is this thread this project's home ?
if not, should we start one ?

edit:
we also need ultrasonic detectors so that the plants can tell us when they need watered.

As of now I've sort of jumped ahead, so while there is a website for professor Ham's lab and they are planning to put more info on there (even hiring another student to update the website stuff from what I was told), as of now there is not much on there. Hopefully more to come soon. I know they are planning a GitHub account at some point for the code and electronics EagleCAD files at some point. But I see no reason this thread could not be a temporary place for some development design.

https://micromet.agsci.colostate.edu/

https://micromet.agsci.colostate.edu/welcom-to-our-lab/learn/

Ideally I'd like to start with the EagleCAD files they have and modify as seem useful, but I might be able to come up with my own board on my own. I'm not sure what code they have written already, but I know they mostly have worked on irrigation stuff and they showed me briefly a webpage with some graphs and stuff collecting sensor data.

I saw in some of their early prototype stuff they were using a board like this one.
https://www.adafruit.com/product/2890
https://www.adafruit.com/product/3417

I really like the idea of the 3-footprint-splitter because I think in many cases you would want two or three modules. The first would be the brain feather itself in whatever format with radio you would want, there second could be a lcd screen or a relay board, and third might be a GPS or something else. The great thing about the feather boards (modeled after the Arduino) is modularity. Not only can many Shields (being called featherwings) be stacked horizontally in parallel, they can  be stacked vertically and soldered that way. Due to one or multiple buses like I2C or another is a brilliant way they have come up with. So that is awesome.

I think I just ordered one of these boards for myself as I like the screw terminals for quick disconnect.
https://www.adafruit.com/product/2926

But here are some other examples of Shields like GPS and Relays:
https://www.adafruit.com/product/3133
https://www.adafruit.com/product/3191
https://www.adafruit.com/product/2923

Ethernet. POE is an option it looks like.
https://www.adafruit.com/product/4003
https://www.adafruit.com/product/3201

I'm going to experiment with these sensors. It sounds like the capacitive sensors are much more resistant to corrosion over the cheap resistive sensors (even with gold plating). I think something similar is what the lab at CSU is experimenting with, though they also have access to expensive industrial sensors as well. But from what I was told professor Ham wants to find affordable but "precise enough" solution. It sounds like they might also be capable of being epoxy or rubber coated yet still sense enough capacitance change to work well but increasing the corrosion resistance.
https://www.seeedstudio.com/Grove-Capacitive-Moisture-Sensor-Corrosion-Resistant.html

And I might need something like this.
https://www.adafruit.com/product/4309
https://www.adafruit.com/product/4357
https://www.seeedstudio.com/Grove-Shield-for-Wio-Lite-p-4156.html
https://www.sparkfun.com/products/15109

But these sensors seem somewhat interesting.
https://www.vegetronix.com/Products/VH400/

What do you mean by ultrasonic sensors for watering? I'm not familiar with those.
« Last Edit: 2019-12-20, 10:18:23 AM by Andrew Barney »

spacecase0

  • Full Member
  • *****
  • Posts: 67
  • Karma: 4
    • View Profile
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #8 on: 2019-12-20, 05:31:20 PM »
plants emit ultrasonic sound when they need watered
https://www.biorxiv.org/content/10.1101/507590v3
that analysis is fairly new, but the use of this idea has been out there for many years.
greenhouses that have converted to letting the plants water themselves have almost no loss while using less total water.
what I had not previously known is that each plant seems to have it's own frequency,
so you should be able to water the plants you want and not so much the weeds.
« Last Edit: 2019-12-20, 05:35:18 PM by spacecase0 »

Ellendra

  • Full Member
  • *****
  • Posts: 33
  • Karma: 5
    • View Profile
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #9 on: 2019-12-21, 11:16:03 AM »

what I had not previously known is that each plant seems to have it's own frequency,
so you should be able to water the plants you want and not so much the weeds.


What I really need is an automated weed zapper. This is giving me an idea on how to better identify them. Thanks!!!
Harsh winters, high winds. Temps on the edge between zones 4 and 5. Steep, north-facing slope. Soil is high in clay and rocks. Fast draining, which is a surprise for clay soil. Indicates a sandy/gravelly layer underneath.

Andrew Barney

  • Full Member
  • *****
  • Posts: 528
  • Karma: 43
  • Northern Colorado, Semi-Arid Climate, USA
    • Pea Breeding, Watermelon x Citron-melon, Purple Foliage Corn, Wild Tomatoes
    • View Profile
    • My blog
    • Email
  • Koppen zone: Dfc / Dfb
  • Hardiness Zone: 5b
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #10 on: 2020-01-01, 08:41:44 PM »
Im not exactly sure how the ultrasonic weed sensing would really work,  but if there is a sensor out there already im sure it could be integrated fine.

Dr. Ham is working on something interesting in his lab that is basically a stem tap that uses electrical signals from the plants directly to detect when they need water. His grad students showed it to me briefly and it sounded cool,  but I don't currently know more about it than that. I bet there is a paper about something like it out there already.

So what voltage source would be coming in to this sensor board? A solar panel with a 12v battery?

How many devices would you need to control in each spot? Just one relay controlling one water hose? Multiple relays?

spacecase0

  • Full Member
  • *****
  • Posts: 67
  • Karma: 4
    • View Profile
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #11 on: 2020-01-02, 02:28:18 PM »
I would assume that most people would set up a 12V solar battery system, the latching valves run on 12V, so not many other easy options.
6V might be more efficient... but if you get many days without sun, you might want the additional discharge ability that you have if dropping down to 5V.. and then up converting for the 12V would be less efficient.
and a DC to DC buck converter does not loose much
the LM2596 converter boards are not much cost and will down convert to whatever is needed.
and you would want regulated voltage for any sensor boards, so, if you can power them off the same voltage as your tiny computer runs at, that would be ideal. (so 3.3V or 5V ?)

the ultrasonic sound sensor is just a microphone with amplification, the output is then converted to digital and then run through an FFT to spit out the frequencies, then check the levels at the required frequency for what plant you are trying to water. at some set level that is maintained for some preset amount of time, then you water the plants.
if you have access to the raw data coming in for the ultrasonic range sensors, you can likely use that board.

Andrew Barney

  • Full Member
  • *****
  • Posts: 528
  • Karma: 43
  • Northern Colorado, Semi-Arid Climate, USA
    • Pea Breeding, Watermelon x Citron-melon, Purple Foliage Corn, Wild Tomatoes
    • View Profile
    • My blog
    • Email
  • Koppen zone: Dfc / Dfb
  • Hardiness Zone: 5b
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #12 on: 2020-01-05, 05:58:25 PM »
I would assume that most people would set up a 12V solar battery system, the latching valves run on 12V, so not many other easy options.
6V might be more efficient... but if you get many days without sun, you might want the additional discharge ability that you have if dropping down to 5V.. and then up converting for the 12V would be less efficient.
and a DC to DC buck converter does not loose much
the LM2596 converter boards are not much cost and will down convert to whatever is needed.
and you would want regulated voltage for any sensor boards, so, if you can power them off the same voltage as your tiny computer runs at, that would be ideal. (so 3.3V or 5V ?)

Yeah, I'm thinking in my head it should be easy to design a board that can accept 12V (or 5V) in and connect to a linear regulator with capacitors or a DC buck converter and regulate it back down to 3.3V. I know there are 12v DC capable water valves and i think there are also 5V capable ones, though i don't think there are 3.3V powered ones, but perhaps something like a transistor connected to the main 12V would be fine to turn on the valves. I know if using one of those relay boards that would also do the trick, so I'm not concerned about it.

Today i was able to get some of the electronics soldered up and playing around with some basic code trying to hack together an early test of something. I think i have some basic code that turns on an LED when soil moisture should reach 50%. How accurate that is and if there are better ways to do it i'm not sure. But at least i have something to test here at home. I currently have about 10 pots on heat mats trying to root and grow rare raspberry cuttings, so maintaining moisture is critical at this point and having even a crude sensor to help me out shouldn't hurt. Might make a good test actually.

The code is pretty crude at the moment. It is constantly polling the soil sensor and blinking a blue led. Without sleep modes and/or polling the sensor intermittently i don't know how long it will last if running off battery power. For this first test i decided to comment out the blue led blinking code to try and save some power.

Here is the Arduino code for the Adafruit Feather (I'm actually using the Sparkfun ESP32 Thing Plus right now):
Quote
//Modified Adafruit seesaw I2C Soilsensor example
//test code with threshold value for turning on a red led when soil moisture reaches about 50%.
//Andrew Barney 1-5-2020

#include "Adafruit_seesaw.h"

Adafruit_seesaw ss;
const int ledPin2 = 12;       // pin that LED #2 is attached to
const int threshold = 400;   // an arbitrary threshold level that's in the range of the analog input
//311=Dry Air, 1016=Hand, and 515=100% Tap Water

void setup() {
  //Serial.begin(115200);
    // initialize digital pin LED_BUILTIN as an output.
  pinMode(LED_BUILTIN, OUTPUT);
  pinMode(ledPin2, OUTPUT);
  pinMode(27, OUTPUT);

  Serial.println("seesaw Soil Sensor example!");
 
  if (!ss.begin(0x36)) {
    Serial.println("ERROR! seesaw not found");
    digitalWrite(LED_BUILTIN, HIGH);   // turn the LED on (HIGH is the voltage level)
    digitalWrite(27, LOW);
    while(1);
  } else {
    Serial.print("seesaw started! version: ");
    Serial.println(ss.getVersion(), HEX);
  }
}

void loop() {
  float tempC = ss.getTemp();
  uint16_t capread = ss.touchRead(0);
  int analogValue = ss.touchRead(0);

  Serial.print("Temperature: "); Serial.print(tempC); Serial.println("*C");
  Serial.print("Capacitive: "); Serial.println(capread);
  //Serial.print("Analog0: "); Serial.println(A0);
  delay(100);
  digitalWrite(LED_BUILTIN, LOW);    // turn the LED off by making the voltage LOW
  digitalWrite(27, LOW); // this pin needs to be low to allow a ground path for the second LED

// if the analog value is high enough, turn on the LED:
  if (analogValue < threshold) {
    digitalWrite(ledPin2, HIGH); // Turn on RED LED (LED #2)
    //digitalWrite(LED_BUILTIN, HIGH);
    digitalWrite(27, LOW); // this pin needs to be low to allow a ground path for the second LED
  } else {
    digitalWrite(ledPin2, LOW);
    digitalWrite(27, LOW);
  }
  delay(50);
/*
digitalWrite(LED_BUILTIN, HIGH);   // turn the LED on (HIGH is the voltage level)
  digitalWrite(27, LOW); // this pin needs to be low to allow a ground path for the second LED
  delay(50);                       // wait for a second
  digitalWrite(LED_BUILTIN, LOW);    // turn the LED off by making the voltage LOW
  digitalWrite(27, LOW); // this pin needs to be low to allow a ground path for the second LED
  delay(50);
*/
}
« Last Edit: 2020-01-05, 06:24:32 PM by Andrew Barney »

Andrew Barney

  • Full Member
  • *****
  • Posts: 528
  • Karma: 43
  • Northern Colorado, Semi-Arid Climate, USA
    • Pea Breeding, Watermelon x Citron-melon, Purple Foliage Corn, Wild Tomatoes
    • View Profile
    • My blog
    • Email
  • Koppen zone: Dfc / Dfb
  • Hardiness Zone: 5b
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #13 on: 2020-04-19, 04:29:20 PM »
An update for those interested in the progress of this project.

I haven't done a whole lot on my own, but I've been integrating more with the Ham lab and team at CSU. In spite of (or rather partially because of) COVID-19 progress on the soil sensor boards and carrier boards have continued to progress at a rapid rate. The timeline for testing here in Colorado is aiming for June. Professor Ham is testing 3 different versions of new low cost capacitive sensors each estimated at a cost of $4.50 USD per soil moisture sensor! Nothing i am aware of currently compares to this low cost. When putting many sensors in a field of crops this number makes a huge difference as the cost can add up quick.

Quote
Here is the summary of changes

 

Soil sensors

Now reads five soil sensors with qwiic style connectors along one side. Can read new CSU/Auburn designs (both T and water content) or more traditional probes (Soil watch 10, Vegetronix, ec-5, DFrobot, etc)

The soil sensors have switched power to conserve amps when not active.  Also they can run on 3.3V or the 3.7V lipo.  All fused for protection

No multiplexer was required to accomplish this config.

 

I2C

The board now has three I2C connectors along the other side of board, all with qwiic style conenctors using the sparkfun/adafruit wiring pattern. This will allow IRTs and RH/T probes (and other sensors) to connect. We hope to calculate water stress indices  and correlate to soil conditions. We think a common config will be 5 soil sensors, 2 IRTS, and one shielded RH/T sensor.   

 

On-board RH/T to determine if enclosure is wet inside.  Switched RH/T sensor to SHT30 so save $. I can get these for $1.90

 

User pushbutton switch changed to surface mount design

The only through hole soldering is the 12 and 16 pin headers and the on/off switch – 3 parts out of 38

 

Other features remain the same

RTC to control sleep using the EN pin (sub 200 microamps in sleep mode)

User pushbutton to control features during startup

On-off slider switch for power

TPL5010 watchdog for emergency reset if system hangs with power on. If the watch does is not “petted” in 66 min, it will do a full system  reset.

Fram memoy still included for long term backup data storage in case communication is lost.

JST battery connection

 

The total part count is about 39 pcs, including the boron and battery.  (recall early designs were about 50)

The cost of the carrier board will be very low. Likely about $1.25 each for the raw pcb. Parts for the whole board are likely $10-12 not including the boron or battery.

The new board is small – 3.35 inches (85 mm) long and 1.9 inches wide (48 mm). Would fit inside a 2” PVC (sch40) pipe for example.

 

The systems is still RF neutral, and uses the feather footprint. It should be able to operate on

Boron -  Cellular

Huzzah32 , Sparkfun Thing plus, ARGON – Wifi

Adafruit Lora M0 -  AeXonis

Adafruit adalogger -  simple data logger (no wireless) with sd card. (use our carrier board’s  RTC for a time stamp)

Andrew Barney

  • Full Member
  • *****
  • Posts: 528
  • Karma: 43
  • Northern Colorado, Semi-Arid Climate, USA
    • Pea Breeding, Watermelon x Citron-melon, Purple Foliage Corn, Wild Tomatoes
    • View Profile
    • My blog
    • Email
  • Koppen zone: Dfc / Dfb
  • Hardiness Zone: 5b
Re: Feedback on OpenAg Sensor Poll #1 - "New" Gardening Tech
« Reply #14 on: 2020-10-12, 02:07:33 PM »
I use radio where I live every day,
often use a radio that is 0.1W out on the 430MHz band (I am licensed for it),
with a large external antenna it talks to every point of land I have (and beyond)
the hand held radios I use have 1W out (or more), the antennas they have are not that great...
at 0.1W and 1/4 wave default antennas, they would have no hope of talking to each other around here.

so it looks like I can get what you linked to to work (as they do say 0.1W max),
to bad they are not higher power... would make things easier
the battery requirements and power usage are perfect

they even sell latching water valves that run on 12V
https://www.adafruit.com/product/997
https://www.adafruit.com/product/996

if it matters, I worked doing electrical engineering for a few years,
prefer analog and CMOS chips, I was away from that field when things like the Arduino came out
I would have no issues designing things like the solar power supply system and interfaces
looks like the hardware is easy now
suppose the software can be figured out with enough effort
I am all in with this project.

Do you have skill at SMD soldering? I have a board layout for the capacitance soil moisture sensors if you want to try making your own. I will be trying to make my own later this month. I have a few prototypes from CSU that seem to work well. I also just helped them design a low-cost humidity sensor board.

The lab at CSU is using particle photon boards that use cellular towers. But a new simplified board and code is being developed for the ESP32.

I think what you may want instead of "repeaters" is a radio based mesh network. There may already be some mesh networking libraries for Arduino already, which would simplify programming. I am not good at programming either, but i can muddy through some Arduino code. The big name in IOT right now are the LoRa radios. I will be testing out a simple WIFI mesh network soon with two ESP32 boards.

The nice thing about these boards (and chips) is they are able to go into a "deep sleep" mode and only wake up once an hour (or whatever you want) to save battery power. For sensors this means these boards can at the very least run on battery power alone for a whole summer, and possibly a year or more depending on settings and batteries. If a solar panel was used this could be entirely self sustaining.

https://www.instructables.com/Simple-Arduino-Wireless-Mesh/

https://nootropicdesign.com/projectlab/2018/10/20/lora-mesh-networking/

https://sandervandevelde.wordpress.com/2016/05/30/cheap-arduino-mesh-using-rf24-radio-modules/
« Last Edit: 2020-10-12, 02:14:07 PM by Andrew Barney »