IoT and automation in homesteading. OR how I'm going to design/build and open-source gardening.

Site Notices

[ARCHIVED THREAD] - IoT and automation in homesteading. OR how I'm going to design/build and open-source gardening.

Posted: 6/8/2019 1:16:24 PM EDT

I'm working on a work project right now that hopefully once done will open up time/resources to start developing and testing automation systems designed specifically for the large family garden and/or small market gardener.

The aim will be to

eliminate as much labor as possible, automating as much as possible, including weeding between rows (and hopefully between plants one day)
eliminate the need for organic/conventional pesticides/herbicides by developing indoor growing systems, even if they are just simple row hoop houses and filtering all air exchange
fully automate irrigation
whatever else reduces labor/cost and increases productivity.

Yes, a lot of this stuff is done already, but much of it that is available for free that I have found isn't too refined, and usually still on breadboard phase. The refined stuff I found is very expensive, aimed at commercial growers. I plan on developing PCBs and BOMs, making those files open-source, along with code to run it all.

First order of business: For a school project, my son will be working on growing carrots in sand substrate hydroponically, and running variable of sand moisture and temp to find an ideal way to grow carrots in sand. He will be running it all electronically/automatically off of an Arduino platform. Its a mix of his botany and electronics classes (we homeschool). Part of that will be building a PCB to put this system on, instead of jumper wires and breadboards.

At some point, I'll start a website just for it, with all the info and files and test results.

Right now, I'm 3D printing custom seed plates for my planter. Technology is amazing.

Posted: 6/8/2019 3:39:57 PM EDT

[#1]

OK, you've got my attention...

(I'm an electronics/microcontroller/maker/tomato-grower hobbyist and so far my only plans for the upcoming gardening season are to control valves to add water to the reservoirs of my Earthboxes based on triggers from small float switches, and maybe take it a step further by monitoring soil moisture in each box - but obviously there's plenty more that could be done beyond that...)

Posted: 6/8/2019 3:57:44 PM EDT

[#2]

Tagged

Posted: 6/8/2019 4:00:54 PM EDT

[#3]

I'm about to start building and installing moisture sensors and additional sprinkler controls for my wife's garden based on them as the stock rainbird devices simply don't cut the mustard.

There's a huge market for this stuff I suspect.

Posted: 6/8/2019 6:08:00 PM EDT

[#4]

Quote History

Quoted:
I'm about to start building and installing moisture sensors and additional sprinkler controls for my wife's garden based on them as the stock rainbird devices simply don't cut the mustard.

There's a huge market for this stuff I suspect.

View Quote

If this work project pans out, my plan is to keep this open source and all the info free. Hopefully I would actually start a non-profit for it.

If my work project falls apart for some reason, I may have to sell the stuff to keep the research going.

Either way I'll be working on automating all my stuff, and that tech will be put out there somehow.

Posted: 6/8/2019 6:50:13 PM EDT

[#5]

This sounds like a really cool project, and I particularly appreciate that you’re able to combine this with your son’s home Ed.

Posted: 6/9/2019 12:08:40 PM EDT

[#6]

So this is what you're wanting?

Posted: 6/9/2019 12:43:44 PM EDT

[#7]

Quote History

Quoted:
So this is what you're wanting?

View Quote

Seen that.

Sort of. I appreciated that it was open source, but their sale price of $4000 was kind of really steep for the amount of garden you get.

Plus lots of other automation devices, like my automatic chicken coop door opener/closer based on daylight.

Also just lots of research on what methods produce best and other research.

Posted: 6/9/2019 4:15:17 PM EDT

[#8]

With a 3D printer and sourcing your own hardware you can build one for half as much...possibly less. They have the full BOM and code available.

Still a big pill to swallow but it's the most complete solution I've seen so far. I'm considering a DIY version myself. Last time I priced it out I got to around $1500 and needed a few more things to finish it off.

Posted: 6/9/2019 7:13:56 PM EDT

[#9]

Quote History

Quoted:
With a 3D printer and sourcing your own hardware you can build one for half as much...possibly less. They have the full BOM and code available.

Still a big pill to swallow but it's the most complete solution I've seen so far. I'm considering a DIY version myself. Last time I priced it out I got to around $1500 and needed a few more things to finish it off.

View Quote

But is the money better spent on an inert substrate hydroponic system where it's only simple, stationary controls?

Some of the questions I'd like to answer.

Posted: 6/10/2019 12:31:32 AM EDT

[#10]

One of the metrics I want to keep track of is all these systems like hydroponic gardens is the comparison to my "traditional" regular dirt garden.

I want to keep track of hours worked in the dirt garden (things global tractor hours for day pre planting cultivating, and then specific for hours worked weeding onions, etc), and cost, including cost of equipment.

That will always give me a control to see what systems have true value vs regular dirt farming.

Posted: 6/10/2019 3:17:01 PM EDT

[#11]

Might want to look into aquaponics as well. High. Seems like an interesting concept to me and you can harvest the fish for meat. From what I've read it can be a bit delicate in terms of balancing the ecosystem. The one thread on here was an interesting read. Just wish I had the ability to do that at my home. I'm pretty sure the HOA might have a shit fit if I attempted something like that

Posted: 6/18/2019 12:39:51 PM EDT

[#12]

Here is the PCB for his hydroponics control. Kiddo helped, until I completely changed the design to save time/cost/PCB size

. The original schematic called for actual relays ($1.60 each), and the PCB reached a size of about 170x170mm (at a cost of $40 for 5 boards). Total cost for the project was creeping up to $80 with components. The new MOSFETS are some 30 cents each, and the size is now 90x70mm (at a cost of $5 for 5 boards). Now, the PCBs will be around $14 in PCB and component cost. Much better.

Its Arduino Nano based, as the Chinese imports are the same cost as a bare controller from Mouser or DigiKey and only slightly bigger than a DIP-28. There is an LED at every "switch" that turns on when that FET is turned on, just for a bit of de-bugging/diagnostics. It also has an OLED screen for some informational displays. We didn't "need" the LEDs or OLED, but this PCB will accept it, and you don't have to use/install them. It also has a buzzer for buzzer type needs (alerts).

It has 6 sensors for soil moisture (or any 5v I2C sensor actually, they are generic COM ports), 6 sensors for temp (or any 1 wire digital temp), and 16 MOSFET "switches" that can handle up to 4 amps (if kept cool; short duration), but we will limit to 1 amp-ish.

Attached File

Posted: 6/18/2019 12:45:51 PM EDT

[#13]

This board is meant for short length sensor wire runs (I2C specifically is no more than 2-3 meters normally). Once it turns into a full system, I will incorporate some bus repeaters to get a hundred yards or so of wire length usability.

Posted: 6/18/2019 12:48:34 PM EDT

[#14]

I have been looking at doing hydroponics in a small greenhouse to see if I even can get it to work!

I planted in raised beds for the past 2 years with no yield just because the damn rain has killed and flooded everything.

Posted: 6/19/2019 8:51:05 AM EDT

[#15]

The board looks really good. I'm still pretty new to arduino, is the nano going to be able to handle all of this? Or are you going to have it relaying data back to a computer elsewhere?

Posted: 6/19/2019 10:01:46 AM EDT

[#16]

Quote History

Quoted:
The board looks really good. I'm still pretty new to arduino, is the nano going to be able to handle all of this? Or are you going to have it relaying data back to a computer elsewhere?

View Quote

Thanks!

Oh yeah, the ATMEGA328 (the actual controller in a nano) can handle this no problem. Its just reading sensors, some IF structures, and turning pins HIGH/LOW.

Fun fact: the ATMEGA328 in the 32 TQFP actually has 2 more pins (A6, A7) than the Uno running the DIP28 version 328.

Posted: 6/19/2019 12:20:45 PM EDT

[#17]

Yeah I figured the Nano was doing all the lower level stuff. Are you going to have remote monitoring capabilities with this? Would figure you'd need either a WiFi chip or possibly LORA if you're looking to stretch out further.

Posted: 6/19/2019 12:35:07 PM EDT

[#18]

Board looks great! What did you design it in?

For someone who doesn't know about hydroponics, could you summarize the functions of the system and the control logic?

Posted: 6/19/2019 3:49:37 PM EDT

[#19]

Quote History

Quoted:
Board looks great! What did you design it in?

For someone who doesn't know about hydroponics, could you summarize the functions of the system and the control logic?

View Quote

I used KiCad for the design.

The controller will monitor the temps and "soil" moisture level readings, and then make corrections as needed.

Posted: 6/19/2019 8:30:47 PM EDT

[#20]

Very cool.

Tag just for interest and education.

Rat_Patrol did you finally perfect the seeder tray you were working on before, showing us here on the forum?

Posted: 6/19/2019 9:20:34 PM EDT

[#21]

Quote History

Quoted:
Very cool.

Tag just for interest and education.

Rat_Patrol did you finally perfect the seeder tray you were working on before, showing us here on the forum?

View Quote

I can't get raw carrot seeds to run through that seeder with any kind of reliability. Any small seed, actually. Either it gets stuck on seeds near constantly, or I'm dropping 4-6 seeds at a time. Can't find any happy medium.

One of my planned projects: i'm going to make an automatic seed tape making machine. Load your seeds into a hopper, machine deposits glue (flour/water), dispenses the paper tape (TP?) and deposits a single seed on the dab of "glue" at a time via vacuum pickup, making as long of tape at whatever spacing you need.

Then just plant your seed tape.

Posted: 6/19/2019 10:25:15 PM EDT

[#22]

Quote History

Quoted:
I can't get raw carrot seeds to run through that seeder with any kind of reliability. Any small seed, actually. Either it gets stuck on seeds near constantly, or I'm dropping 4-6 seeds at a time. Can't find any happy medium.

One of my planned projects: i'm going to make an automatic seed tape making machine. Load your seeds into a hopper, machine deposits glue (flour/water), dispenses the paper tape (TP?) and deposits a single seed on the dab of "glue" at a time via vacuum pickup, making as long of tape at whatever spacing you need.

Then just plant your seed tape.

View Quote

Very interesting concept. I'm familiar with seed tape, and the tp idea has merit, though I would guess you'll need to experiment with kinds of tp. They seem to vary WIDELY in strength and thickness, and their tendency to dissolve completely in water varies also, which thickness/resistance may or may not present any kind of significant barrier to the radicle as it emerges from the seed.

In the floriculture industry, we used needle seeders to fill seed trays, and some of them are capable of planting the tiniest of uncoated seeds (that I have seen, anyway).

I have wondered if that needle concept could be adapted to a row planter, kind of like the tobacco setters I grew up using (transplanter with rubber "fingers" (to receive the transplant) on a belt-driven wheel, timed with gears for spacing and opening/closing the fingers)

So if the needles were placed on a wheel....

Anyway, I think your seed tape is simpler and has a lot of possibility.

Posted: 6/20/2019 11:11:38 AM EDT

[#23]

I'm thinking RV toilet paper, since its thin and dissolves "rapidly".

That machine will actually be easy, electronically. Have a function to move a vacuum needle into a cone shaped hopper, sensor to read when vacuum is formed (indicating a seed is attached to the needle tip) which stops the travel into the hopper, function to move and deposit seed onto a pre-deposited "glue" (another function, but that applicator will be just traveling in the Z axis), steppers to move the TP along, and a collection spool that is a small electric motor with a comically loose friction clutch (just enough to roll up the seed tape). Simple folding dies and a roller will fold the TP over onto itself. I'm also thinking a very slight burst of air will shoot the seed away from the needle (very small breath of air) so the needle never has to contact the "glue". The vacuum can actually pull on a spring loaded plunger to create the little bit of back pressure needed.

A planter specifically built for it would be easy enough too. Just anchor the start of the row with a clump of dirt or rock or something, drive along (most seeds this small are only covered with 1/4" of soil anyway), and have some sweeps push some soil onto the seed tape.

Posted: 6/22/2019 9:14:37 PM EDT

[#24]

@Rat_Patrol,

I am wormholed on Arduino/RasPi currently for work on a number of significant initiatives.

Hopefully some of this may help:

1. Check out MySensors open source for a framework to be able to rapid develop your software + sensor network. It is FANTASTIC and I would run everything through MQTT if I were you.

2. Check out mcsSprinklers. It’s a stand-alone software but also a plugin for Homeseer. INSANELY capable system that makes use of EtherRain controllers for irrigation valves. The EtherRains are cheap enough that you should maybe consider running those to get going.

3. Check out IOBroker, an open source automation system that has heavy use in Germany. It’s pretty great, it has a MQTT plugin and it also has a InfluxDB plugin.

4. If you don’t know it already, check out ‘TIG Stack’ for storing and presenting your logging data. :)

5. I may end up going with some custom PCB shields for my Arduinos for my project. If I do I may PM you and see if you want to help design for some $$$. I’m a software jock and too old to start cold with hardware.

Cheers!!

Posted: 6/22/2019 11:20:55 PM EDT

[#25]

Quote History

Quoted:
@Rat_Patrol,

I am wormholed on Arduino/RasPi currently for work on a number of significant initiatives.

Hopefully some of this may help:

1. Check out MySensors open source for a framework to be able to rapid develop your software + sensor network. It is FANTASTIC and I would run everything through MQTT if I were you.

2. Check out mcsSprinklers. It's a stand-alone software but also a plugin for Homeseer. INSANELY capable system that makes use of EtherRain controllers for irrigation valves. The EtherRains are cheap enough that you should maybe consider running those to get going.

3. Check out IOBroker, an open source automation system that has heavy use in Germany. It's pretty great, it has a MQTT plugin and it also has a InfluxDB plugin.

4. If you don't know it already, check out 'TIG Stack' for storing and presenting your logging data. :)

5. I may end up going with some custom PCB shields for my Arduinos for my project. If I do I may PM you and see if you want to help design for some $$$. I'm a software jock and too old to start cold with hardware.

Cheers!!

View Quote

I'll be checking those all out, thanks!

I'm more of a hardware guy, just slugging though the software.

Truth be told, I lean what I need to know to accomplish the task I have in front of me. I'm 100 percent self taught.

I may be up for trade: some software work in exchange for some PCB layouts

. Shields aren't bad, and assembly by hand isn't bad as long as you stay at 0803 and bigger stuff. I'm hoping to get a pick and place machine for work before long too.

ETA: so, in theory, how hard would it be to create a fake engine cam signal that can be adjusted on the fly for different "rpm"? I'm going to be grabbing the cam signal from an engine once I upgrade to a digital scope, and need a simulated signal for a project.

Posted: 6/22/2019 11:28:57 PM EDT

[#26]

Quote History

Quoted:
Yeah I figured the Nano was doing all the lower level stuff. Are you going to have remote monitoring capabilities with this? Would figure you'd need either a WiFi chip or possibly LORA if you're looking to stretch out further.

View Quote

Sorry I missed this earlier.

No data logging or remote monitoring at present.

I don't like WiFi, so once I want remote monitoring, I'll either put it on my LAN or run Murs radio to a receiver in the house.

I have cat 5 everywhere around here...

Posted: 6/24/2019 1:41:58 AM EDT

[#27]

You've seen the TED Talk from Caleb Harper right?

If not, sounds like he went down this same road a few years back with a team at some big name college. He's open source too.

Posted: 6/24/2019 2:17:23 AM EDT

[#28]

Quote History

Quoted:
You've seen the TED Talk from Caleb Harper right?

If not, sounds like he went down this same road a few years back with a team at some big name college. He's open source too.

View Quote

I just checked it out

Neat, but fun what I saw, they are more focused on nerd than food.

If I read their wiki right, $500 to build a box to grow a single head of lettuce ???

That can't be what mit put out as a solution.

Posted: 6/24/2019 4:14:45 PM EDT

[#29]

Sending out for an order of PCBs.

Some are for work, some are for other projects yet, but also the sensor/FET board will be among them.

I added clamping diodes across each output channel for inductive loads (solenoids). Don't have to populate the pads, but the pads are there.

The "J" connectors (output channels) line up to the Arduino IDE pins: J3 is D3, JA1 is A1. Allows to easily keep track of what channel is controlled by what pin.

At some point in the future, I think I will convert to BJTs instead of MOSFETs to current control to each channel. Its just that I have these MOSFETS on hand anyway, so may as well use them instead of buying more components.

Last render before I send out the gerber files:

Attached File

Posted: 6/24/2019 4:16:58 PM EDT

[#30]

Very simple schematic actually.

ETA: You can probably tell I like using pins/labels instead of wires

. Just makes for a clean schematic, especially for bus lines.

Attached File

Posted: 6/24/2019 4:55:23 PM EDT

[#31]

You can add a PAR sensor to that as well...

https://www.apogeeinstruments.com/original-quantum-sensor/

Posted: 6/24/2019 4:59:01 PM EDT

[#32]

Quote History

Quoted:
You can add a PAR sensor to that as well...

https://www.apogeeinstruments.com/original-quantum-sensor/

View Quote

looks like a broad spectrum light meter?

ETA: Oofta, just saw their prices.

I think the point is to see how much of what spectrum you are getting. That could be configured with a multitude of narrow spectrum sensors, put onto their own sensor array, and comm the assembly as an I2C into this board. Probably build something for around $30 instead of the $500 they are asking.

Once I go for a full system control, I'm going to configure it on a much larger microcontroller (such as Arduino MEGA platform) to allow for tons of I/O and data.

Posted: 6/24/2019 9:51:30 PM EDT

[#33]

Quote History

Quoted:
Sorry I missed this earlier.

No data logging or remote monitoring at present.

I don't like WiFi, so once I want remote monitoring, I'll either put it on my LAN or run Murs radio to a receiver in the house.

I have cat 5 everywhere around here...

View Quote View All Quotes

View All Quotes

Quote History

Quoted:

Quoted:
Yeah I figured the Nano was doing all the lower level stuff. Are you going to have remote monitoring capabilities with this? Would figure you'd need either a WiFi chip or possibly LORA if you're looking to stretch out further.

Sorry I missed this earlier.

No data logging or remote monitoring at present.

I don't like WiFi, so once I want remote monitoring, I'll either put it on my LAN or run Murs radio to a receiver in the house.

I have cat 5 everywhere around here...

@Rat_Patrol,

MySensors has a mesh topology WiFi system and can also do 433mhz radio. It is elegant and quite ingenious. Dive on this one I promise you.

Also, look at the Arduino battery shields that have a library and allow the whole Arduino to sleep. So you can set the thing to wake up every 30 seconds, 5 minutes, 1 hour, whatever and report it. It will DRASTICALLY lengthen your battery life between recharges.

https://www.mysensors.org/

https://www.amazon.com/dp/B01N7YJL12/ref=cm_sw_r_cp_api_i_V3xeDb3S1CF0X

Posted: 6/24/2019 10:24:00 PM EDT

[#34]

Quote History

Quoted:
@Rat_Patrol,

MySensors has a mesh topology WiFi system and can also do 433mhz radio. It is elegant and quite ingenious. Dive on this one I promise you.

Also, look at the Arduino battery shields that have a library and allow the whole Arduino to sleep. So you can set the thing to wake up every 30 seconds, 5 minutes, 1 hour, whatever and report it. It will DRASTICALLY lengthen your battery life between recharges.

https://www.mysensors.org/

https://www.amazon.com/dp/B01N7YJL12/ref=cm_sw_r_cp_api_i_V3xeDb3S1CF0X

View Quote View All Quotes

View All Quotes

Quote History

Quoted:

Quoted:
Yeah I figured the Nano was doing all the lower level stuff. Are you going to have remote monitoring capabilities with this? Would figure you'd need either a WiFi chip or possibly LORA if you're looking to stretch out further.

Sorry I missed this earlier.

No data logging or remote monitoring at present.

I don't like WiFi, so once I want remote monitoring, I'll either put it on my LAN or run Murs radio to a receiver in the house.

I have cat 5 everywhere around here...

@Rat_Patrol,

MySensors has a mesh topology WiFi system and can also do 433mhz radio. It is elegant and quite ingenious. Dive on this one I promise you.

Also, look at the Arduino battery shields that have a library and allow the whole Arduino to sleep. So you can set the thing to wake up every 30 seconds, 5 minutes, 1 hour, whatever and report it. It will DRASTICALLY lengthen your battery life between recharges.

https://www.mysensors.org/

https://www.amazon.com/dp/B01N7YJL12/ref=cm_sw_r_cp_api_i_V3xeDb3S1CF0X

I'll check out the mysensors page.

If I need battery, I figured I would run a pair of 18650 batteries (about 4000mah total capacity) and just use a standard charge controller to re-charge from wall wart or small solar panel. I planned on 18650's just due to cost over the shields (unprotected 18650's are only a few bucks each now), the added capacity is just bonus. Checking on things every 30 minutes even if nothing is on (run the code where it only goes to sleep if its not "doing something", sleep, wake up and check stuff, if good sleep again, etc.)

All good ideas though, I'll be making a list here to keep track of this stuff. I imagine the final product will be a decent sized board with all sorts of goodies on it, or able to accept all sorts of goodies. That way only 1 board has to be produced, populate it with what you need.

I do want to try to keep as much as possible built on PCBs no bigger than 100mmx100mm though. That size and smaller are MUCH cheaper to have made. I just placed an order for 5 different PCB designs from my *cough*Chinese*cough* PCB build house. The shipping cost more than the 25 PCBs I have on order

[ARCHIVED THREAD] - IoT and automation in homesteading. OR how I'm going to design/build and open-source gardening.

Outdoors » Homestead, Farm & Garden

Warning

Confirm Action

About AR15.COM

Stay Connected

Newsletter

Contact Us