What's new
  • Podcast

    Our video interview between Gypsy Nirvana and Soma is now available to watch in the podcast section. Click here to check it out.

Automation

M

medi-useA

Good one Hephaestus!


Hehehehe

What would you recommend as a good book for beginners in electronics of this type?

muA
 
M

medi-useA

Searching around and I found these...Electronic bricks...shields if you will...
ebcover.jpg

ebkit.jpg



from http://www.seeedstudio.com/depot/electronic-brick-c-48.html

muA
 

DarkLance

Member
Hey all, joining this thread a bit late, but I'd like to chime in.

I had been working on a database driven application (Java&php&mySQL) to log parameters within my grow when I stumbled on the growtronix website.

So I figured I'd solve my problem like any good american and throw money at it until it goes away. Unfortunately, I've read quite a few comments along the lines of gardenlover's, so I held off.

I think developing an in-house, open-source ICMag solution is just what is called for.

That's just how we do, 'round heer! BTW, check out VorSprung's Sprung Bags (link in sig). We ordered a set and they are just as good as the old BubbleBags I used to have.

So, my development skills are quite rusty, but I just left a job in industrial controls, working with the Atmel ATMega128, so I want in with this project!

medi-useA, you seem quite keen to invest in this project, are you trying to build this for yourself or did you intend to develop a solution to resell? If this is the latter, I think we'll need to lay down a set of requirements. Based on the growtronix product this is what I've come up with:

-Sensors
Air Temperature
Air % RH
Air % CO2
Solution pH/TDS/Temp
Liquid Level

-Warning Sensors
Water (Flood)
Smoke Detector
Motion Detector
Door/Window Contact

-External Controls
20A Switched Duplex Electrical Outlet
5V Contactor Switch

We would also need logging either onboard or externally.

I think we should constrain ourselves to one 20x4 display for now, I could not see a need for more than that. I think disfunktional asked for multiple temp levels to be visible at once, I think this could be accomidated in a simple manner with a small screen. If we build the core, expandability will be inherient, as this Arduino seems to have that ideal at its core.

Speaking of the Arduino, I seem to have a lot to read... I like how the "shield" aspect works, especially its plug and play functionality.

So the giant, 5000 kg, rabies-infected mutant gorilla in the room is the software. I'm no longer the code-junkie I was back in 1999, medi-useA, you mentioned something about mature dogs learning new tricks... From what I read, the language is based on something called 'Wiring' and is similar to C. I'll try and get ahold of some documentation and reference material and figure out what we need to do.

I'l try and be of help, I remember how to write logic statements, at least.
 

Hephaestus

Member
Once the shield is fab'd (still waiting on one chip to test the design - UPS hates me apparently) I think I've got the software handled... Know a programmer who loves this stuff - we were talking a nice windows based front end; let you just program in your parameters and settings into a nice window... Press play and the arduino gets itself a program... Keeps all the code away from us non-genius types; but the code would still be there for tweaks if needed...

Do you think we really need the 20Amp contactor? I've toyed with it repeatedly - but I'm thinking with individual power control over all important devices - it might not be totally necessary; Personally I'm a bigger fan of the big ol mushroom panic switch; for a manual shutdown if needed.

Personally I'm thinking SSRs for power control - I really like the American DJ PC-100A hack (basically mount the SSRs across the top; you get the lighted switch on the front showing current state and power state (ie switched on - it could function, lit - it is powered)). Although it's not beyond reason to have them done up in china for cheaper once the design is more worked out...
 

darkhollo

Member
Rather than making a single box that would have all sensors and software in a single unit, why not make individual sensors that just report their readings on a ethernet based network. You could easily power them over ethernet via the PoE standard. Then you could have configurable relay boxes. These boxes would have software which is configurable to listen to messages the sensors are putting out over the network. They would simply drive TTL outputs based on the logic configured by the user.

This would allow you to scale your setup to 100s of sensors if need be or just a few. This is a similar approach to the growtronix but with the use of microcontrollers in the relay controllers and sensor boxes you could remove the requirement for a traditional desktop to be controlling the mix.

I write embedded software for a living. Both on microcontrollers and larger systems. I just don't see how you are going to get around software knowledge on this. An open source icmag solution would be awesome but I have my doubts as to how that could really happen. I'm sure there are some other software nerds among us, as well as the required electrical engineers to really make this happen.

Count me in!

--dh
 
From a spectator stand point, count me in. I'm a total automation nerd although (currently using X10 but don't really trust it much) I have no electronic or programming skills my background has always been in project management. I do "get" it all so if you need a noephyte to help think it through from a noob standpoint, I will have opinions :D

I really like the system Growtronix has to offer but with digital ballasts I'm afraid to spend $1200+ on a system that may or may not work with the RF interference digital ballasts give off. The Growtronix website clear states his system may not work with digital ballasts so you take your chances there.

Rather than making a single box that would have all sensors and software in a single unit, why not make individual sensors that just report their readings on a ethernet based network.

This is a great addition for scaleability purposes.

My 2 cents.

YGB
 
I saw that some had mentioned Growtronix earlier, and I just wanted to say

DO NOT BUY GROWTRONIX!!!!!!!!!!!!!

I had ordered the system and then started to order components that were always being deliver way after ordering. I ordered the $400 TDS and PH meter and finally called after 6 months and he said he would ship it soon. anyway long story short a year later he said he shipped is but could not provide shipping receipt. in Short this is one guy building these things once he receives your money with piss poor customer service! Please do not waist your time with this company! I have emails I cam PM if you would like to decide for your self too. So basically I spent about $1000 for an incomplete system because I refuse to give home any more money. Sorry for the rant guys but that sill gets my goat.

Thank you for your review. I was looking to buy this system but have been getting a bad feeling about it from other posts I've read. It's a shame because it really does do everything.
 

Hephaestus

Member
I could start looking at going that ethernet sensor route - it'd require a laptop to do though... Bit beyond a microcontroller scenario...

But the word of caution I'd have - DS18B20 sensor is >3$ - bit of bell wire - .50$ -> none of growtronics sensors are under 30$ :joint: pH and TDS are 500$ by themselves... While it wont be completely plug n play - theory I have is to make it extremely easy to add in... You'll be able to control hundreds of sensors/relays if you see the need - but you should be able to put a system together for >200$.

Nice thing is - this system wont be affected by rf like the growtronics and digital ballasts; SSR will do a fine job of isolating. However won't have dimming capabilities on the ballast - that could be good or bad depending on your point of view...
 
I could start looking at going that ethernet sensor route - it'd require a laptop to do though... Bit beyond a microcontroller scenario...

But the word of caution I'd have - DS18B20 sensor is >3$ - bit of bell wire - .50$ -> none of growtronics sensors are under 30$ :joint: pH and TDS are 500$ by themselves... While it wont be completely plug n play - theory I have is to make it extremely easy to add in... You'll be able to control hundreds of sensors/relays if you see the need - but you should be able to put a system together for >200$.

Nice thing is - this system wont be affected by rf like the growtronics and digital ballasts; SSR will do a fine job of isolating. However won't have dimming capabilities on the ballast - that could be good or bad depending on your point of view...

Can't a laptop be used to program the micro controllers and then disconnected once programmed?

Don't get me wrong. I'm not complaining or anything, I'm just looking for the "perfect world" scenario. Some people may want to control several rooms from a single source and having all the sensors spread across the rooms feed back to a central PC for monitoring and programming would be convenient.

What you guys are doing here is awesome work and I wish I could help more. Perhaps as non-tech tester :)

YGB
 

Hephaestus

Member
Typically yes - but you still need adequate memory/processing speed to handle all those ethernet devices... Quick look tells me - the programs for handling the ethernet + sensors starts rapidly eating up available memory. (leaving less space for control/etc)

Probably what you'd do in that scenario - is a multiple room scenario - is use one arduino to each room, then have them reporting status updates back via serial back to a main pc. That I'd have to look at; not quite sure how I'd accomplish it - but there are a few options... :joint:
 
Sounds logical. Does this sound right?
1. Each room can have it's own controller that has the appropriate sensors/micro controllers needed to control that room.
2. The controller can be accessed and monitored by a central PC.
3. A central PC can monitor x number of controllers/rooms.

It seems to me that with today's computers and cheap memory it could be done using the PC power we currently have. There are similar (but expensive) PC based systems that control large professional and university greenhouses from a single PC. They even have access and control via mobile devices.

I'm no computer expert by any stretch so I don't know the limitations of PC power or how many ethernet nodes or sensors a single PC can handle (256 comes to mind). I'm also talking about my specific needs which may not be the needs of others. I'm simply following medi-useA's lead and asking for my wish list. :)

My system would control all of the sensors that everyone else asks for (lights, temp, humidity, rez water level, etc) with the ability to program/monitor/log via PC multiple rooms (clone, veg, flower)

YGB
 

Hephaestus

Member
PC control is one thing - but like I previously stated; I've done the PC control in the past - while computers have gotten better; now we have all the 'junk' that runs in the background and has a bad habit of causing unplanned shutdowns... This is why I think a micro controller is best. You could go linux where you have better control over such events - but in the end - a standalone system still seems best. Anytime you interconnect systems you're creating a instability - which is something I'd rather avoid myself...

Ethernet sensors are nice - but damn are they expensive...
 
Just thought I'd throw the ideas out there.

Forgive my ignorance, is it possible setup your parameters for all the micro controllers with a PC then disconnect and let the micro controllers take it from there until you wanted to alter a timing or temp etc? And can the micro controllers store usage data until a PC is connected and that data is downloaded to the PC for review?

Because I'm over my head here technically, I'll step back and look forward to seeing what you put together.
 

Hephaestus

Member
Yup, only need the laptop to download the program - in most cases, probably be able to adjust off the units keypad afterwords; only really need to connect to do major changes (ie adding new devices that are not already in there).

Arduino will be setup to store data - not a huge amount, I was more thinking lcd readable - it's not like we really need to have datalogged temps by the second - 30 minute trends and high/lows etc - are handy but don't need to know the temp every second over the grow... Probably could dump it to a pc if needed but; personally I dont see the need...
 

GrowerGoneWild

Well-known member
I'm working on a new grow myself, and I plan on automation to some extent, not fully automated, but more like assisted with the machine.

I quickly skimmed over the thread, and I saw interest in nutrient dosing systems, I'm not convinced that this is a great idea. Over time I've kinda streamlined this idea..

I've worked with GH, and the literature suggests solution changeout every 2 weeks. So, I think that the optimum solution is a feed to waste system, this would provide perfect feeding conditions, except the cost would be high.

I'm going assume that most of us like to keep nutrient cost under control. The more economical solution that I've come up with is auto-topping. It simply uses a float valve and a set level in the tank. When the water is used by the plants, the level drops, then the water is added back to the tank. As with most hydro systems, this is how I maintain it, watering the water.. the irony.

Now if you wanted to add more sophistication to it..

Since we know that 2 weeks is the useful life of the nutrient ( for GH) before it needs to be changed, we would use a PLC to pump out the tank, and refill it with fresh mix, and
for simplicity, I would choose a 1 part nutrient. I could simply calculate my dose for the next tank full and I could literally run 1 month without having to touch water or any nutrient.
 

darkhollo

Member
The sensors don't need configuration. An onboard dipswitch tells the micro what type of sensor it "is" and uses proper tables for reading the sensor. All the sensor box does is make a reading every 30 seconds and makes a message to be sent on the network. No configuration required so no requirement for EEPROM on that board.

The meat of the processing is done in relay control boxes. They will need EEPROM to remember their configuration logic. So you can say look at temp sensor id 38384 and light sensor id 8383 to do some logic. This box sits on the ethernet network and waits for those sensors to send out readings. It then processes that logic.

Microcontrollers can handle a shitload of lines of code. microchip.com has micros with a built in ethernet chip. As far as the network load goes.. you need 100 bytes tops per message. You could have hundreds of sensors running w/o any issue.

With EEPROM you can have configure once run w/o a computer attached designs. You could have a computer attached to the network to log if desired but it would not be making any decisions or providing anything to actually make the setup run.

You could actually have a logging computer on a network and just have sensors. You could pay the price for that.. and as your grow needs change add automation. Keeps the price down on sensors, relays etc. Each sensor should an lcd or 4 character display to show the reading so if you are in the room you can read temp, ph or whatever w/o the need for a puter.

--dh
 

Hephaestus

Member
Dark - interesting idea - not so easy to implement, I'm thinking my method is simpler and easier for all to hack...

GGW - Better solution; why not just run ala aquarium - run automated nute changes... Setup the sump to overflow, then once a week - have the pumps shut down, open a valve to drain... Close the valve and turn on a pump (or open a valve) to refill with new, close when full.
 

GrowerGoneWild

Well-known member
GGW - Better solution; why not just run ala aquarium - run automated nute changes... Setup the sump to overflow, then once a week - have the pumps shut down, open a valve to drain... Close the valve and turn on a pump (or open a valve) to refill with new, close when full.

I already defined that. Besides you would have to continually add water to the tank to maintain PPM.

Anyways. I'll explain again.

My tank is currently rigged up to fill to a set level using a mechanical float. If at any time the level drops, water is added back to the tank

There is no PPM/PH dosing. But I dont believe I need a doser.

So heres the method. But more explicit.

-I can set my PLC ( I use Allen Bradley PICO controllers) to dump nutrients buy using a pump in the tank every 2 weeks.
Simple 14 day timer.

-My dedicated drain pump will run for a designated ammount of time, say 4 hours. Timer triggered event, where I specify the duration of the event.

-The mechanical pump will continually fill the tank for 4 hours, so in effect this will "flush" the tank.

-At the end of the 4 hours I can signal a secondary pump with a premeasured ammount of nutrient, say NSR Greenleaves.. A one part bloom nutrient, that I premeasure and dilute.

For example: Say it takes 200 ML of NSR to bring up a tank of 40 gal to 1500 PPM. I would simply premeasure 200 ML and dilute it with some water, and put in a container and that pump would pump the entire contents of that to the main tank.

My dump pump and my dosing pump would be cheap, they dont have to be anything special, they would only be used a few times a month.

So the only thing I would have to do is just keep refilling the premeasured nutrient ammount, and verify PPM's/PH in the tank.

The cost for something like this is reasonable.

Toilet tank valve- 10 USD
Hardware and lines 30
Fountain Pump (2) 20-80
PLC 50 - 300


Now there may be a system out there, but I'm not aware of it.

BTW- I was thinking of posting my PLC program here, but theres not alot of people interested in this style of PLC, I'm not sure if anybody would care.

For what its worth.. I'm already using some of these ideas in my system. My auto-topper was completed today, and I just need to attach a hose to it..
 

darkhollo

Member
Hephaestus - I agree this is not the shortest path. Just trying to add my .02. The more generic you make the design, the more it will be usable by the masses. Try to think of all the use-cases for these devices and go back from there.

Also I'm an embedded c programmer. I'm nearly bare-metal in my line of work so sometimes I lose scope of what's "complicated" in this area of interest. :) I'm doing this kinda stuff everyday at work. I just don't have a good EE friend to enlist to make my boards. :(

--dh
 
M

medi-useA

I go away a short while, off chasing info 'cross the 'net, and when I get back to my 'lil 'ol Automation thread I find out...

It's Alive!
young_frankenstein_doc_small.jpg

young_frankenstein_kobal-2145.jpg



DarkLance
darkhollo
YouGrowBoy
GrowerGoneWild

Welcome to the Discussion!

I started this thread as a learning tool and to put something together th@ others could use as a reference.

As you can see if you've been reading the thread, I got a bit ...side tracked on the Arduino...

an idea...motion sensors to detect the borg! or IR..and a sound repellant!Sorry, they come to me and I blurtem out!


I think developing an in-house, open-source ICMag solution is just what is called for.

DarkLance...You just made my Sig :)


So, my development skills are quite rusty, but I just left a job in industrial controls, working with the Atmel ATMega128, so I want in with this project!
Funnily enough, the Aduino uses the Atmel ATMega128 and others in the series...:smokeit:



medi-useA, you seem quite keen to invest in this project, are you trying to build this for yourself or did you intend to develop a solution to resell?
The idea is not repellant to me, but I aim to learn to build and use the controller first...and make info all available free online...It is the very freedom of the open source/licence system th@ is allowing this to be possible...but if a suitable compromise could be made, great, but if not...others will still benefit from the research and efforts of all involved.

...I think we'll need to lay down a set of requirements. Based on the growtronix product this is what I've come up with:

-Sensors
Air Temperature
Air % RH
Air % CO2
Solution pH/TDS/Temp
Liquid Level

-Warning Sensors
Water (Flood)
Smoke Detector
Motion Detector
Door/Window Contact

-External Controls
20A Switched Duplex Electrical Outlet
5V Contactor Switch

We would also need logging either onboard or externally.

I think we should constrain ourselves to one 20x4 display for now, I could not see a need for more than that. I think disfunktional asked for multiple temp levels to be visible at once, I think this could be accomidated in a simple manner with a small screen. If we build the core, expandability will be inherient, as this Arduino seems to have that ideal at its core.

Speaking of the Arduino, I seem to have a lot to read... I like how the "shield" aspect works, especially its plug and play functionality.
I do like the logging idea...I was going to go basic for the display until I am comfortable with the unit and process of building it...KISS!

I also like the plug'N'play aspect...I've been looking into the Xbee system, the rf Adunio system and a few others. I want to be able to place a unit in or near my cabs, place sensors and walk away...I want to be able to monitor the system from a set place mounted outside the cabs and I want to have a portable remote device th@ sounds an alarm, sends txt mssgs and can control/shutdown remotely....I don't want much, do I??? :tree:

There are about 6 different ways to make Aduino Controllers talk to each other and to a computer or us!...all are valid.

So the giant, 5000 kg, rabies-infected mutant gorilla in the room is the software. I'm no longer the code-junkie I was back in 1999, medi-useA, you mentioned something about mature dogs learning new tricks... From what I read, the language is based on something called 'Wiring' and is similar to C. I'll try and get ahold of some documentation and reference material and figure out what we need to do.
I've found a few sites with wiring tutes...will post em soonest.
Thanks for volunteering...I yell and swear @ my pooter in 4 different Languages, and I still can't get it to work properly! :muhaha:


Once the shield is fab'd (still waiting on one chip to test the design - UPS hates me apparently) I think I've got the software handled... Know a programmer who loves this stuff - we were talking a nice windows based front end; let you just program in your parameters and settings into a nice window... Press play and the arduino gets itself a program... Keeps all the code away from us non-genius types; but the code would still be there for tweaks if needed...
Great to hear Hephaestus...There are so many ways to control, so many things to control and so many styles to control th@ a point and click interface will be prized!.:xmasnut:


Do you think we really need the 20Amp contactor? I've toyed with it repeatedly - but I'm thinking with individual power control over all important devices - it might not be totally necessary; Personally I'm a bigger fan of the big ol mushroom panic switch; for a manual shutdown if needed.
I do like the 20 amp contactor, if only for paranoid safeties sake!...as I LOVE the BRB idea...[Big Red Button]....everything is optional! :)


Rather than making a single box that would have all sensors and software in a single unit, why not make individual sensors that just report their readings on a ethernet based network. You could easily power them over ethernet via the PoE standard. Then you could have configurable relay boxes. These boxes would have software which is configurable to listen to messages the sensors are putting out over the network. They would simply drive TTL outputs based on the logic configured by the user.

This would allow you to scale your setup to 100s of sensors if need be or just a few. This is a similar approach to the growtronix but with the use of microcontrollers in the relay controllers and sensor boxes you could remove the requirement for a traditional desktop to be controlling the mix.
I too embrace this idea...I like the modularity of it...scale it up or down as required...interchaneable bits'N'bobs, sensors and comm devices.


his would allow you to scale your setup to 100s of sensors if need be or just a few. This is a similar approach to the growtronix but with the use of microcontrollers in the relay controllers and sensor boxes you could remove the requirement for a traditional desktop to be controlling the mix.

I write embedded software for a living. Both on microcontrollers and larger systems. I just don't see how you are going to get around software knowledge on this. An open source icmag solution would be awesome but I have my doubts as to how that could really happen. I'm sure there are some other software nerds among us, as well as the required electrical engineers to really make this happen.

Count me in!
Thanks dh....You're IN! :) welcome to the great experiment..hehehe


From a spectator stand point, count me in. I'm a total automation nerd although (currently using X10 but don't really trust it much) I have no electronic or programming skills my background has always been in project management. I do "get" it all so if you need a noephyte to help think it through from a noob standpoint, I will have opinions
You're coming armed with more exp than I, ..welcome aboard the Idea & Concepts Train...Next stop, PROTOTYPING!...hehehe


I really like the system Growtronix has to offer but with digital ballasts I'm afraid to spend $1200+ on a system that may or may not work with the RF interference digital ballasts give off. The Growtronix website clear states his system may not work with digital ballasts so you take your chances there.
Th@'s a good discussion point!


We need a Do's and Don't's list for things like this.

#1 Buy From Reputable Dealers if @ all possible.


Thank you for your review. EDIT: Originally Posted by gardenlover I was looking to buy this system but have been getting a bad feeling about it from other posts I've read. It's a shame because it really does do everything
I think Th@ was more the dealer than the product. Unfortunately in this world the predators, idiots and fly-by-nighters are there to snag the unwary...Fuck 'em...lets build something better!


I could start looking at going that ethernet sensor route - it'd require a laptop to do though... Bit beyond a microcontroller scenario...

But the word of caution I'd have - DS18B20 sensor is >3$ - bit of bell wire - .50$ -> none of growtronics sensors are under 30$ pH and TDS are 500$ by themselves... While it wont be completely plug n play - theory I have is to make it extremely easy to add in... You'll be able to control hundreds of sensors/relays if you see the need - but you should be able to put a system together for >200$.

Nice thing is - this system wont be affected by rf like the growtronics and digital ballasts; SSR will do a fine job of isolating. However won't have dimming capabilities on the ballast - that could be good or bad depending on your point of view...
As I mentioned before, Hep...Put me down for 2!...even a bare boards and a list! hehehe...First I learn...Then I experiment!...Getting my self an Aduino starter kit for Christmas...I Hope!


Can't a laptop be used to program the micro controllers and then disconnected once programmed?

Don't get me wrong. I'm not complaining or anything, I'm just looking for the "perfect world" scenario. Some people may want to control several rooms from a single source and having all the sensors spread across the rooms feed back to a central PC for monitoring and programming would be convenient.

What you guys are doing here is awesome work and I wish I could help more. Perhaps as non-tech tester

YGB
The 'Perfect World' scenario is different for every one...best bet is to lurk and learn...see what mistakes or right things others are doing, and take it from there.
One thing I would recommend is get an Aduino, power supply and and an intro to Aduino kit and complete the projects...they should give a basic overview of Aduino...It's Wh@ I intend to do.

Search for Aduino PC's...there are some out there


Now there may be a system out there, but I'm not aware of it.

BTW- I was thinking of posting my PLC program here, but theres not alot of people interested in this style of PLC, I'm not sure if anybody would care.

For what its worth.. I'm already using some of these ideas in my system. My auto-topper was completed today, and I just need to attach a hose to it..
I WANT YOUR IDEA'S HERE GGW!...this is the forum I started for 'em!...I've not looked hard @ PLC's yet myself...caught up in my Aduino Hunt...but I am going to look into them in the new year and start posting info on their use...Wh@ you have provided is great.


Also I'm an embedded c programmer. I'm nearly bare-metal in my line of work so sometimes I lose scope of what's "complicated" in this area of interest. I'm doing this kinda stuff everyday at work. I just don't have a good EE friend to enlist to make my boards. :(
Stick around dh...your services and views will be needed. I appreciate your time...




Everyone has been civil and th@'s GREAT!

But, late, I admit, LET ME STATE THIS for all who come after...
gee, I feel like a grump for saying this! :(

We are not here to Argue...we're all coming @ it from a different angle/background/agenda/POV...if PLC's aren't your thing, just follow the conversation and you might learn something enlightening...or applicable to your situation. Same for all the other Automation Methods. By discussing this in open forum we can codify Automation so th@ all may approach it....be it someone in a 3rd world country who's building something from scratch with salvaged components...to hobbyists....to hardcore sumbitche$!...
We all get along here we all share..we all play nice! :)

Drop in anytime.!:eggnog::smokeit:

muA
 
Top