Grow Room Automation: Starting With Light Control

[Old Salt]

I'm building an 8 COB fixture with supplemental wavelengths and multiple drivers. There's not much on the market for light control. On/off is easily accomplished with timers, but I want more. I want independent intensity control of the various light strings in the fixture. I bought a TC421 programmable light controller, but can't connect to it by WiFi. I looked at Coralux's Storm and Storm-X. They have no way to switch the power to the drivers on and off. You can get around this by dedicating one of its analog outputs to driving a relay. That can tie up channels which could be used for better purposes. There are other controllers used by folks here I've had a look at, but they all seem lacking in one way or another. That makes this a DIY project, an I looked around for an easy way to do this.

A microcontoller seems to be an ideal solution. I have a Basic Stamp, but that's a little slow and has limited I/O capabilities. That left the Arduino or Raspberry Pi. Either one would do, but I chose the Arduino. I've gone ahead and ordered the UNO with several modules that I think I'll need for this project. To start, I won't include any RF communications such as WiFi or Bluetooth. That may be added later, once the base functionality is proven. For now it will be programmed at the front panel, or through the USB port.

So, what functionality do I want:

• Power on/off switching for Meanwell drivers
• independent intensity control for each light channel
• options of instantaneous, ramp, or sinusoidal change for each light channel
• daily cycle, with multiple cycles for each light channel possible
• power failure recovery with soft start

I'm sure others have different requirements, and I'll consider them when I play the part of a code monkey. Browsing through the available modules, and published projects for the Arduino, I had several ideas for later implementation. Automated watering, environmental controls, nutrient mixing are just some of the applications that I'm interested in down the road.

I'd like to hear your ideas. Please keep it to lighting for now.

 

[Ripe1]

Well this is freakin cool! I have a buddy who has programmed a live light show, programmed to the best of his abilities with available technology to simulate not only the exact hours of the sun, but the gradual change still in spectrum. All with midi controllers leds and live sound lights. That is about all I can offer this project outside of interest and support. Conquer!

 

[Old Salt]

I'm not sure if I'll include local sunrise/sunset timing. At the moment I'm more interested in fixed 12 or 16 hour on times. I could add it later if there's enough interest, and room on the chip. I'll keep it in mind while programming, as it might involve grouping the channels.

 

[Ripe1]

I’m not sure how he even wrote and programmed it. I think you’lol be fine with 12 and 16 hour times. It seems to work for everybody else. You just automatically remind me of a tinkering engineer like my dude. Great journal by the way!

 

[dadums]

look up gas light

 

[Old Salt]

gas light?

 

[kelticBlue]

Gas lantern routine, a unique light schedule. A link for more info. Gas lantern

 

[Old Salt]

The Arduino and relays arrived today. It took a bit to get communications established with my 64 bit Windows 7 based computer. I managed to compile and upload the Coralux Storm-X software. The other components I need haven't arrived yet, so it's still non-functional.

The first step will be getting the Coralux software to run. Once that is done, I'll start with the redesign of both the hardware and software. At this point I'd like some input as to what you'd like to see the light controller morph into.

Is the 16 X 2 LCD display good enough, or would you like to see something else?
Is the rotary encoder acceptable or would you like another method of input?
- keypad
- capacitive touch LCD
- something else?

Are eight PWM channels too many, or can we use less?
Are eight digital outputs for relays enough? Do we need more or less?
Is it worth incorporating fan control for light cooling? If so, how many channels?

The Coralux Storm-X implements lightning bolts and cloud cover. Is this desirable?

Are there any other features you'd like to see in a light control?

Note that this controller is meant to be installed in the light fixture. In the future communications to a central control unit, and then out to the world via WiFi.

 

[Old Salt]

Here are my answers to the above questions:

The 16 X 2 LCD display is good enough for local control.

Is the rotary encoder acceptable or would you like another method of input? I don't know. I'll have to try it and see for myself. I'm thinking the rotary control or push buttons for local control.

Are eight PWM channels too many, or can we use less? Four is enough for me:
- main
- supplemental deep red
- supplemental far red
- supplemental IR

Are eight digital outputs for relays enough? Do we need more or less? Two is enough for me, one for the main lights, another for UV.

Is it worth incorporating fan control for light cooling? If so, for how many channels of control? Not for me. I'm using passive cooling.

The Coralux Storm-X implements lightning bolts and cloud cover. Is this desirable? Not to me.

Other: There is a possibility of having light fixtures slaved to the primary. If this is done, the electrical connection would be made by a standard parallel port cable.

 

[Old Salt]

I'm still working on this. I switched my focus to an automatic watering system. Lights can be controlled by a timer, but watering a coco medium requires more control. It's the need to water twice a day that keeps me from making overnight trips.

Currently I have a core processing system. It consists of:
- Arduino UNO
- real time clock
- 16 X 2 LCD display
- rotary control
- 15 available I/O ports

The I/O ports are used for:
- water level sensors
- Pump and valve control

Port usage will vary by application, and will be user configurable. For a 4 plant coco grow system, with individual feeds for each plant, and a shared overflow drain pump it could look like this:

Port Use
1 Feed Pump (for ports 3 - 6)
2 overflow resevoir drain pump (shared with all plants)
3 feed valve for plant 1
4 feed valve for plant 2
5 feed valve for plant 3
6 feed valve for plant 4
7 available
8 available
9 available
10 optional soil moisture sensor for plant 1
11 optional soil moisture sensor for plant 2
12 optional soil moisture sensor for plant 3
13 optional soil moisture sensor for plant 4
14 nutrient resevoir level sensor
15 overflow resevoir liquid level sensor

As you can see, the ports get used very quickly. At this point in time it would be very easy to add more I/O ports. The design will be difficult to change once the circuit boards have been made.

If anyone else is interested in a system like this, I'd need to know how many ports you'd need. Note that multiple feed resevoirs, for example veg, flower, and flush, can easily be accomodated by this design.

So let me know how many plants you need to take care of, and I'll add to the design as needed.

BTW: once proven, I'll make this open source hardware and software.

 

[Latitude17]

Any progress on your watering system, @Old Salt ?

 

[Old Salt]

Much of it is done. All basic hardware interface routines are complete.

 

[drummer69]

Hey Old Salt
I am planning a system out right now for my grow but having a difficult time finding something with enough I/O on it that I might just go full blown PLC at this point but I'd rather the ardunio route as I have been dying to get my hands on one or 10 LOL . I cant wait my problem is I have no idea how to code lol

 

[drummer69]

I am thinking about 12 dosing pumps for my H2O needs , I will have reg plumbing in my area come spring. tow level soil moisture sensors per plant , air temp , RH , fan control , and humidifier and possibly and H20 misting air ring too we will see
I am on my first ever grow right now and my set up is crude top say the least but all my ladies seem to be doing great so I am going to make it better and better as $$$$$ provides.
Leys not forget auto CO2 and main pumps a various valves LOL
I might be over complicating this but ehh just thinking now while its cold here

 

[DrewT]

@Old Salt I regret I've missed this post until now. How did your light experiment go? I did something similar a while back and had success building a 350W array with 10 independent channels. This was all driven with a ATMega2560 because it has 12PWM outputs to control the LDD drivers.

I think there are enough of us sniffing around this (and other projects) to get it done!

 

[cadwal]

Take a look at the SONOFF modules if you have not already

 

[drummer69]

I started looking at those yesterday after someone mentioned them and almost seems like if I have to tinker with it then why not just make something
Just random thoughts without pot

 

[cadwal]

i am going to build a smart power bar with temp / humidity to the second timer for lights and pumps all controlled with a smart phone. then i will look at complete aeroponics control as well .


i have the parts shipped and will get here next week

 

[DrewT]

@cadwal , Those are pretty nifty devices. There are two reasons I'm not planning to use them and I'd appreciate any feedback you have that could persuade me otherwise:

1. Out of the box these units are cloud dependent. If the internet in China goes offline, so does your grow. This can be overcome with custom firmware (tasmota for example), but it's not a simple off-the-shelf solution.
   
   
2. These modules require users to cut into mains power for many useful applications. I have reservations about suggesting products that may easily result in injury or property damage. Add to that the sonoff is not UL/CE listed and literally melts down around 20A instead of having some sort of thermal protection circuitry.

I'm just not seeing why these warrant inclusion in projects destined to be shared with the community. Perhaps the design will change down the road but as for now I just cant get excited about suggesting them in general.

 

[cadwal]

these are very good points, one version is cloud based but the 433mhz ones can be run on all of the home automation software completely dependent of wifi. with programming changes they have the capability of being stand alone. they are CE certified and with the plug in modules you really do not need to cut mains . and a properly located breaker of fuse in any DYI system would ensure safety. But its your project, your system, and your style of fabrication. everyone has their own thoughts and abilities.