DIY Grow Control: Fans, Lights, Hydro Pumps, Temp and Humidity

subscribed

randomseed said:

So your saying you could run
Master -> flower room slave board
Master -> veg room slave board

and then all the sensors for each room to that rooms control board?
I like it, big issue I have with my growtronix system is that everything has to be serial so I have cables doubling back to where they came from in a few places to get to the outlaying power controls.

Click to expand...

There is no need to doubleback this system can be configured as a star topography or daisy chained, you wouldn't even technically need a master and 2 slaves you could run one in each room and decide which you'd like to use as a master.

As far as the Xbee thing goes I have made some changes to the board allowing you to output to any serial device bluetooth, wifi or xbee. I've also updated the board to make it cheaper to produce, I replaced the "Arduino Mini Pro" with a 28pin Dip model and the necessary compenets reducing production cost significantly. at this point the expensive parts will be the temp/humidity sensors and the casing which I would like to have 3d printed.


I've updated the design a bit for th

I like how your going to design cases, that is one of my major hiccups and just chose to COTS most of them. As for topology while my beta testers are star with dumb units(relays, sensors), I decided to shift most of the units into "smart" units. This way you could operate in a distributed mode and while most major operations would fall under a central controller(i.e timing and data gathering) you could in theory have everything operate independantly. This makes it pretty easy to send data over 415mhz,6mile xbee, ethernet, wifi basically however you find it best to do.

being able to squeeze an atmega onto a board less then 2sq inches and still have room for instrumentation and able to fairly easily assemble is one hurdle that keeps cropping up, but I think I have some of that sorted where possible. I have found most of the smaller SMD components to be cheaper, but you trade off with ease of assembly. Between small boards, good SMD component choices and small to medium bulk orders tend to keep prices in check without dumping thousands of dollars. But then you find yourself playing that lets shrink it untill its an IC game (which I have considered, pH and eC on a single SOIC 8 chip, how awesome would that be) hehe.

Kdn said:

I like how your going to design cases, that is one of my major hiccups and just chose to COTS most of them. As for topology while my beta testers are star with dumb units(relays, sensors), I decided to shift most of the units into "smart" units. This way you could operate in a distributed mode and while most major operations would fall under a central controller(i.e timing and data gathering) you could in theory have everything operate independantly. This makes it pretty easy to send data over 415mhz,6mile xbee, ethernet, wifi basically however you find it best to do.

being able to squeeze an atmega onto a board less then 2sq inches and still have room for instrumentation and able to fairly easily assemble is one hurdle that keeps cropping up, but I think I have some of that sorted where possible. I have found most of the smaller SMD components to be cheaper, but you trade off with ease of assembly. Between small boards, good SMD component choices and small to medium bulk orders tend to keep prices in check without dumping thousands of dollars. But then you find yourself playing that lets shrink it untill its an IC game (which I have considered, pH and eC on a single SOIC 8 chip, how awesome would that be) hehe.

Click to expand...

That is the beauty of the hardware design I'm going with its completely customizable based on individual needs! I would like to COTS the boxes too but thats just not going to happen as I want it to look professional and me cutting up plastic boxes with a dremel just aint worth it even though I think it will add to the cost significantly (I may come up with a COTS and custom solution as the first version will most def not be in the enclosure I want


now I just have to get some scratch together to get the boards made... BTW for those PCB designers interested, the cheap PCB manufacturer I found was ( iteadstudio.com ) The Water pump and lighting relay control systems have been working flawlessly for over a month now so I'm pretty happy with the design and stability of the system at this point but honestly that's the easy part.

Hey everyone thought I'd check in, I'm working up a Bill of materials for a run of 10 boards (I'm hoping to get funding to do them by christmas but unfortunately I'm currently unemployed and do little side work only infrequently at this point). While going over things I ran across this website www.polycase.com they do custom enclosures any amount... I'm going to get together an enclosure design and see what they quote, and I have yet to look through their catalog which they say they can customize ANYTHING with them so this may be a good option, as the 3d printing was looking at being the most expensive part of the process this could reduce that cost significantly. Even if it were I still think the cost would be worth it but I'd also consider selling bare boards no case to keep costs down to customers. but as I mentioned before I feel the case is necessary because of the possible exposure to moisture though I'm not going for waterproof (too many connections for that and it would be impractical).

Watching this one you are doing exactly what i wish i could do, hope to see how much i would have to pay!

poppagetsbaked said:

Hey everyone thought I'd check in, I'm working up a Bill of materials for a run of 10 boards (I'm hoping to get funding to do them by christmas but unfortunately I'm currently unemployed and do little side work only infrequently at this point). While going over things I ran across this website www.polycase.com they do custom enclosures any amount... I'm going to get together an enclosure design and see what they quote, and I have yet to look through their catalog which they say they can customize ANYTHING with them so this may be a good option, as the 3d printing was looking at being the most expensive part of the process this could reduce that cost significantly. Even if it were I still think the cost would be worth it but I'd also consider selling bare boards no case to keep costs down to customers. but as I mentioned before I feel the case is necessary because of the possible exposure to moisture though I'm not going for waterproof (too many connections for that and it would be impractical).

Click to expand...

hehe, Poly is who I get some of my COTS cases from hammer turned me onto them a while back. They have some cool stuff that works pretty well vanilla. But theyre custom prices can be expensive, better to find a local shop to machine them for you. I posted a pic with a controller in a poly case it works really well for its type, it has enough room to house a wallwart a couple relays and "base" station logic(arduino, wifi, sensors, gps/rtc, etc...). their outletbox cases are perfect for recycle timers, with a good amount of design you can fit an arduino, xbee and 2 relay with outlets in one.

trooper, its pretty cheap to get started depending on what you want to do, how many spare parts, etc... Lets say you are a beginner and just want to control your pumps, lights and a fan. You really only need an arduino uno(or clone) and a few relay boards(or a single quad one for example). the total should only be around 50-60 bucks if that and you have a very good starting point to learn from and expand off of. The cost comes in the time it takes to learn about everything, but its not really that hard and once you start its hard to stop

I don't know if you have seen this, but this is where I'm headed with mine, I especially like the humidty of soil check using the resistance... http://www.pc-control.co.uk/greenhouse.htm

Nice will take a look... I finally think I have it all together so here is a parts list:

iteadstudio.com
Custom Designed PCB

PRT-00132
RJ11 6-Pin Connector

COM-00107
Voltage Regulator - 5V

COM-08589
Diode Rectifier - 1A 50V

COM-00536
Crystal 16MHz

COM-00097
Mini Push Button Switch

DEV-10524
ATmega328 with Arduino Optiboot (Uno)

COM-08357
Resettable Fuse PTC

COM-00096
Electrolytic Decoupling Capacitors - 100uF/25V

PRT-07942
DIP Sockets Solder Tail - 28-Pin 0.3"

PRT-00119
DC Barrel Power Jack/Connector

COM-08374
Resistor 10k Ohm 1/6th Watt PTH
Options:

XBee Medium Range Wireless Communication:
1 x XBee Explorer Regulated
2 x XBee 2mW Chip Antenna - Series 2 (ZB)
1 x XBee Explorer USB
1 x 3.5" Cable, 6 Conductors

XBee Long Range Wireless Communication:
1 x XBee Explorer Regulated
2 x XBee Pro 60mW Wire Antenna - Series 1
1 x XBee Explorer USB
1 x 3.5" Cable, 6 Conductors

Bluetooth Wireless Communication:
1 x Bluetooth Modem - BlueSMiRF Silver
1 x Bluetooth USB Module Mini (Optional)
1 x 3.5" Cable, 6 Conductors

WiFi Wireless Communication:
1 x GainSpan WiFi Breakout
1 x 3.5" Cable, 6 Conductors

Solid State Relay (Add-on):
Not Recommended to Exceed 7A
1 x Solid State Relay Kit
1 x GFCI Outlet
1 x GFCI Outlet Shroud
1 x GFCI 3 Prong Replacement Appliance Cord

Temp + Humidity Sensor (Add-on):
Not Recommended for wet areas
1 x Humidity and Temperature Sensor - SHT15 Breakout

Cabling is 6 wire Cat 3 RJ11

Of course there is a lot of soldering etc in there too

attached is a zip with the Eagle files...

View attachment PGBGrowControlv1.zip

I know alot of the parts can be had cheaper elsewhere but I'm a big fan of SFE and if I were going to go into production I'd certainly find cheaper suppliers but for a few boards I'm not worried about it... Glad you like polycase... I'm waiting to hear back from them I wasn't expecting a cheap solution I have some machine shops in my area but they are generally expensive as well... anyways enjoy!

You could make bank off this. The cheapest alternative is like $300+

have you done any research for a cheap c02 ppm transmitter? This would be nice to integrate as well

trooper55 said:

Watching this one you are doing exactly what i wish i could do, hope to see how much i would have to pay!

Click to expand...

another option if you have more time than money is the picaxe. The boards are not too bad in price but I just use the chips with 2 resistors and a serial port to program. (of course you would still need the relays etc..)The arduino is good and if I had the money I would get one because of the support and the community of arduino and there are libraries for everything. The picaxe does have a support fourm too.

I posted before I read the last post what you are doing is an order of magnitude greater in professionalism than what I am doing. Great Thread. I scavenge most of my parts and use 12v car relays with transistors driving them and leds to suppress back emf so to see this is amazing.

Not bad at all poppa, looking good. Just missing a few high speed filtercaps(some .1uF close to the IC) but other than that it looks like it’ll workfine, and easy to build to boot! On your next revision make sure you make a groundplane poor (use the polygon tool, and type GND in the naming bar and hit enter,then build a polygon around your board) not necessary but makes a few thingeasier on the routing end and helps clean up noise. I practically live atsparkfun hehe, they are cool dudes and are quite helpful. I most definitely wouldn’tbe where I am without them and also a few other resources/areas that willremain unnamed

blazeallday9898 said:

You could make bank off this. The cheapestalternative is like $300+

Click to expand...

blazeallday9898 said:

have you done any research for a cheap c02 ppm transmitter? This would benice to integrate as well

Click to expand...

I will have a c02 sensor shortly, I just found a supplier and have startedmaking a board for it. It has a few issues but nothing too serious to dealwith, in fact it’s a lot like interfacing a pH probe.

In the mean time I will leave some pics of my latest design, I finally gotthe boards in and built one of each today. Just got done testing them andreading pH over bluetooth and to think just the pH boardalone could run a small room. The smart relay board is going to be great, 2onboard relays(rated themselves to 20 amps at 120v I wouldn’t do more than 10really) and most of the pins are broken out for expandability. Now to order abunch of parts to build a few of them out hehe.

So overkill for what I need and I still want a couple.

wow... why do I feel like I should go back to school before starting to read this thread lol

wo wo wo... those relay boards from sparkfun are not meant to actually control that much power. You will start a fire. The traces on those boards is way too small to handle even 3a. they are fine for a trigger chord for a bigger relay box (like MLC or titan Helios), but not to put a 5a dehumidifier on. If you want to use those relays, put them directly on the wire, not on small traces on a dc board. I use some I was able to pick up for cheap (crydom d2425) straight out of arduino (internal opto-isolation). they will flip with 3.3 too, and handle 25a. So running my 5a dehumidifier (with a 12a peak when compressor kicks in) doesn't even need a heatsink in the junction box the relay is in.

my system is using ds18B20 Dallas OneWire temp sensors , DHT-22 humidity sensors, and 5 25a relays. Everything is triggered from my control software I have written in c++ using boost libs for serial. so arduino sits there listening on serial and when it gets a line, it parses the command and switches to what it needs to do.

I am waiting to order my xBee parts due to the flooding in Tiawan a while back. the costs have gone up over twice what they were in Augest, so hopefully xBee modules will become cheaper soon.

Would love to join forces, as this has become my main project the past month. I havent gotten on these forums for quite some time, and it would be good to get involved in the community again instead of sitting in my basement 14hrs a day making things alone for myself..

I also have a bunch of scripts to go with motion, for multicam security motion detection, for starting reggea music 5 seconds after the door is open, also upload all still images and video to offsite sftp server as they are created.

I am hoping to port all of my project onto a beaglebone, change sensors for i2c (although more expensive, better compatibility), and use 3.3v GPIO to flip relays, also get some network cameras instead of cheap webcams I had lying around..

and ps. sparkfun free day was awesome this week. Im all fired up waiting for new parts..


pps. is your parts list distributer mostly sparkfun... cause you could save a lot of money going with digiKey or mouser as a supplier.

Sounds like a nice system ommp! Def like you idea of the 3.3 system all around will make xbee and more accurate sensor readings easier. Most of my system is 5v but I like to interface MRs, and I use level shifter when needed to step it down to 3.3 or 1.8v. I designed an Xbee breakout that can be used as a level convertor as well (octal bidirectional auto level shifter) for easier interfacing with Arduinos. I should have some board samples in a week, but it may take a while to get a bulk order due to the new year. Its really a shame not more people are doing this, as I see so much money wasted on things like recycle timers, Something a $1.45 IC can handle quite well I keep meaning to put some writeups up but have been quite busy lately.

yeah I have also been meaning to do some documentation on my project. It is easier to just go into the basement and get to work than to create a wordpress site or something to display the work I have been doing.. It is also still very much a work in progress. Today I am working on getting temp and humidity data onto a 2d plot (graph). Thinking about making a web front end and using FLOT (pretty easy), as well as QWT (head games) for desktop program.

I am waiting for my ethernet shield in the mail (sparkfun freeday score). And then I will change the internal logic from serial to ethernet. I know some people get scared and think this is dangerous, but my home network is pretty secure and I am not port forwarding this to the internet, it has to be accessed via ssh tunnel, and there is at least basic authentication on this apache directory.

I will post svn access when I find another power supply for my beagleboard (rev c4). My repository sits on beagleboard, and earlier this week I got some new webcams (I recomend the logitec c525, great 1280x720 security camera) that needed a power supply at the USB bus to get enough juice to each cam... used my beagle power supply to power cams leaving beagle sitting there off.

This reminds me... if using a beagleboard and arduino setup, both take very little electricity -> a UPS (uninterupted power supply) rated for small amounts can keep power to your arduino and server for a long time (small in UPS ~ 200-300 watts... beagle+arduino = ~20watts). This leaves timers running and security system running for about an hour with no power to the grow area.... I dream of getting a used serverRoom UPS that can handle a few lights if the power goes out during a storm.

I have a question about your c02 sensor... what is the cost of the sensor and is it an infrared sensor? I have looked and looked for cost effective c02 sensors and just haven't found anything with the accuracy I am looking for (infrared digital signal instead of resistance).

I am glad to have found a few like minded individuals in the world. Have fun and stay safe

oh and Kdn... blur out the MAC address of your wireless module in photos and such.

Oh good catch ommp, never even thought about that, even with me looking at it like 5 times oh well, Ill reupload it in a min Thanks for the heads up!
I am using the hanwei mg-811 its of the gas reactive type prolly what you have tested but it seems to work rather well in our perfered range. As you have found they are rather a pain to interface and the IRD type that are good for our wide range can be quite pricey and seem to be rare. I'm still playing with this thing and hope to have a little unit with a atmega onboard that will report humidity, temp and co2 however you need(i2c, spi, serial exposed) in a few months. I need to finish the temp breakout/eC and a couple others first as they are further along on their life cycles.