Grow room automation...Whatcha think???

[Guest]

I hesitate to recommend hacking a cheap fan speed control. Often the cheap ones are basically a modified lamp dimmer, and the circuit isn't isolated from line voltage. Here are some examples. http://home.earthlink.net/~jimlux/hv/dimmer.htm
I ran across this. Schematic attached. http://www.elecdesign.com/Articles/Index.cfm?AD=1&ArticleID=6256

It's based on a PWM regulator, and limits the duty cycle to 50% which helps avoid issues when trying to run a fan too slowly. Looks like the unit cost is under $2




However, that will only work for DC fans. Not completely satisfactory. I'm going to have to finish later, as I'm getting ready for work.

As far as smoke alarms, I've been using 4 wire detectors, which have a set of relay contacts that can be used. It's possible to tie onto the alarm of the cheaper units, but it might be necessary to use a RC filter and an analog input as the signal is probably pulsed.

 

[Guest]

The TDA1085C looks like a good candidate for a relatively easy to use, inexpensive motor control, but for "universal motors", shaded pole and one other type which slips my mind. AC induction motor controls are a whole other animal. Variable frequency drives are about the only way to control those. They're kind of like an DC regulator w/ a variable frequency inverter after them. Methodology is relatively complicated and the components are expensive. I'd found a site w/ a "cookbook" schematic using the TDA1085C, and it wasn't bad at all.

A microcontroller does increase the complexity, but at minimal costs. In some ways the complexity is reduced, because the device only performs one function. After a while, the control computer is likely to have a large amount of control algorithms which can interact. I think most people are going to want to implement a function. i.e. a pH controller, temperature control, EC control, or a light control. That gives them the ability to start w/ just what they need. Some will want to implement the functions w/ a central control PC. Not a big deal. We can do both. There are good arguments for doing either, or both.

I'm not suggesting that the 1-wire devices shouldn't be used. If we can get this going as an open source project, I say let everyone generate or request anything they want. There are going to be some devices or functions that you or I may think aren't needed or trivial, but may be of high importance in specific applications. There are also some x10 devices I use, simply because they're readily available. Although I don't like to use them in critical areas, you can get away w/ it w/ monitoring.

As far as the fan speed control, if continuous speed control is needed w/ induction motors a VFD is about the only way to go. If discreet speeds will work, it's possible to use two or more cheap manual controls, adjust them to the desired speed and switch in the one that's set to the desired speed. If a VFD is required, I recommend buying a commercial unit. A surplus, or used VFD can be bought for a couple of hundred dollars, plus or minus. It's very tricky designing a reliable VFD.

 

[Guest]

Minds_I, I couldn't quite read the labels on the diagram. Could you post another? W/ the level of integration of todays ICs, and the easy availability of data sheets via the web it's not too difficult to come up w/ functional circuits. A lot of this stuff is simple enough so that the application notes will often have a circuit diagrammed that will work w/ only small modifications. IMO software design is harder, just easier to change if/when an error is made.

 

[Guest]

Minds_I, I believe I'm out of here. You should have plenty to work w/ if the project picks up again. I'll wait a week or so before deleting my account, just in case. I would have sent you a PM, but apparently the rules have changed. Good luck.

 

[BakedBeans]

I'm still hanging by. I'd like to mess with the code when I get a chance, though I have no time/space anymore for hardware...

I just have to wrap up this project....

 

[Ono Nadagin]

wow guys... have not read everything posted since my last visit to this thread, but it looks like y'all have been busy.

Ill stop back in a day or so after I have gotten a chance to catch up

 

[minds_I]

Hello all,

Sygh, I am sorry I have not put much attention into this thread as of late.

The truth of the matter is that I have none of the necessary equipment to construct the necessary sensors. However, the 1-wire devices in my above diagram (yes, I know it is pixelated) appear to be designed to be quite plug & play.

The interface between the 1-wire system is, from my point of view, a cake-wake.

What needs to be considered in the communication aspects between the software and the 1-wire interface.

You have mentioned many open source software packages that will facillitate the communication protocols and from what I understand provide a "front end" for user friendliness.

So, heres my situation...I have not the equipment to begin construction. In addition, my needs do not include the hydroponic aspects that the system would be targeted for.

I do not have the sofeware necessay and again I have no equipment with which to communicate anyways.

I have been looking into obtaining said equipment ( O.scope, v/o meters, power supplies, parts....) for another project using the 1-wire system for X-mas light shows on the house this year. But that is a wish list at the moment since I have other priorities.

The beauty of the 1-wire system is that each chip does the signal conversion and the communications. So, the system componant design is where attention needs to be focused... the output of the particular sensor must be preconditioned to fall within the 1-wire chips input range. Again, this seems fairly straight forward.

The software design can be made as complex as you wish but an old saying from back in the days when we would program in FORTRAN using punch cards (yes, i am that old...barely) is "garbage in - garbage out".

Also, from this side of the monitor, simplicity is key else for all the systems fuctionality will be of little use.

Here are some blowups of some of the system componant concepts. I hope these will be better.

Anyway, I am still here. Sygh, my apologies for not getting back to you sooner.

minds_I


The humidt sensor componants are listed in pdf associated with the chip (already done)





The pH probe may need pre-conditioning.





digital output used to switch 120volts


PS: I am also tinkering with a pH dosing machine (parastaltic) that can be garage built using off the shelf parts.

 

[Guest]

Looks decent. You should be able to eliminate the opto isolators w/ the addressable switches and solid state relays. IIRC a lot of the SSRs could be switched w/ 4 mA, and they normally have an internal opto isolator. Ebay often has some good deals on used equipment, although I haven't found a 'scope to be too useful in the last few years, there are some good prices on older units. An old AT computer power supply makes a good cheap supply. Also there are quite a few web pages describing the construction of a small hot air soldering iron. Here's one.http://www.engadget.com/2006/03/07/how-to-make-a-surface-mount-soldering-iron/

 

[green_tea]

update:

Just picke up a phiget 8/8/8 kit...

working on the design of the console app to poll the inputs and have a namedpipe server running to accept messages from the main app about which outputs to turn on/off etc..

ideally, the goal is to:

poll all the sensors every 5-10 seconds
store the data polled in a SQL database.

the main app grabs the data from the SQL database and will be where all teh good stuff happens (and also the hardest to code)

anybody know where I can find some real cheap perilistic pumps / solenoid valves?
(if solenoid valves, preferably ones that are made for gravity feeds, and that are somewhat calibrated or have a small ID)




the BIG picture:

monitors:
- water temp
- ambient temp
- any other temp you want to monitor (intake for air cooled hoods, exhaust for hoods etc)
- Co2 (found a cheap analog co2 sensor that i can get to work with teh phiget)
- Ph
- EC/TDS/PPM (whichever will be easiest to calibrate with teh phiget)
- hi / low water level sensors in ALL required areas
- some others I can't think of right now


Control:
- pump that pumps the nute solution into the ebb&flow tray
- pump that pumps R/O water from the R/O tank to the nute tank
- solenoid valve taht turns on/off the water hookup to the RO system (IE control thefilling of the RO tank)
- solenoid that is at teh bottom of the nute tank and goes to the drain (automated weekly nute tank draining)
- 8 solenoid gravity fed valves for ph up/down nute micro/grow/flower, extra supplements



Future things to control:
- Co2
- agitator devices for the nute bottles so that they are always shaken
- agitator device for the nute tank (to help mix the nutes / ph up/down etc together for faster, more accurate readings

- plenty more... I don't ahve my sheet i wrote all this down on.

 

[green_tea]

well the data gathering module is complete. along with the named pipe module (so the interface can talk to the phidget if need be, IE turning on / off things such as pumps)

time to interface it with the SQL server, and then work on the daunting task that is the UI

maybe if I ever get the UI to something that looks amazing, I'll post the code.

found some really cheap OEM CO2 sensors that work in the 0-4v range...

 

[green_tea]

here is a quick pic of one of the windows of the UI I am going to be working on.

(still in the design phase for the GUI)

(image above, takes a while to load, since uploading here is disabled currently; at least for me)

 

[green_tea]

no one is interested in this at all?

were talking total control from the computer.

24/7 monitoring of temp/ph/etc

hell I plan on making a security version of this with door/window sensors... where the GUI is a floorplan of your house, with each monitored area colored green if ok, and red if tripped.

imagine being able to get a SMS text message saying "code X" with each value of x being a different problem (1 could be break in at front door; 2 could be we had a power outage, 3 could be a smoke detector alarm; 4 could me movement on a camera you have somewhere, etc etc)


god i cant wait for my graduation this summer so I can dedicate all my free time to coding this.

 

[Stay Puft]

I am interested.
It looks like you may end up short in the I/O department. (Phidget 8/8/8= 8 digital Inputs, 8 analog Inputs and 8 Digital Outputs) Have you mapped all of your I/O?
(IMHO it is best to allocate all I/O up front in the design, else you end up with the infamous "Never Ending Feature Creature" scenario.)
Do you have plans for expanding I/O capabilities?
I wish you were doing a 1-wire interface....Maybe you can once you are done with the Phidgets thing?
Thanks for sharing your work. Nice job, Please keep up the good work!
Regards,
Stay_Puft

 

[green_tea]

each module (think tigers cab) would require 2 phidgets each.

(at least with the way my setup ideally would work)

I have a list here somewhere...


well when i can find it again, ill post it... (i'm moving soon, so i haven't really had the urge to clean my room up, thus the disorganization)

anyway, phidgets are cheap in the long run, and really easy to put together IMO.


Ive been working on visio diagrams for this, so Ill post some of the flow charts / Activity diagrams as I go along.

Oh, and I'll post a list of everything it will eventually monitor / control.

 

[minds_I]

Hello all,

GT, Staypuft has a point with the 1-wire devices. I too looked at the phigets and other DAU units. One-wire has much more adaptabiity and connectivity as well as a stable interface.

Looking forward to your front-end code.

minds_I

 

[green_tea]

how would a one wire device give you data about say, PH ? are they digital and sending the ph "value" as a data stream?

I see the benefits of making the controller all chips and no interface to the computer, but im trying to be able to use the software for long term data analysis.

Maybe explain the one wire device a bit better for me (im an EX engineer major, not a current one!) haha

 

[minds_I]

Hello all,

Gt, pH sensors use a difference in voltage to measure pH changes. The key is using an off the shelf pH probe.

Look up a few posts and you will see the basic 1-wire schematic for such probes.

One wire devices have have great potential for grow room automation.

Just google 1-wire and read the 1-wire device data sheets...when I first saw them..it was like a kick in the head...Sygh turned me on to them as I was also interested in data acquisition----to much hassle.

minds_I

 

[Stay Puft]

Hi,
Glad to hear from you Minds_I. (Now where is Sygh?)
Sygh was the one who turned me on to 1-wire as well.
I believe he was also the person that mentioned Hobby-boards.com. HBs has very reasonable prices for the 1-wire interface Hardware. They sell the 1-wire modules pre-built or in kit form and they also sell the individual components. They provide schematics on their website. (worth checking out even if you "roll your own")
I purchased the bare PCBs/Dallas Chips from HB and did the kitting/stuffing myself. I would definitely suggest getting their pre-built/tested units instead. (I did not save ANY money by kitting it myself, as a matter of fact I think it cost me MORE in the end! ) HBs is a good alternative for those who do not want to(or cannot) make their own circuits.

GT,
The 1-wire system is connected to the PC via Serial (or USB) port. The 1-wire data is sent back to the PC via the interface so you can store the data for analysis. To attach a analog sensor(Like PH meter) to the 1-wire buss you would attach the sensor output to one of the the inputs on a DS2450 (1-wire quad A to D converter).
The easiest way to do this would be to get a HVAC2-R1-A from hobby-boards for $30.50. Note you will have 3 more inputs remaining so you can use for more sensors!

I have tried to use Mr. House for software but i did not get far. (It kept crashing my system)
I just discovered Jhomenet and am going to give that a try this week. It looks like my request from way back has been fulfilled! (Java based software for 1-wire interface) Why didn't you guys say anything? LOL -kidding
I am excited to try it! I will report back. Anyone else try JHOMEnet yet?

Regards,
Stay_Puft

 

[green_tea]

Do they have APIs for the standard programming languages?

IE C++, etc?

I think thats the nice part of phidgets, easy to use APIs are already created and guaranteed to work. no messing around with assembly and such.

though there are apis for the serial port etc... hrmm

 

[Stay Puft]

Here is a good starting point:
www.maxim-ic.com/products/ibutton/software/resources.cfm
Regards,
Stay_Puft