Aleister93
Active Member
There are a couple of great threads on growroom automation on this forum, but most of them seem to me to be high cost and/or require vast technical skill to pull off.
This thread is to present ideas on how to do monitoring/control as cheaply as possible. One good way to achieve this is to use old PCs as the heart of your monitor/control system.
I'm fairly comfortable with PCs. Using a PC to gather data and control devices/processes has the advantage that it facilitates making your information remotely accessible by internet. Plus, I already have a lot of old computers laying around, which thrills my inner Scotsman beyond measure. A few random notes in regard to this follow.
My criteria are: a) cheapness b) cheapness c) ease of implementation d) flexibility e) accessible by internet if at all possible. I welcome all contributions to this thread that are in aid of these goals.
1) Dataq makes some cheap ($25-$50) hobby model data loggers (multi-channel A/D with either serial or USB inputs) these seem to me like they would be highly adaptable to simultaneously measuring temperature, humidity, ph, and EC/TDS. They also have interfaces available in Active X and in the case of the serial model, C++, so development could be done without re-inventing the wheel.
http://www.dataq.com/products/startkit/di194rs.htm
http://www.dataq.com/products/startkit/di148.htm
I used a similar device in an old soil grow I did, to monitor soil moisture. It was the ultimate in ease and cheapness: I attached steel spikes to the device inputs and shoved them in pairs into the soil. As soil moisture decreases, resistance increases, input falls. At the time, Dataq had a much more closed interface, so I didn't do remote monitoring, I just displayed the graphs on a local monitor next to the grow room. (running under Windows 2000)
If you value your labor at zero, you can also build your own A/D boards cheaply. the Gadre book referenced below contains information on how to do this. There was also a good article in Popular Electronics many years back, titled "Build an Eight-channel A/D Converter" by James J. Barbarello. (June 1995, page 45) This project was specifically aimed at temperature monitoring, using a cheap ADC chip, a multiplexer, and a voltage regulator. It was just three chips and a few resistors and capacitors. This project has the disadvantage that it ties up your parallel port, which in this context I was trying to keep free for device control, later. (If I recall correctly, it is cheap to add extra parallel ports to your computer, provided you have a card slot open.)
I know that at least SOME ph probes have built-in amps, so you only have to measure the voltage from them, and convert that voltage to a ph number. There are some notes on the Dataq site about this, but they dealt with an obscure part. I'll try to get more information about ph probes as we go along, so anyone following this thread can use a standard, easily-available (hopefully cheap) part.
I don't know for sure, but I suspect most TDS meters are only measuring the electrical resistance of the water they are attached to. (water with solids dissolved in it conducts better, hence lower resistance) Anyone who knows, please explain if there is any difference between EC and TDS meters, other than the display scale? perhaps the easiest-to-implement input out of the basic four i mentioned, (temp, humidity, ph, and TDS/EC) would be the last.... the probe on at least one of my TDS meters consists of two pieces of metal protruding from a piece of plastic. MAYBE nothing more is needed to measure TDS than to have two steel spikes a fixed distance apart, sticking out of a piece of epoxy, attached to insulated wires, feeding into our cheap dataq box. plus, of course, some way to convert the number the box gives us to either an EC or TDS scale. There might be some electrical hazard in this scheme, since this clearly involves mixing water and electricity. CAVEAT EMPTOR!
I don't know anything about humidity sensors at this time, so that will be addressed later as I learn more.
2) A lot of security can be gotten by use of a simple webcam monitoring program. The house where I kept my grow was burglarized last year, and "Active WebCam" emailed me some lovely photos of my burglar, as well as alerting me within a minute that someone was in the house. There are other programs with similar functionality, allowing you to accomplish a great deal for the price of a cheap web cam, plus software, if you pay for it. If you're very lazy, you could make this double as a temperature monitor by putting a honkin' big thermometer where your webcam can see it.
3) My personal choice for device control would be the parallel port. My reasons for this are simple, it's cheap to work with, and I understand parallel ports well enough to program them, if i have to. I recommend :
"Programming the Parallel Port" by Dhanajay V. Gadre,
"PC Hardware Projects Vol. 2" by James Barbarello, and "PC Intern" which I believe is available online, hint hint. Also, There are plans online for light show controllers meant to be driven from the parallel port, I don't have the references in front of me.
Here is a kit that might be useful:
http://www.electronickits.com/kit/complete/elec/ck1601.htm
the relays are only 10 amp, I would be loathe to run lights on that without checking carefully with an electrical expert. I would personally look for something with 20 amp relays or more per 1000W light, but what do I know? A 400w light nominally draws a little over 3 amps @ 120v , IF I remember correctly, but might use a lot more when firing up?? My general inclination would be to put lights on their own circuit, on a dedicated mechanical timer. But, I would feel fine using a board like this for lighter stuff, pumps, fans, etc. If you're away for long periods of time, I'd get a UPS for the computer you're using as a controller. and for your router/dsl modem/cable modem, if you want to be sure you can check on the computer.
This thread is to present ideas on how to do monitoring/control as cheaply as possible. One good way to achieve this is to use old PCs as the heart of your monitor/control system.
I'm fairly comfortable with PCs. Using a PC to gather data and control devices/processes has the advantage that it facilitates making your information remotely accessible by internet. Plus, I already have a lot of old computers laying around, which thrills my inner Scotsman beyond measure. A few random notes in regard to this follow.
My criteria are: a) cheapness b) cheapness c) ease of implementation d) flexibility e) accessible by internet if at all possible. I welcome all contributions to this thread that are in aid of these goals.
1) Dataq makes some cheap ($25-$50) hobby model data loggers (multi-channel A/D with either serial or USB inputs) these seem to me like they would be highly adaptable to simultaneously measuring temperature, humidity, ph, and EC/TDS. They also have interfaces available in Active X and in the case of the serial model, C++, so development could be done without re-inventing the wheel.
http://www.dataq.com/products/startkit/di194rs.htm
http://www.dataq.com/products/startkit/di148.htm
I used a similar device in an old soil grow I did, to monitor soil moisture. It was the ultimate in ease and cheapness: I attached steel spikes to the device inputs and shoved them in pairs into the soil. As soil moisture decreases, resistance increases, input falls. At the time, Dataq had a much more closed interface, so I didn't do remote monitoring, I just displayed the graphs on a local monitor next to the grow room. (running under Windows 2000)
If you value your labor at zero, you can also build your own A/D boards cheaply. the Gadre book referenced below contains information on how to do this. There was also a good article in Popular Electronics many years back, titled "Build an Eight-channel A/D Converter" by James J. Barbarello. (June 1995, page 45) This project was specifically aimed at temperature monitoring, using a cheap ADC chip, a multiplexer, and a voltage regulator. It was just three chips and a few resistors and capacitors. This project has the disadvantage that it ties up your parallel port, which in this context I was trying to keep free for device control, later. (If I recall correctly, it is cheap to add extra parallel ports to your computer, provided you have a card slot open.)
I know that at least SOME ph probes have built-in amps, so you only have to measure the voltage from them, and convert that voltage to a ph number. There are some notes on the Dataq site about this, but they dealt with an obscure part. I'll try to get more information about ph probes as we go along, so anyone following this thread can use a standard, easily-available (hopefully cheap) part.
I don't know for sure, but I suspect most TDS meters are only measuring the electrical resistance of the water they are attached to. (water with solids dissolved in it conducts better, hence lower resistance) Anyone who knows, please explain if there is any difference between EC and TDS meters, other than the display scale? perhaps the easiest-to-implement input out of the basic four i mentioned, (temp, humidity, ph, and TDS/EC) would be the last.... the probe on at least one of my TDS meters consists of two pieces of metal protruding from a piece of plastic. MAYBE nothing more is needed to measure TDS than to have two steel spikes a fixed distance apart, sticking out of a piece of epoxy, attached to insulated wires, feeding into our cheap dataq box. plus, of course, some way to convert the number the box gives us to either an EC or TDS scale. There might be some electrical hazard in this scheme, since this clearly involves mixing water and electricity. CAVEAT EMPTOR!
I don't know anything about humidity sensors at this time, so that will be addressed later as I learn more.
2) A lot of security can be gotten by use of a simple webcam monitoring program. The house where I kept my grow was burglarized last year, and "Active WebCam" emailed me some lovely photos of my burglar, as well as alerting me within a minute that someone was in the house. There are other programs with similar functionality, allowing you to accomplish a great deal for the price of a cheap web cam, plus software, if you pay for it. If you're very lazy, you could make this double as a temperature monitor by putting a honkin' big thermometer where your webcam can see it.
3) My personal choice for device control would be the parallel port. My reasons for this are simple, it's cheap to work with, and I understand parallel ports well enough to program them, if i have to. I recommend :
"Programming the Parallel Port" by Dhanajay V. Gadre,
"PC Hardware Projects Vol. 2" by James Barbarello, and "PC Intern" which I believe is available online, hint hint. Also, There are plans online for light show controllers meant to be driven from the parallel port, I don't have the references in front of me.
Here is a kit that might be useful:
http://www.electronickits.com/kit/complete/elec/ck1601.htm
the relays are only 10 amp, I would be loathe to run lights on that without checking carefully with an electrical expert. I would personally look for something with 20 amp relays or more per 1000W light, but what do I know? A 400w light nominally draws a little over 3 amps @ 120v , IF I remember correctly, but might use a lot more when firing up?? My general inclination would be to put lights on their own circuit, on a dedicated mechanical timer. But, I would feel fine using a board like this for lighter stuff, pumps, fans, etc. If you're away for long periods of time, I'd get a UPS for the computer you're using as a controller. and for your router/dsl modem/cable modem, if you want to be sure you can check on the computer.
