Greenhouse automation

Hello

Although I have some experience in programming PCÕs and mainframes, I am completely new to PLC-programming, but very eager to learn how to use them. There is a lot of interesting material on the internet about this, including this board and the book that goes with it. But, after reading much of this material, I think the only way to learn is probably to start with a real world application.

There is a tradition in our family to cultivate our own vegetables, and for that purpose we have build us a little (70 sqm) greenhouse where we grow different vegetables as salad and tomatoes...
To maintain the ideal climate in the house, I would like to do some greenhouse-automation. (Now we must do this manually, but it is one of my tasks in the household and I forget it to many times.) I know such systems exists for professional purposes, but they are far to expensive for our little operation.

In a first phase I would like to control the ventilation system (window vents in the roof of the greenhouse) in relation to the temperature and the relative humidity in the house and the wind-velocity outside of it (When there is excessive wind the greenhouse should be closed, whatever the temperature and the humidity).

To manage this, I was thinking of an OPLC from Unitronics. The presence of the operating panel makes it possible to easily change some parameters and to initiate some instant actions, like starting irrigation for a certain amount of time (what I would like to add in a second phase.) The cost of this PLC is within the limits of what I want to spent on this, and there is reasonable support for it here in Belgium.

Logically I think I need 2 analog-inputs for the temperature- and the humidity-sensors, 1 high-speed counter to count the pulses of an anemometer and 2 digital inputs for 2 switches that can tell the PLC when the limits of the window vents opening and closing are reached. On the output I think there must be 2 relays, for driving an engine in 2 directions, for the opening and closing of the vents.

Till now all seems logical to me. (Maybe to logical?) Even the ladder-program I think I have an idea how it must look like. (To much pride?)
But all that is of course rather theoretical. Now it is essential to know which kind of input sensors I need and how they must be connected to the PLC. And for that I have to little experience and knowledge about the technical aspects of this connections.
The documentation of the OPLC for example says about the analog inputs: Two 10-bit, multi-range inputs: 0 - 10V; 0-20mA, 4-20mA; conversion method: successive approximation; input impedance >100KOhm for voltage, 500KOhm for current....and so on. Is it really necessary to understand all this? Is there no form of standardisation in this connections? (Maybe I make someone laughing now?)

Can anyone give me some advice about this? Or tell me what is the most efficient way to learn more about this?

Kind regards.

Marc

Those are standard anlog input types. The one you use depends on the sending unit you use. As far as temp measurments using a thermocouple you need to convert the ma signal to one of the common analog types. THis is usually done with an amplifier that cleans up the signal and will scale it to some engineering unit. Unfortunaly analog is the most expensive side of gathering data. Do some searching on thermocouples, thermisters, etc to see what is out there for sensors. When you find what you need and the output of the sensor is compatable with the PLC analog inputs you are almost there...

John

Marc,

I can assure you that no one is laughing - you have just done an outstanding job of outlining an excellent project and a perfectly acceptable plan of attack. I’m sure that you’ll get some highly useful responses to your post. Since I have no experience with the brand of PLC that you mentioned, I’ll let others fill you in on how to proceed with programming it.

But to answer your specific question,

Is it really necessary to understand all this?

Click to expand...

I know that many people are going to respond and say that I’m wrong - hopefully they’ll be good enough to explain some of the technical aspects of what you posted. I’m going to keep it short and just say “no” - you don’t have to “understand” all of the information that you posted - in the same way that my little wife doesn’t need to “understand” gasoline octane ratings. Believe me, her ignorance in such matters doesn’t keep her from filling up the tank with unleaded regular. Still, you’ll definitely need to keep those specs handy as you try to find the sensors that you’ll need to connect to your PLC. Basically, you’re going to be looking for instruments which will provide signals which will be compatible with your PLC inputs. By the way, have you looked at any prices for this type of stuff yet? If not, you might be in for a case of “sticker shock”.

What you’re going to run into next is the fact that the instrumentation (sensor) part of your project is likely to cost much more than the PLC controller itself. Along those lines, let me mention a couple of ideas for you to think about. Both of these ideas would be too “unsophisticated” for an industrial application - but for your home greenhouse project, they might save you enough money to make the project economically feasible.

As for the wind speed - are you sure that you need to know the speed of the wind? - or would just knowing that the wind is “too strong” be acceptable? Suppose that you were to hang a wind-chime near the greenhouse - and that immediately below the “vane” (the weight hanging at the bottom) you mounted a proximity sensor. If the wind is strong enough to move the weight away from the sensor, then it’s too strong. Now your “wind-speed” input is just a simple discrete input and costs only as much as a proximity sensor. Of course you’ll need to calibrate this “instrument” by increasing or decreasing the weight of the vane - or making the vane larger or smaller - or making the string longer or shorter - but comparing the cost of this arrangement against the price of an anemometer, the time and trouble taken to experiment might be very well spent.

You said you’re going to need to know the temperature - and the humidity - do a little research on a wet-bulb psychrometer. Basically, it’s just two thermometers mounted side by side - but one has a piece of cotton wicking (think shoe-string) stretched over its sensing bulb. The other end of the wick hangs into a jar of water. As air flows over the bulbs, (mount this next to a fan) the wet bulb cools faster than the dry one - due to the effect of evaporation. The difference in the two temperature readings is related to the relative humidity in the air. A quick web search should find a chart showing the conversion between the temperature readings and the humidity. Plug this wet bulb reading into your PLC and have it calculate (or just use a data look-up table) to react to the humidity. So instead of breaking your budget with an industrial grade humidity sensor, you might be able to get your project working just by adding a second temperature input (remember, you already needed one for the temperature reading) and a wet cotton shoe-string.

What I’ve tried to do here is give you some ideas so that your excellent project can proceed in spite of the high cost of industrial quality instrumentation. Later, as more funds become available, you’ll definitely want to upgrade and improve your system (and learn and understand all of those confusing specifications that you listed). Even so, I think that you’ll enjoy doing things the “Fred Flintstone” way to begin with - and I guarantee that you'll learn plenty in the process.

Marc,

I agree with Ron that no one is laughing.

I also agree with him that you don't really need to "know" all that stuff, but you it is still pretty important. This is the stuff that lets you know whether a device will operate well on you system or possibly burn it out. It also tells you the stuff you need to know to scale your input values and so on.

My advise would be to not worry too much about those details yourself but instead be sure to contact your local support people and then let them check out your choices before you plug things in. This should go along way in preventing the release of any imbedded smoke.

Steve

Marc,

I think you did an excellent job of describing your application. You may want to look at using the Omron Zen smart relay for this application. This is a striped down PLC with a built in operator interface, in fact in the Omron pamphlet they show using the Zen for greenhouse automation. The DC unit comes with two analog inputs, four digital inputs, and 4 relay outputs. You can program the unit either from the built in keypad or with software.

The software even has the ability to simulate your program before you ever wire it up. The Zen unit is around $125 US for the unit I described; it also has a real time clock built in. The software is around $50 US, but I am sure you can acquire it at no cost.

I hope all goes well.

Mike

I wish our student questioners could learn from your example if how to clearly and completely describe your project.

I have a couple of suggestions. First of all, for a novice it is probably easier to work with voltage signals, so you should try to get your sensor outputs to be 0-10 VDC. These will also be less expensive in most cases. With your short wiring lengths voltage should work just fine.

Humidity sensors are often un-reliable, so exercise some care in selection. The sling psychrometer approach previously described is an elegant solution, but as a beginner it may be too complex - a canned unit might cause a lot less frustration.

I suggest you organize your inputs and outputs in a point list - search the site for examples. It is essentially a list of each input and output, the voltage or signal level, a description of the sensor or device attached, the range of the instrument, and maybe the location.

Finally, I have never heard of that brand of PLC. The guys here will be happy to help you with program problems, but the answers may be generic in mnature because many of us won't know the details of your unit.

Pulse divider

Marc:

High-speed counters can give even experienced programmers trouble. An alternative work-around to Ron's is to use a fairly inexpensive pulse divider. This could take the pulses from the anemometer and generate a single pulse for every 100 (for example). This would go into a single discrete input. The PLC would see a pulse in normal scan time, and can then count pulses over a time period (or perhaps even time between pulses)

It's not as "accurate" as a high-speed counter, but for your application, it will be fine.

You ask "Is there no form of standardisation in this connections?" The answer is - there is, and there isn't. 0 to 10V, -10 to -10V, 0 to 20 mA, and 4 to 20 mA are the standard signal types for analog inputs. And while 4 options doesn't seem like much of a "standard", at least it's only 4. And that's why the PLC that you are picking has the ability to do both voltage and current sensing.

Hi All,

Thank you so much for all your help and guidance. This is a great forum!
I need some time to digest this.
Concerning the unit I want to use: all specs are at the site of the manufacturer at http://www.unitronics.com/products.htm. I am thinking of using the V120-12-R2C (or the older M90-R2-CAN). The kind of temperature-sensor I am thinking to use, is something like the ATS2000A TSPOT, which specs are also on the net. I suppose this is already one good combination?

Thank you again.

Marc

Marc,

If you decide to use thermocouple inputs to measure temperature, a "regular" analog input (0-5, 0-10, 4-20 mA) will not suffice. You will require a signal conditioner. The cost ranges anywhere from $100 to $300 US.

Other things to consider for your application:

1) Your output devices and the current loads they require. Most PLCs operate on minimal loads (2 amp relay). If your current requirements are greater than the rating, you will require additional relays to handle the load.

2) Having the operator interface built into the controller is a nice feature but limits you to the I/O that the controller provides. A packaged product does not usually provide the same flexibility as a modular or embedded solution. The controller will be doing all the "grunt work" processing, turning on and off. The operator interface makes it user friendly. You will want to determine your I/O requirements.

If you have already chosen the Unitronics product, then you will need to see what additional components you will require, ie signal conditioner for thermocouple. All these additional devices add to your overall controller cost. If your controller costs $500 but you have to add $400 in additional components, when you can get a controller and operator interface that will do what you want with no additional components for $700, you are actually spending more.

We obtained a new customer that was recently considering Splat and Automation Direct (DL06). Our customer required two RTD inputs. In both instances, he would have required signal conditioners to convert a standard analog signal into an RTD and additional relays for output loads. Eventhough the controller was under $300, he would be spending $300 in additional components to adapt the controller to his application. We were able to provide a controller that would do RTD inputs directly.

3) One last thing to keep in mind. When you are sizing your application, get more I/O than you think you will need. There may be a time (when you are developing or 2 years from now) when you will want/need to add to your solution.

God Bless,

Stephen Luft

Hi Marc....I'm from the 'old school' of down and dirty / cheep, so here is another idea on your temperature controls....use a photo sensor to look thru the glass thermometer, set so that the light is blocked when the temp. rises,; then you can just use it as a DC switch input to your plc. you could do the same with the humidity .....just an idea....David

Marc,

I have seasoned engineers working for me that can't outline an application as well as you have. Good job! It really gives us something to work with.

The M90 units are great for cost and ease of use. It should be fine for your application. If the budget will allow you may want to try to use a pulse to analog converter for your wind speed. Your sensor supplier may offer such a device. You may want to take a look at WWW.omega.com and see if they have any of the sensors that you need. The humidity device I use is an Omega HX92C. It is a loop powered 4-20ma device and has proven to be quite reliable.

Sounds like a fun project.

Good Luck

Lenny

Marc,

We have used several Unitronics products. Very nice product. Powerful, and very easy to use. Free programming software!!

You might consider using the V120-12-UN2 instead of the V120-12-R2C as this has two universal input channels that can be configured for analog input or Thermocouple or RTD. The downside is that this module has transistor outputs for 24vdc instead of relay contact outputs, so you might have to provide interface relays to drive your real world loads (solenoids, contactors, etc.)

Just remember that you 'lose' some of the discrete digital inputs when you start to use the counters and analog/universal input channels i.e. each high speed counter consumes 2 discrete inputs and each analog channel consumes two inputs.

If you are worried about cost, the expansion I/O modules work with both the V120 line and the cheaper M90 product line.

My only gripe about the analog input module is that it is strictly unipolar. We had to resort to using a resistor bias scheme in order to 'read' bipolar signals. Not expensive (4 resistors) but awkward. The analog output modules are bipolar.

Otherwise, it is a very nice product for anybody interested in a 'beginers' PLC.

Hi All,

Don't think I gave up on this project. Since many answers where positive about my choice of the Unitronics OPLC, I decided to ask some advice to Unitronics headquarters. I am still waiting for an answer. (I know they are moving to a new buiding, so this can take a while.)

Kind regards

Marc