1

I'm noob to Arduino and electronics, but here is a problem from life which I need to solve for myself:

We grow strawberries and early spring freezes can take most of the flowers from the plants. One way to protect this is to have some kind of heating in solar houses, but if you grow outside, your best option is to watering the flowers on the crops when the freezing comes.

I want to deploy at least 3-4 temperature sensors, which are placed at every parts of the land, which will send the data for an Arduino which will turn on the pump with a relay (for this part I also want to use a pre-built relay solution with socket plugs already built, I don't want to mess around with high electricity myself).

For a cable solution (like catKIT system) I need at least 100m cable for the farest part of the land, which can be managable, but later on, scaling the land lots of cables everywhere can be a problem.

So the best solution would be if every temperature sensor is a standalone station, data transmitted by wireless to the main Arduino which is triggering the pump. This way there will be lass cables on the land, less change to cable break by tractor, etc.

Also, with a wireless system, I don't have to deal with signal loss over the cable, just only wireless signal loss. Is it possible to correct this by middle wireless signal transmitter stations? Or even better, all of them connected to each other!?

Like if the farest sensor reading Celsius 0.8, middle sensor reading Celsius 1.2. Middle sensor have a variable to also consider any reading from the farest sensor, which is reading 0.8. If that reading under Celsius 1, than don't bother his own output, send 0.8 to the main Arduino which is triggering the pump.

These station's power input still need to be solved, but maybe solar panels can be an option? However I have a clue that night they will not work, most of them freezing comes at night.

I saw wireless shields for arduino which offering even a km range in ideal conditions. But how degrading when the conditions are not ideal, like a house wall, rain, fog, snow?

Is it possible to feed multiple wireless inputs too one arduino? 3-4-10-20 standalone stations which transmitting temperature reads for the main arduino. If one of them sends lower temperature than +1 Celsius, than the pump will be turned on. Is it possible to solve this?

Thank You for your answers!

  • 1
    Is temperature really that different, in different locations? I'd go with a single sensor, and maybe a larger safety margin of e.g. 1.5 degrees C. Don't make things too complicated. – Gerben Jul 7 '16 at 16:04
  • Probably not, but in altitude can be a difference (0 to 200cm, if we implement vertical towers). Also consider the environment (open vs tree shades). Happened in our garden that the flowers at open field frozen, the ones was in little shade won't. Covering the entire land with sensors just saves it from partial freezing, which can be still relevant. – Lanti Jul 7 '16 at 16:53
  • 1
    Seems like you already know where the coldest spot is. Just place the one sensor there (on ground level). – Gerben Jul 7 '16 at 18:39
2

There's as many ways of solving your problem as there are of skinning a cat (those poor poor cats - always getting skinned... ;) )

As you have already noted there are two basic solutions: wired or wireless.

Wired does have one major advantage that you haven't really touched on (only mentioned in passing) and that is power. With a wired solution you can provide power over those same wires so you don't need to worry about solar panels, batteries, whatever. But wires can break, yes. So they're not ideal. But then what is?

With wireless you then introduce the major problem of power. Solar panels are one possible way of getting power, though they are expensive, and you need to "over spec" them to charge batteries so you continue to have enough power for the times when is little or no sun. Complementing them with wind power is also an option to boost the battery charge when its windy. It all adds up to very expensive sensors though.

A compromise might be to consider both options together. Turn the sensors into small clusters of sensors that are near by with a base station for those sensors that both gathers and stores the power needed for them, and performs the wireless communication. The sensors would then be wired to the base station. Your wires are then shorter, and your power is more centralised.

But what to then use for the wireless itself? Well, there it's really all down to frequency. There's two basic gropings here:

  • High frequency - 2.4GHz - is great at unobstructed communication. That's the frequency your WiFi works on, and as you probably know it's great if there's no walls in the way. Add walls and the signal drops off considerably.
  • Low frequency - 433MHz - Far far better at penetrating walls and things but much less bandwidth.

I have had distances of hundreds of meters using a 2.4GHz nRF24L01+ module with power amplifier and antenna. That's in open air with (pretty much) line of sight. Indoors the range is much much shorter. Put buildings in the way and your hundreds of meters drop down to tens of meters.

So if there are obstructions you're pretty much going to want to use the 433MHz range (there is another frequency, which is different depending on where in the world you are - 868MHz or 915MHz, so we'll ignore that for now).

There's plenty of modules around that run at 433MHz, some better than others. There is, however, a new standard emerging now that is designed specifically for this kind of thing - long range, low bandwidth, just for sensors (as part of the IoT bandwagon) called LoRa (short for Long Range) which aims at distances of 15km using very small amounts of power (tens of mA) which is ideal for solar powered operation. It also can use the 868/915MHz region as well.

The final option that you touched on without knowing it is that of a mesh. This is where all the nodes form a network and each node is capable of relaying data from other nodes to get to their destination. Kind of like how the internet works. There are some mesh libraries available for some radio modules on the Arduino, but I have no idea how good or reliable they are. The best ones would be completely self-healing with auto-discovery - add a node and the whole mesh works out how to route messages to the destination node; remove a node and the mesh works out new routes to pass the data around the hole it's created.

Creating a good mesh stack is not child's play, so you would want to research it properly and find the best modules with the best library support if you wanted to take this route. That's why LoRa is taking off - it's so much simpler to get wide coverage without the need of complex mesh systems.

2

Personally I wouldn't use an Arduino as a wireless sensor controller, I would use an ESP8226, which comes with WIFI interface built in. You can program this through the Arduino IDE in the same language, and its compatible with most Arduino sensors.

The ESP8266 also has the advantage that it can create a WIFI network of its own, because it can act as a WIFI client, a WIFI access point or both at the same time. So if you do have an issue with the range of the signals then you can program your own WIFI repeater, which could also measure the temperature at the same time. They come in many different shapes and sizes, I think you might be better having an external aerial capability, have a look around and see what you can find. They cost around $2 so they should break the bank.

As for powering it you could use solar or wind power to charge a battery. Probably the best sort of battery for this is a Lead Acid battery because they don't object to being trickle charged. Assuming you sensor station runs at 5V, you want at least a 6V battery. You charge a 6V battery at 7.2V, so your solar panel(s) should generate at least 7.2V in average sunshine. You also need to work out how much power you sensor station is going to need, factor in an extra for working in cold weather, etc and then multiply it by the number of hours per day want it to work (24?). This is the total power you need per day, so the output of you panels must be = (Required Power / Hours of Sun). At this point you are best looking for a COTs device to charge the batteries and achieve your needs.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.