# Project specifications

- This is an **Arduino** project - This is supposed to be internationally applicable

I am working on a home automation project which might go into smaller production. The project is wireless and the connections would be managed as followed:

• You would have multiple slave nodes, and one master node in the house.
• All of these nodes would be transmitting and receiving data. E.G.
• Receive: command to switch relay, turn on the AC etc.
• Transmit: optocoupler data to see if a light is on, temperature etc.

# Bands

I have 2 band options:
• 2.4 GHz band
• 433 MHz band

## 2.4 GHz

Pros:
• Big data transfer rate(not really that important)
• Cheap transceivers with antennas available here

Cons:

• Small range, solved by making each node a repeater. It bounces the signal if it isn't its, but writes down the packet id so if another node bounces the packet the node wouldn't get stuck in an endless loop)
• A lot of interference

Questions:

• Do any government regulations limit duty cycles for this band?
• Would it be that badly affected with other wireless networks?
• Would the other signals make the signal hard to read(I have never worked with this band before)?
• Would the repeater idea work?

## 433 MHz

Pros:
• Big range
• Good wall penetration
• Little interference

Cons:

• Really freaking expensive transceivers
• If I would use no transceivers, but 1 receiver and 1 transmitter(cheaper), I would need 2 antennae, and it would cost me about the same as the 2.4 GHz transceiver w/ antenna

Questions:

• Would the 1 receiver + 1 transmitter interfere badly with each other?
• Is this band crowded?
• Do any government regulations limit duty cycles for this band?

## Suggestions

Feel free to give me any suggestions!
• I fly RC airplanes and 2.4GHz RC equipment can typically transmit up to 1.5km. Improving the receiver antenna have been known to increase the range to 2.7km. Boosting the signal strength (illegally) have been known to increase the range to 24km. So 2.4GHz isn't really limited by range if your packets are small and you transmit at low baud rates. Dec 18, 2014 at 3:23
• Just to add some options, nordic does both 2.4ghz and sub-1ghz stuff. banggood.com/… is a nifty MCU and wireless comm with chip antenna for 2.4ghz, and banggood.com/… is a good sub 1.0ghz transceiver. Dec 18, 2014 at 3:32
• Have you looked into IEEE 802.15.4 (eg zigbee)? It's specifically made for applications like home automation, and there are open network/OS stacks which could save you a heap of time (contiki, tinyOS etc)
– sapi
Dec 18, 2014 at 6:29
• Zigbee is really expensive. Dec 18, 2014 at 6:52
• I tuned out at your very first bullet point. If it matters that the project has a arduino in it as apposed to a bare microcontroller, then it's very likely not really a electronics problem. We are not the arduino help desk. No, I'm not going to read the rest of the question now. Dec 18, 2014 at 21:44

Putting this in an answer just because it's easier to type here. I did a similar automation project at home like you describe, so I'll share how I felt when the project was finished. I did custom boards with a TI MSP430 @ 900Mhz, just point to point with a MSP430 with ethernet as the access point. I wish I had picked 433Mhz, I expect it's range would have been better for me.

I picked 900Mhz, like you I was looking for something that was not in 2.4Ghz which I figured would be crowded with wifi and cordless phones. I wanted a small antenna though so I chose 900 over 433. Using my little TI SA there were other 900Mhz sources in my neighborhood but it wasn't crowded.

There are definitely power restrictions, and transmission time restrictions in all those bands. TI has a nice app note about regulatory compliance in these ISM bands. Most interesting to note is that if you do frequency hopping you'll be allowed to use more power than if you just pick a channel and stick with it. That's a little more work, and for me it had battery life implications. My nodes sleep and poll to look for commands from my phone. If I used hopping they'd have to wake up and find the AP's frequency first which burns more power.

I think if you look at some of the micro controller solutions you can find pretty cheap 433Mhz transceivers. That was another reason for 900.

Ah for your other questions about interference it depends, certainly it would be easier to design a system where you always had access to your channel and no one else interfered. Knowing you have to deal with interference will cause you to handle things like re-transmissions, acknowledgments that messages have been sent, maybe error checking like CRC, and collision hold off.

Really it's not a bad thing, you will end up making a more robust product. You don't want to send your switch a signal to turn on and have it never get there. Better to expect interference and make your system robust enough to try again.

Repeaters can definitely work, people use them all the time. To me a repeater infers that it will be on all the time so not good for a battery powered EP, but there are other mesh networking strategies that could work too. I'd start with just point to point first to get yourself going, but you can always dive right in.

Just my 2 cents, good luck with the project.

• Thanks for the 2 cents. Now if I could only find such a cheap transceiver XD. Have you got any links to such cheap transcievers. I want them to be $4 max w/ the antenna, and I am not finding any like that. Dec 17, 2014 at 20:54 • Maybe a 433Mhz CC1101 module from China off ebay?$9 a pair plus $2 shipping :) Can't vouch for the support you'll get ebay.com/itm/… Dec 17, 2014 at 21:30 • Thx I'll have a look :) Dec 17, 2014 at 22:09 Do any government regulations limit duty cycles for this band? Yes, depends on modulation and power output, both FCC and ETSI have duty cycle limits in the 2.4GHz band for narrow band signals. Wideband (500KHz and up) have PSD requirements under FCC rules. Would it be that badly affected with other wireless networks? Depends on the congestion levels in your area. all users of the ISM bands are supposed to be good neighbors, that is, not to cause too much interference and be tolerant to others. Would the other signals make the signal hard to read (I have never worked with this band before)? Depends on your transceiver. Some have good sensitivity and selectivity at certain modulation schemes, some do not. Would the repeater idea work? Mesh networks work, but the data bandwidth is halved with each hop. Zigbee is an example of a mesh network. Would the 1 receiver + 1 transmitter interfere badly with each other? Not if you don't do them simultaneously. Is this band crowded? 433MHz in Europe has ~1.7MHz as ISM, it can be crowded in parking lots (car FOBs). Do any government regulations limit duty cycles for this band? Yes, depending on your modulation and power levels. • So the 1 trans 1 receiver would work as long as i dont receive and transmit at the same time Dec 18, 2014 at 6:54 • Yes. Full duplex RF is quite hard to implement. Thanks for the edit. Dec 18, 2014 at 20:01 I had problems with these small 443MHz transmitter and receiver modules in the past. It turned out that Wii-U uses this frequency. I would suggest using the ESP8266 WiFI modules if you don't mind the low throughput. I just started using one of those for a large wall clock I'm building. They are dirt cheap (<$4).

If you're using 2.4GHz, you should remember that microwaves, WiFi, and Bluetooth devices use these frequencies, all of which may cause unwanted interference in a home.

Also, having your data packets hopping can get very cumbersome. It takes very careful work to prevent packet collisions on the next hop and nuclear packets that never end.

There is no reason for a 433MHz transceiver to be expensive, as Silicon Labs' Si4432 is about $3. Newegg has something from China for about$14. (I will not vouch for its quality, though!)