Hi there,
Wanting to set up my first iot network, but curious on what framework to go for.
I have several lamps on my property that I would like to turn on from motion detected on several motion sensors placed in different spots.
The will be approx 15 - 20 m between lamps and motion sensors.
There will be approx 30 - 40 m from the place where a wifi router is placed (and where a potential hub will be connected) to the lamps / sensors placed furthest away.
There will be no wired power to the sensors, so they will need to run on battery.
I prioritize privacy and thus had the thought of setting up HA on an RPI.
I have tested that my phone can catch wifi at all places where sensors are to be located, however, from reading lots of threads I understand that wifi devices and battery power is not a good mix of sensors are possibly trigger several times a day.
Devices have to be cheap (isn’t that always a criterion? ) and be able to communicate to European standards if difference in frequencies (eg z-wave).
Any good suggestions for which frameworks / setups to go for?
If it were me, id plan on running direct burial low voltage wires out to each location to carry either 5 or 12vdc. With low voltage DC available, you can now leverage cheap wifi hardware at each location (like a nodemcu or esp32 dev board). It’s also fairly easy and cheap to set up a few outdoor access points and boost the crap out of your signal if needed.
Going with zigbee or zwave is also an option, but the hardware is more expensive, and I imagine you’ll need more repeaters between your hub and the remote devices, vs blanketing your home with a few wifi access points.
You’ll have a lot of losses on the kind of distances OP needs with such low voltages, unless you use really thick cables.
Technically, the best approach is to lay high voltage cable (230/110) and ethernet for control. But practically, that’s not always an option. I myself have been looking into PoE based lighting. Low voltage (48V-52V) but with much lower losses than 12 or 5V, established standard and you implicitly have a data line for control. So PoE++ switch / injector on the house side and PoE splitter plus small MCU board at each light fixture. Of course that still requires you to dig and lay cable.
Z-Wave or Zigbee, as sparky said, if you don’t want to dig.
12GA landscaping wire will handle like 8-10A up to 30 meters easy. These little esps barely draw anything. Like 1A@5V tops is what I see on mine. And in the US at least, a 100’ spool of 12/2 direct burial line is like under $50 on Amazon.
I have a mix of everything and find I get the most flexibility and best performance out of my wifi stuff. As others have said, any of them will work. I’m a hardwire kind of person.
Edit; don’t forget running line level AC is a completely different beast and at least in the US requires permitting in most places. Direct burial low voltage wire is safe and something any home owner can do. Also don’t forget there’s not a lot of options out there when it comes to POE and ESP’s. It’s possible, but it adds cost.
Where is all this talk of buried cables coming from? Unless I missed something the OP never mentioned that the devices were going to be out in the garden…
I must have completely misread the op. I saw lamps and long range and immediately thought of outdoor lamp posts. Yeah if this is indoor then pick your poison: wifi, zwave, or zigbee.
Just to clarify the situation: wifi router / hub is placed inside the house. Motion sensors and lights are places outside in the yard / garden.
It was not a bad idea running low voltage cables to the sensors, but for several reasons it would not be practical to burry cables in the location.
PoE was mentioned, what is that?
Several of you mention Zigbee or Zwave. Do you think that it would be possible due to the distance? The signal has to run 30 - 40m from hub inside through walls and then to the devices.
If I set up a repeater device (can devices used as repeaters be run off battery?) just outside my front door, then it would still be roughly 20m from that repeater to the sensor far most out.
I looked at the IKEA Tradfri, which supposedly run on Zigbee, but they declare a distance of max 10m between devices, and I would need far more than that.
And one additional question - is there an easy way to test if the range is good enough? Maybe buying two zigbee products to see if they respond (but guess I would also need a hub). Any good way to do this without investing too much in a framework that may not work with the range?
I guess the minimum is a hub and one device, still an investment. Next week I can do a range test for you if you like, but it’s not an exact science as our houses are very different.
Hi Dave,
Thanks a lot, that is really kind of you!
If possible that would be really great. I know it is not exact since, but it would give some kind of indication whether it would work.
My house is 10 cm AAC, 7,5 cm rockwool, 11 cm brick as shown in this picture.
FWIW, I did some range testing here at my house last year when trying out my first zwave motion sensor (a Fibaro FGMS001, that sensor has a bultin range testing feature). The limit of signal loss was at about 25m from the Aeotec Z Stick controller (no repeaters). There was a 90cm thick natural granite wall in the signal path, the rest was direct line of sight, the controller stick was about 3m above ground. When I changed the angle somewhat, so that a second granite wall got into the signal path, range would drop to around 15m. I ended up digging a ditch and laying Ethernet cable for better reliability.
Your wall is much lighter and less dense. It should absorb less signal. Also keep in mind that high moisture conditions (fog, rain) tend to increase absorption and signal scattering rates in air and may reduce range significantly. Zwave is a little less prone to that than Wifi or Zigbee, due to the lower frequency.
Hi Alex,
Thanks a lot for the input and sharing your experience.
I think a range of 25m is pretty impressive if first going through a granite wall. You mention that the controller was set 3m above ground. Does this give better signal spread? I was just planning on placing mine beside my wifi router at table level in the room.
My wall is less dense, yes, but the signal has to pass through two additional inside walls of 10 cm thick AAC. I do not know how much this will impact the signal.
Good point about the weather impact. That would need to be taken into consideration for the system reliability.
As I understand Zigbee runs on the same freq as wifi. I don’t know how much of a difference that higher freq will mean in practice compared to Zwave. But, if I have wifi connectivity to the wifi router in all places where I plan on setting up battery powered motion sensors, will that in theory mean that I will also have zigbee connectivity in those places? Or will a zigbee stick connected to an RPI have a lot less signal strength compared to a wifi router? Also I do not know if there is difference in battery powered zigbee devices signal strength? Can one device/brand have a less range than another or does the zigbee standard ensure that it is the same?
It depends on the antenna type and frequency, there are physical reasons why higher antennas have better signal propagation properties (ground plane reflections, 2nd order lobes for dipoles, etc). But this is more relevant for higher power long range propagation (like radio or cell phone towers, for example). For our type of use, the most important thing is to get as little dense or conductive material into the signal path as possible. Higher up antennas will usually make for a more direct line of sight with less stuff in between, like vegetation, parked cars, people. But that depends a lot on your specific setup of course.
Probably not that much, unless there’s lots of metal in the walls (drywall studs or plumbing for example). Some type of thermal insulation have an aluminum foil on one side, those are deadly for RF signals.
Physically, the difference is very significant. Water absorption (water vapor in the air, human bodies, humidity in walls, etc) is about 4 times higher for 2.4GHz than for the frequency used in zwave (around 850-900 MHz, depending on country). There is a reason microwave ovens work well at 2.4GHz This is usually compensated by higher transmit power, different and more error tolerant modulation, more sensitive receivers and more advanced antenna design and signal shaping techniques (like MIMO). My super thick granite walls are real wifi killers. A wifi signal with a transmit power of +19dBm will barely go through one of these walls. But a zwave signal at -5dBm (that’s 260 times less power !) will go through it without issues. Zigbee is probably somewhere in between. While it shares absorption rates with wifi, it is much lower bandwidth. But again, this is super dependent on your specific setup, construction materials, etc. In the end, only testing can tell.
Not necessarily, as tx power will be very different. A good wifi AP will have around 200mW (or more depending on local regulations), while a Zigbee or zwave stick will have around 10mW. A battery powered sensor will usually transmit in the sub-1mW range to conserve battery. Antenna design can also differ. But wifi and Zigbee are not really directly comparable, because of the different modulations, bandwidth and error corrections. So Zigbee will work at longer ranges than wifi at comparable tx power.
That said, you say you have adequate wifi coverage at the sensor locations. Maybe wifi based sensors could be an alternative.
If you decide to go with Zigbee or zwave, is there a possibility to plug in a repeater node somewhere ? Maybe in a dry compartment or along a wall ?
Regarding the placement, I was planning on running a RPI with a Zigbee USB dongle (if Zigbee will be useable in this case). I guess it would be possible to run a long USB extension cable and put the dongle up high, but would that make any sense? Or do you think a dongle will be too weak a signal compared to a dedicated hub?
I have also seen someone mention a USB stick that supports both Zigbee and Zwave, wonder if that would be a good way to future-proof the system?
I do not thing that wifi will be an option because I will need some of the sensors to be run on battery. From what I can read, wifi will kill the battery in a very short time (also I’m a bit paranoid about wifi device privacy issues). On the contrary, I have read that Wifi 6 is coming, but I do not know if that new standard will make battery wifi devices a realistic alternative or if there are yet any devices available to suit it. I guess you will also need a dedicated wifi router to support it.
Below I have included a picture of how I imagine the system. There are actually quite a few places with wired power that could maybe be used as repeater nodes. Can all zigbee devies act as repeaters or do you need specific devices for that? It is easy to configure a device as a repeater, is it done in HA or on the device?
Is it possible to buy a zigbee device that has much lower range than another (so that you have to be careful what to buy)? If yes, what to look for to tell ‘good’ from ‘bad’ devices?
By the way, about the picture and placement of the RPI that will act as the Zigbee hub - does it actually have to be close to the existing wifi router? I suppose that the RPI will not be dependent on the existing wifi router in any way?