Home Planning Question

Hi All,

I’m a new OH user and am beginning to plan out what sort of devices I want to implement and would love any feedback from the community on their experiences and choices.

I don’t foresee a very complex home automation setup but would love the option to expand. I think right at the beginning I only intend to do temp and light control, with a possible garage door integration in the med/long term.

For the temp control I am leaning towards two Nest Thermostats since I already own two Nest Cameras.

For lighting I’m still somewhat undecided on implementation. I think at first I want to only have the main family room and kitchen can lights controlled. I don’t really have an interest in color options, just plain white light On/Off control and dimming. Following some brief research I feel like to do this one could have dumb bulbs and smart switches OR smart bulbs.

I am leaning towards smart switches, perhaps something along these lines. Any one with experience using this dimmer product?

Any suggestions for garage door integration? What about z-wave as a bridge? Do you think it would be hard to expand the lighting to the whole house if I wanted( Max 12 switches and possibly 5 outlets)?

Thank you in advance for any advice,

Nest has good integration. I’d probably choose Echobee were I to buy a new HVAC controller today, but Nest is well supported.

Zwave is a good choice. It would probably be easier to just replace the whole switch though depending on where you are there may not be a neutral which can be tricky. There are tons and tons of options and devices with Zwave. And because Zwave is a mesh network, the more devices you have the better the network becomes. It supports up to 256 devices on a single controller.

Pretty much any of the Zwave controllers will work. The Aeotech Gen 5 is really popular as is the RazBerry HAT for those running on RPi.

On the garage front, there are some Zwave controllers and sensors. There is a device called Garadget that has a binding. I’ve gone down the DIY route.

My main advice is to start small, and ease into it. There is a lot to learn and it becomes overwhelming if you try to do it all at once.

1 Like

Thanks for the quick reply Rich. Your advice on starting small is exactly my stance. Its been 10+ yrs since I’ve been anywhere near a linux and/or embedded system so implementing OH2 on a RPi3 B+ has sorta been my easy welcome back project recently.

I believe with your feedback I’m getting more confident in choosing the Aeotech Zwave controller and maybe a single Aeotech Nano Dimmer and giving it a shot. I am firmly in the ‘completely hidden’ camp when it comes to home modifications. So the design appeal of having a standard looking wall switch with the hidden zwave dimmer behind the plate intrigues me.

Thanks for the garage suggestion, I really have not looked into those options at all. This is a great zwave off the shelf product!


There are plenty of zwave switches that look just like any other switch.


I even bought one that had a way to replace the button and surround so we could use dark brown instead of white since all the rest of our switches are dark brown.

You are not limited to retrofitting something behind the switch.

You don’t say where you are in the world but if you have water radiators for heating, take a look at evohome. IMHO, far supior as I can control rads individually. Integration with OH is also very good.

Thanks for the suggestions @CDriver I unfortunately do not have water radiated heating in my home. Nor do I have (at the moment) controllable registers.

Yes, ZWave is a good choice. Aeotec and Fibaro make a lot of pretty good devices, but openHAB also supports pretty much every other device on the market. Since you asked for a garage door opener, the Fibaro FGRM-222 rollershutter actuator is very popular among openHAB users, and it can be configured to be a garage door controller, too.

So am I. Then tightly coordinate your smart home ambitions with your remaining home planning and renovations works.
Put your brainpower into wiring considerations. A KNX bus would need to traverse all rooms, and centralized wiring is even better and more future proof. But obviously pretty hard to achieve in a retrofit scenario.
I’ve ZWave-equipped my whole house, so I essentially didn’t have to add wires for command & control.
Still I had to put in or change wiring at times, often leading to a need to redecorate, and at times we ended up redecorating the whole room or order the painter to get that done.
Just recently, I became fed up with ever replacing batteries of my ZWave sensors so I started to put them on 5V mains power, but I had to hide the power supplies and cables somewhere, so I did that right before the decorator got in anyway.


Just to throw a curve ball in, on the off chance that you’re open to the idea of a complete re-wire (or someone else is reading this thread as research before committing to a home control installation)

It’s worth knowing that the Velbus glass panels are fully programmable and contains all the logic required to run anything from the most basic heating system, right up to a complex multi state HVAC.
Each button is assignable to anything else in the Velbus installation.

A quick example would be an “All Off” button, where one button press sets an entire property to a pre-defined state.

More information about the whole Velbus system can be found here - http://www.velbus.eu/products

It has been around since roughly 2007, with more than a million modules now in service around the world.

A useful example wiring diagram can be found here - Wiring example PDF

Now that @cedricboon has created an amazing OpenHab2 binding for Velbus, you can do things like pass “now playing” data to the display of the top of the range Glass Panels.

Yes, I sell the kit, so I have a vested interest and a clear passion for this hardware.

One tiny detail to consider, especially for a re-wire situation is…

How many questions can you find that ask for help fixing a Velbus installation VS other manufacturer’s products ?

It’s the same reason why I would never buy a Ford. (Again)

The spare parts are easily available for a really good reason :wink:

Good luck with your retrofit, If you decide you do want to go for a completely hidden solution, please get in touch.




Just recently, I became fed up with ever replacing batteries of my ZWave sensors so I started to put them on 5V mains power, but I had to hide the power supplies and cables

I must confess that is an issue I quote quite a lot to new clients.

If you’re going as far as running a low voltage power supply to each control point, you might as well add 2 more cores and install a data bus based system.

No I’ve put in decentralized power supplies.
I agree to have a fully centralized wiring for each and every potential load and sensor location is a nice idea.
But in practice, a) your plans will keep changing fast as new technology and ideas keep emerging and change as you learn what’s working working well and what’s rather not (e.g. motion detector coverage or battery lifetime).
And that’s not even where the OP is yet . And you and me and him will keep coming up with more and more automation ideas (and thus wiring locations) once the first parts successfully made it into operation. And b) you can centralize in a greenfield approach (i.e. if you build your home from scratch or buy one and have the time to fully renovate it before you move in) but not really if you need to retrofit your home while you live there, in your spare time.


You are correct, in every aspect.

Greenfield projects are always easier, but so is tapping into a bus cable and adding new features :smile:

As far as retro fitting is concerned, then Yes, wireless is the logical approach.

Once you exclude the future battery &
sleep state issues :wink:

In the defence of a wired system, if redecorating is part of the plan…

How hard it is to run a single channel down a wall (with an angle grinder or pearls chaser) just deep enough for a tiny bit of data cable?

I’ve done it in my own property (just to prove to myself that it is possible), whilst we lived in the property.
It wasn’t the easiest thing I’ve ever done, but I would do it again if needed.

Also, who said that a wired system had to be centralised?

Velbus can be split into as many nodes as your project requires.

I’ve got 5 in my own home, as an example.

Some of my clients only have 1 node, while some have a node per zone (IE hotel rooms or apartments)

As you rightly say…

I’m sure between us all we can think up a thousand solutions, and a thousand more tomorrow.

My suggestion is just that, a suggestion :smile:



One pitfall thing to be aware of is N-way switches/wiring. (usually in stairwells or large rooms where switches on either end/side of the stair/room control a common fixture).

There are multiple legal ways for a construction electrician to have installed such circuits and not all of the legal wiring patterns are easy to automate. You have to know about the specific wiring of each n-way circuit. I have two stairwells with N-way switches. In one of them I was able to replace the existing switches with standard GE zwave N-way packs. The second stairwell is wired in such a way that it is incompatible with a smart switch strategy. BASICALLY, if the fixture is powered directly from the mains and the N-way switches make/interrupt the circuit, you could be in trouble. (as opposed to mains-power arriving through one of the switches and then to the fixture.)

just FYI…

1 Like

I couldn’t agree more. :smile:

Maybe we should be clearer and say that ANYTHING over 50v is a complete NO go for the end user and any work should be referred to a qualified and registered contractor?

It hasn’t to be but that’s the optimum I would be going for if I was to build a new home.
Centralized wiring allows for exchanging the whole smart home technology WITHOUT rewiring if you ever need to, for whatever reason. 10 to 30 years to live in a building without changing its basic infrastructure is a long time and we’ll see technologies and vendors rise and fall in that time.
That’s btw the reason I would never go for Velbus or any other proprietary system and stick with standards such as KNX or ZWave.

Then again, the beauty and power of openHAB is that you can combine and even replace transmission technologies without the need to change your command & control layer at the same time. You can put wires wherever efforts are low and put up a wireless system, mains or even battery powered, where appropriate, or whatever you get for the best price. Just don’t combine too many different technologies, I’d stick with 2 or 3.

What do you call the opposite of a flame war? This is good reading. :smiley:


Indeed, a maximum of 2 or 3 technologies/hardware platforms is a must for a short term integration with openHAB (most of my projects have at least 4, and yes it is time consuming)!
As for the wiring, if we are talking about a new building, a bus system backed up by a power distribution system is more than enough (for the last 35 years things have been starting to settle and it doesn’t seem they will change in the next 50 years). In the case of an existing wiring setup, yes RF is the only way to go!

1 Like

Arrrr, well yes…

I agree with 90% of all that :smile:

From experience with a huge array of customers, both private and commercial, I’d always recommend a staged or zoned installation.
For all the reasons you have stated not to :slight_smile:

Also, for large installations, the saving on larger cables and reduced risks of data errors (or at least tracking down faults).

Just for reference, while KNX is a standard, it’s not an open standard like Velbus.

There are quite a few companies that build complimentary hardware that leverages the Velbus protocol, or write software around it. (I’m not allowed to say any more)

In my professional opinion, while KNX is a well adopted standard, it’s far from a well met standard when you drill down into each device.

Where as a Velbus module is fully compatible (where practical) with the rest of the range.

From an installer’s point of view, I’ve found out that it’s over £1500 to complete the KNX training course, and at the end of the 5 days, the installer is allowed to put the KNX logo on their paperwork.

Whereas, if an installer attends a FREE training session for Velbus, for only a few hours, they will have all the skills & knowledge they need to deploy a system.

If they attend a paid for full day training event, they will walk away with >£500 worth of Velbus kit mounted on a counter top stand, which acts as a development kit and a powerful sales tool.

I can tell you are very comfortable with KNX & Z-WAVE, which I respect.

If you would like to spend some kind with the (Free) VelbusLink configuration software and connect it (remotely) to one of our training kits, please let me know and I’ll give you the connection details.

Wrong, wrong, wrong!
It is an open standard! It is publicly available! The tools required to commission can be built by anyone!
It’s just that KNX Association Developed the tools faster and better than others!
But, AFAIK anyone can build their own tool!

No one requires a training for a knx system commissioning, it’s just a training so people understand how to deal with knx!
It’s simply not true! I’ve commissioned thousands of devices without having any formal training from knx association!

Okay, maybe that’s just the feedback I’m getting from UK installers.

I’d be very interested in getting you hands on with VelbusLink, as you would be able see it from the view of an experienced KNX configurator.

Okay, again, not the feedback I’m getting in the UK.

I don’t doubt that anyone can build to the standard, so it must be the charge to use the KNX logo that is hiking the retail prices ?

One of our cable suppliers charges almost 70% more for their KNX branded cables, compared to a comparable EIB spec. (Power pair and twisted data pair)

Just for information sake, Velbus modules can work on the same cable types as Mode, Lutron, KNX, AMX etc, with the added bonus that the overall cable length can be in excess of 2KM.

Due to the higher voltage and lower speed (16Kbps) of the data pair.