Home automation planning and costs

And what is the big disadvantage of the radio-based systems now? If you look at, for example, Eltako, it looks very clear, you can buy piece by piece what you want at the moment, upgrade one after the other when you have the money, and you are not dependent on any complex star wiring. When I think about how many kilometers of power cables alone accumulate when you have to pull every socket, every light, every roller shutter individually to the main distribution? o.O Where is the catch of the radio systems? Serious question, I would really like to know. For me as a layman, it sounds very simple. Receiver in the flush-mounted box, radio switch (which doesn’t even need batteries anymore) stuck on the wall, done. At least sounds great?

Disadvantages are:

- Often central units or nodes are necessary, which unnecessarily increases the cost of the system
- Ranges sometimes suffer due to other electromagnetic fields (refrigerator, television, DECT, WLAN, microwave)
- Telegrams get "lost," meaning they simply do not arrive, and after x attempts the process is aborted
- Batteryless push buttons are quite loud and unattractive due to the piezo element
- Many radio systems still require batteries or power, so no advantage is seen here
- One is usually tied to one manufacturer (of course there are also cross-manufacturer radio systems, hence "usually")

and so on and so forth...

A couple of kilometers of cable, on the other hand, can be quickly installed in a new building... and a star wiring is much simpler than a conventional one, as distribution boxes and connection points are omitted and thus complicated circuits are not necessary. The cables simply go from A to B and then switching/dimming/driving is done from A to B

It's not just about laying the cable; the price of the cable alone is quite a big deal. One meter of 3x1.5 costs quickly €0.32 even on the large roll, and that adds up with the many meters. When I think about it... 4 sockets in a room (which I think is still few) already means four separate lines into the room. Plus 1 line for the light and 1 for the roller shutter, making 6. 6x 50m is already 300m, which is almost €100 for a single room, just the purchase price of the cable. Not a single switch connected yet and no BUS cable laid (another €0.3/meter).
In addition, a huge main distribution board, deluxe cable mess, and masses of cable ties in the ceilings and walls. That's a bit too much for me...

Radio:
I need a central unit and/or node at KNX anyway, otherwise, I can't use all the advantages, right? Not to mention the programming software for KNX; you want to be able to adjust things sometimes, otherwise, I might as well stick with conventional.

I accept the range, but you can fix that with a repeater for €50. That probably also equalizes the lost connection attempts, which are certainly due to too long distances or interference. But even my partly used cheap Kopp radio system in the current house works perfectly through three floors.

Whether the switch is "loud" is personally not important to me at all. Mine conventional ones are too^^ You could replace them with normal switches/buttons and corresponding transmitter modules, but why? They also need power and cost another €40 each.

Manufacturer dependency is not a problem anymore with the mentioned eltako ->Enocean because (finally) a standardized protocol and several manufacturers across the board.

I just thought that KNX flush-mounted actuators might be an alternative, i.e., basically only replacing the radio link of the initial installation with a BUS cable but still doing the switching of consumers on site. Then only one power supply would have to go into the room, the switches would sit there and save a lot of cabling with expensive power cable. But the flush-mounted actuators in KNX are even more expensive than the radio actuators in EnOcean, if I read that right (100€/pc KNX compared to about 60€/pc EnOcean/Eltako).

I think KNX is too expensive for me. If you want to expand something later, you have to chisel channels and pull cables again.
That's just my amateur educated opinion after a short research.

Well, you can also use 5x1.5 for the sockets and then it only costs €0.65 per meter for every 3-piece combination for the average buyer. And I don’t know what kind of huge house you are planning, but in my house the longest cables are no longer than 20m, so it turns out that a "normal" room with 6 sockets and one connection point for power- and bus-cables costs about €40.

With the KNX central unit you are on the wrong track... exactly that is THE ADVANTAGE of the system. You don’t need one; everything works autonomously after commissioning. With radio you need nodes/centrals/repeaters, which collect the telegrams and distribute them to the correct participants.

Programming software for KNX is free for up to 20 devices, so usable for the average consumer, and with radio you also need programming software, which is always included; with KNX you have to take care of that yourself.

With the repeater you introduce yet another point of failure in your house and what do you do if it fails?

Lost telegrams increase with the number of participants, and with radio you automatically have more participants than with a wired system, because, for example, you need a motor and a window contact at every window, which means 20 participants for 10 windows, and each one transmits repeatedly. Add lighting, etc., and you have a bunch of devices sending signals all over the place, which causes telegrams to be lost. With a wired system you have, for example, 8 windows on one device, which counts as one participant.

The systems are also designed so that after a certain number of failed attempts (usually three), communication is simply stopped because the receiver is unreachable.

Conclusion: As the complexity of the system increases, the failure rate with radio also increases... this can still be observed with mobile networks at concerts or crowds... eventually reception is only sporadic. The same can be applied to a house full of radio participants.
Whether switches feel good is of course subjective, but you have probably never experienced an EnOcean push button, i.e. Eltako... these make a really "crisp" click, which I personally find disturbing.

Flush-mounted actuators with KNX are more expensive for two reasons.

1. They are usually only used for retrofitting, so simply less demand.
2. All the technology that otherwise controls 4/8/16 channels has to fit into a flush-mounted box to switch/dim only 1-2 channels.

Of course the price per channel is significantly higher.

Later, to expand something, you rarely have to lay new cables. You simply use the existing ones and you have to plan ahead and not just build recklessly. A well-planned system needs no additional cables even decades later. Usually it is enough to replace the bus components and thus simply get new functions.

To each their own, EnOcean is simply too limited for me, that’s why I ended up with KNX.

KNX is only expensive at first glance. A few years ago there were comparisons showing that from a certain size on KNX becomes cheaper compared to conventional systems. Here is a not quite up-to-date info on this topic... unfortunately I can’t find anything comparable for radio...

Please don't get me wrong, I don't want to badmouth the system, on the contrary. I want to form a proper opinion for decision-making. So, thank you very much in advance for the detailed information.

You probably exceed the 20 participants quickly, but at least it's a start and an important piece of information for me.

The note about the multiple actuators with KNX was very interesting; although they are not cheap either, an 8-fold actuator for around 400,- is somewhat more pleasant than the single ones.

Did you do all of this yourself? Or what did you have done? And what did you pay afterwards for what scope? Your avatar looks pretty intimidating in terms of effort at first glance...

What speaks against a KNX solution with small sub-distributions per room that house the actuators and thus significantly reduce the cabling effort and amount in the walls? Then you could also expand it gradually, as desire/time/money allows, which would make the whole thing much more attractive for me.

And how do relevant companies view the topic of own work in this area? Do you get approvals for self-wired but provider-planned systems? I have already wired several sub-distributions myself according to instructions and had them approved; is this possibly also feasible here?

Thank you very much for your detailed answers, it helps me a lot!

You don’t have to buy a Gira. 20-gang switching actuator MDT for 360 euros. I don’t think you’ll end up (much) more expensive with KNX. You can also bypass the 20 project limit of the software with multiple projects.