Is Smarthome KNX automation possible based on the floor plan?

  • Erstellt am 2016-08-27 00:02:48

Grym

2016-08-29 11:47:25
  • #1

If all experts say that you should install KNX, at least prepare it, and that you should not rely on proprietary/conventional solutions, and so far no one has complained about their KNX system at home, then at some point I have to realize that my previous opinion might be wrong and I should be more open to the whole thing.

Now a question about star wiring – would you really wire every socket individually from the distribution cabinet or would you at least combine a double or triple socket? Would you then also install correspondingly many fuses, so with 100 sockets 100 fuses, or still about 40-50 fuses when combining double and triple sockets?

And can I then decide afterward in star wiring which sockets are switchable? Just by changes in the utility room? Because another 7 16-channel switching actuators to make around 100 sockets switchable would already be quite expensive.
 

Sebastian79

2016-08-29 11:50:23
  • #2
Because 2-3 people tell you in an anonymous forum that it is important, you let yourself be "convinced" within days/hours?

Whether that is the right way... You have long not yet understood KNX, but are already going into detailed planning.
 

Uwe82

2016-08-29 12:29:13
  • #3
If all experts say that jumping off a bridge is a great sport, would you do it too? If you haven't recognized the added value, you will implement exactly the same with KNX as with a conventional installation, it won't bring you any added value. For example, roller shutter control at will or no constant light control.

The art is usually to find a certain common denominator and then implement it automatically. Whether the slats are 25% or 40% closed doesn't matter at all once you get used to it.

Regarding the wiring, this is how we did it: We thought about how many sockets we want to be switchable. In the bedrooms, it's a maximum of 1 or 2. Then a corresponding cable is brought into the room (NYM5 or NYM7) and daisy-chained from socket to socket. Then you can connect the switchable wires at every socket wherever you need them. Then wire it up in the cabinet (we have terminal blocks there, which makes it easy), parametrize, and done.

Usually nothing changes with the fuses, one fuse per room, blinds are on a separate common circuit. Only for the kitchen we have three triple sockets, each on its own fuse, because the most power-hungry devices are often connected there.
 

Mycraft

2016-08-29 12:53:59
  • #4
A central unit is not really necessary... however, if you want access from outside, complicated logic, a lot of automation... then you need some kind of server whether it is a Raspberry Pi with Smartvisu or an EIBPC or Wiregate, that you have to decide for yourself... every solution, as always, has strengths and weaknesses...

Considering dealing with the components of the system is not a bad idea... however, it is the fifth step before the first... as I already wrote to you, mentally go through the rooms and write down which functions are needed. Not every room light has to be dimmable and not every socket has to be switchable.

For the sockets, you do it as Uwe already said, a set of three or the whole room if it is not too large gets an NYM 5 or 7 and then you can wire the sockets as needed. That way you don’t have to turn your house into a copper mine, even though it often looks like that.

You will most likely not need 2000 meters of bus cable...
Also, the power supply you suggested is oversized and rather for large installations. With 640mA for the inner line and 160mA for the outer line, you have more than enough.



yes, you can then do that by simply rewiring the sockets and in the distribution cabinet as needed, but as I already said... you don’t have to wire every single socket separately. I have, for example, about 100 sockets and 48 actuator channels.
 

Alex85

2016-08-29 17:00:15
  • #5
Similar to Grym, in the meantime I’ve also read a bit more on the topic.

With this subject, you really go from one thing to the next... the wish list keeps growing and growing, you want to do it “right.” So I scrapped everything again and started over small. So much is possible and (still) much of it is just gimmicks, in my opinion, or the intended purpose is questionable.
For example, for the “all off” switch at the entrance to work, I really have to have the entire house covered with KNX in terms of lighting and sockets. That is, of course, a considerable investment. To save how much electricity a year from forgotten lamps? Hmm. In general, the electricity saving arguments seem pretty unrealistic to me.
What stuck with me was the image of the “thousand” switches in the kitchen/dining/living room. For that, you at least need KNX for lighting and blinds in this room. But then you quickly get into window contacts (terrace door open = blind there doesn’t go down; blinds down and terrace door opened = blind at the door goes up). With blinds, you also quickly come to the weather station and then ... ah! It just never ends!! So just leave it at that and live with the one (!!) room with the switch cluster — like 99% of other homebuilders?!

The integration of heating and household appliances also seems uneconomical to me. If the manufacturers have modules for it, they are crazy expensive (500€+). I’m also not so sure what controlling the heating is supposed to achieve. There are examples like “if the window is open, the heating turns down.” What does that do for me in a new build with underfloor heating? Exactly nothing.
I would find the ventilation system on KNX interesting; then you could, for example, install “power buttons” in the toilet or kitchen to boost the output when needed. But even here, the KNX modules are crazy expensive and a ventilation system with separate room control is inherently associated with a surcharge and would make sense for this purpose.
Miele offers washing machines with optional KNX in their top models. I wouldn’t buy those myself, let alone pay triple digits more for the KNX module. My washing machine beeps when it’s done, you can hear it throughout the house. That’s enough. I don’t need an app alert for that.
Homeservers range from cheap and fiddly to nice looking and expensive (2,000€). With those, you get some colorful stuff on a tablet (which I would consider highly optional) and more complex logic becomes possible. Basically, the important functions should probably be kept away from the homeserver because it is prone to failure. That means only comfort goodies like visualization and scene control come on there, which you can easily do without in case of malfunctions.
The absolute limit is then the 1,000€ for the software for parameterization or the crutch via the lite version.
Generally, the prices of the modules, software, and offered services imply that at every corner a hefty license fee is included along with a big surcharge for “luxury.”


also consider whether indeed all sockets have to be switchable. The socket for the Christmas decorations in the window, okay, or the one in the corner for the floor lamp. Maybe one for the coffee machine so it turns on in the morning (although I already find that contrived, that there is really any added value in that). But not throughout the house, where 50% of the sockets are either permanently unused or used sporadically. A current TV uses under 1W in standby. Calculate when the “all off” pays off. Your grandchildren won’t even experience that.
There seem to be real religious wars about presence detectors regarding what is sensible and what isn’t. The fact is, they are not cheap, you at least have to prepare the installation (because pulling cables to the ceiling afterwards is annoying) and — that’s my opinion — those things are very ugly.
 

Sebastian79

2016-08-29 17:14:10
  • #6
Just BTW: I installed "Kacketaster" in the bathrooms, so that pressing them causes the ventilation system to run at higher power. Otherwise, nice post, I see it the same way.
 
Oben