[ha] Comments on home automation

Adam Davis (adavis.at.baladyne.com)
Tue, 24 Nov 1998 15:07:39 -0500

I just joined to list today, and have gone over the majority of the
messages in the archive.
I'm glad this list is here!

At any rate, I have a few comments regarding my concept of the ha
system:

When I come home, the lights should go on. Assuming I command the
systems to turn on some music(via panel, voice, remote, etc) it will
pipe that music into the room I'm in(or the nearest room equiped with
sound) and adjust it to the level which I last set. If I move to
another room, the lights will go out in the previous, go on in the one
I'm entering, and the music will likewise follow. I want provisions for
audio, video, and data communication between rooms for future
expansion. Of course, security and monitoring would be included. Have
the microphone in the baby's room piped to whatever room I may be in.

With this vision, I see these parts:
1) physical data paths:
Need 2 seperate data paths. One for sensors and small microcontrollers
which would be relatively slow. I don't imagine my sensors needing
packets larger than 5-10 bytes. I would not use X-10, but I don't see
why it can't be used as the slow data path for those who would like to.
The second could implement a faster standard, such as ethernet.
Need audio and video paths. These would be seperate lines for each
room, and terminate in a multiplexing center. This center could route
them to each other, and to cd,radio or other sources(I'm imagining 50 or
100 disc cd players, and a few recievers) as well as video sources, such
as antenna, cable, and DVD changers. This multiplexor would be
connected to the slow data path. Then again, perhaps MP3 would be
suitable to replace the cd-changers. a computer, or SBC with an 8GB HD
would hold my entire collection. It would then be able to support
several users at once, whereas each cd-changer could only support one
listener per changer. hmmm...

2) protocol:
There are already millions of protocols out there. I imagine that we
should decide what the system will do, and what data will be needed and
transported before we can start talking about a protocol for the slow
data path. The fast, I imagine, would do well with TCP/IP.
The system I would like to have would be handled with small uC, such as
the PIC or AVR, with a computer taking notes and managing the larger
items(such as adjusting the vents so that each room can have its own
temp setting) Thus the protocol needs to be small. Each controller
could be a master, slave or both. Each controller could require a
confirmation from the recieving party(as security, or user request) or
not(as one just reporting the outside temperature). It would need to
support variable packets to 50 bytes or more. It would need to support
broadcasting to all(turn off lights) and to anyone(temperature or
clock). It needs to support addresses to 10 bits. Anti-collision would
need to be addressed.

3) misc:
I'm not sure how to handle power. Long lines of dc end up being lossy,
and on the other hand having a transformer or other supply for each uC
is expensive. But IMO the benefits are greater if I were to run a
dedicated DC line through the house. I wish someone would make an
isolated AC-DC power supply the size of a 34-pin DIP in a sip package
that would supply 200mA at 5V for a few bucks apiece. Wishes aside,
it's not too much more expensive to route another pair with the cat-5,
coax, and audio cabling, huh? ;-) I suppose I'd simply have to limit
what I can run off it.

Of course, I don't have the money for this. But I would want the system
to be able to support it. It would be fairly easy to implement, given
time & money. (I'll take donations!)

Well, I hope to be able to contribute to this group.

-Adam