Re: Re: Lighting Control (was Re: Adverts)

[Nigel Orr <Nigel.Orr@xxxxxxx>]

-------------------------- eGroups Sponsor -------------------------~-~>
GET A NEXTCARD VISA, in 30 seconds!  Get rates
of 2.9% Intro or 9.9% Ongoing APR* and no annual fee!
Apply NOW!
http://click.egroups.com/1/7872/9/_/2065/_/967726039/
---------------------------------------------------------------------_->

Would it be better to take these design discussions elsewhere?  I'm aware
that some folk on here are only really interested in off the shelf
equipment, and the minutiae [sp?] of new designs (whether of webservers,
audiovisual distribution or light dimmers) isn't really of interest, but
others spend a lot of their time and effort in trying to 'do something
better'.  ha_design@xxxxxxx is there for use if it's felt appropriate!

At 17:11 30/08/00 +0100, you wrote:
>and am trying to identify best practice. I assume the IR receiver unit
will
>be intelligent enough to understand dim up/down as well as on/off

Can do.  How many dim steps do you think are necessary?  I'm not one for
dimming lights normally, so I'd plan to just have on, off or dim (maybe
with the dim level presettable to a fixed % instead of stuck at 50%).  IIRC
X10 has 8 steps?  In practice, do people really tweak their lights by 10%
or just stick to a couple of fixed levels?  Switchable soft on and off
would be an advantage, it looks nicer and is better for the lamps too (but
it's no good for fluorescents!).

>would certainly be a requirement for me. Could there be a problem with
the
>directional nature of IR, in that if you were the wrong side of the
room you
>may not be able to control the light with your remote.

Possibly.  I'll try it.  I know IR does tend to bounce a bit, but that and
lampshades might be a problem.  Things that look transparent normally often
reflect or absorb IR, and things that are reflective often transmit or
absorb it... I'll build a simple prototype before I do any mains switching
stuff and see how well it works in different rooms.

>  I don't quite
>understand the advantage of this approach over the IR light switches
that
>the likes of Lutron supply (other than maybe the cost)

Do you mean the Telume that letsautomate sell?  The cost is definitely one
thing (though it does seem reasonable for what you get), but they do look
nice.  Personally, I'd prefer to keep existing switches, and add the
control in the ceiling (where it's also electronically more
straightforward, as you have unswitched live, switched live and
neutral).  The actual switches do look very small, though I _do_ like the
LED ladder showing dim level :-)

They also need a separate slave controller for multiway switching, which an
intelligent ceiling rose (or junction box for those without ceiling roses!)
doesn't.

>  and these units are
>possibly more intuitive in that if a single light switch controls four
>lights then the single IR light switch would do the same, with the
ceiling
>rose approach, you would need to know which light to point at.

Sort of.  Each light would need it's own ID, but if you press a button on a
local remote, it will go to other rooms too, so you could have macros for
multiple lights if you wanted.

Pointing a remote at the light you want to control with it seems fairly
intuitive though... instead of at a switch on the wall.

>every detector point) but are the lenses suitable for transmitting IR
>through an could there be conflicts between best direction for
detecting
>movement and best direction for transmiting IR.

PIRs traditionally go on a wall with the outside window on it.  That also
happens to be a good place to receive IR, because you are shielded from
direct sunlight.  It's a slightly less good place to transmit from (if you
had a directional receiver, it would be pointed close to an outside
window), but if the receiver has to have 360 degree vision anyway, it
doesn't matter.  Interference from CFL lights could be a problem, but my
experiments with the TSOP receivers suggest that they cope extremely well
with a CFL at close range and IR from a few metres.

Not sure about transmitting through the PIR, though it might work- IR
controls are around 1um wavelength, which is pretty close to visible light,
so I'd expect they might be blocked by the light filtering which is needed
to make the PIR (10um wavelength) work reliably.  I'd plan to have the
receiver and transmitter sticking through the PIR case.

>  Again if the IR receiver is
>built in to the light switch then this is likely to be near the door
which
>is often the direction that PIR's point in anyway.

True again, maybe the Lutrons are better for most folk.

>Anyway these were my thoughts, I am definitely in favour of IR control
for
>lighting (even with x10 I would be controlling many of my lights from
my
>Pronto) but I am probably more in favour of the IR receiver being built
in
>to the wall switch rather than the ceiling rose (unless I have missed
>something obvious).

Better save up for a Lutron then :-)

BTW, in response to Mark,
>I thought the status of the lights was one of your prime objectives? 
You
>were complaining you couldn't do this with X10.

The feedback issue isn't really important to me, someone else asked about
replacements for X10 which could do feedback, and I replied that the IR
system I was considering building would be able to have that added quite
easily.  From other comments, it looks like it would be necessary for a
reliable IR system though.

I'll go back to my previous hard-wired ideas.  One further thought- it
could still use an IR protocol, so it could connect into the IR bus or
receive local commands from a remote but use hard-wired commands for
reliable control, it's either that or the Dallas Semi microLAN.  Any
thoughts on that idea?

I still reckon the ceiling rose is the best place for the control
electronics, assuming it can be quite small!

Thanks for the suggestions- they've been very useful

Nigel