Re: X-10 Mister House Motion sensor problems

["Robert Green" <ROBERT_GREEN1963@xxxxxxxxx>]

"Marc F Hult" <MFHult@xxxxxxxxxxxxxxxxxxxxx> wrote in message

<stuff snipped>

> If it's any consolation, the problem of accumulated delays in motion
sensing
> and its application to smart lighting control is not confined to X-10.
Your
> application is simple one, consisting in turning on a single light in
> response to a single motion detector which requires minimal smarts and so
> Dave's simple fix may suffice. ( Another fall-back which may be even more
> dependable may be to use a switch with a built-in sensor and eschew
> centralized HA altogether. )

That's an interesting point and certainly a pathway a number of people have
taken.  I find it makes me uncomfortable to have such a situation because
almost always you want to be able to control the light in a context other
than detected motion.  The X-10 stuff is quick enough if you position the
sensor and transceiver close to each other and close to the load.

Before I did that, the RF or PLC commands might not always get through and
that means either waiting until the PIR circuitry "looks" again (several
seconds) or reaching for the manual switch.  The time that it takes to
react, reach for the manual switch and hit it coincidentally is nearly equal
the time it takes for the second PIR scan to occur.  Often, the user ends up
hitting the switch a millisecond or two after the 2nd PIR scan has turned
the lamp on and, you guessed it, the user actually turns the lamp out.

> Cyberhouse software (and ABIK, Crestron, AMX, Lutron, I think now Homeseer
> and doubtless others. See eg http://www.freepatentsonline.com/6912429.html
;
> http://www.patentstorm.us/class/700/277-Multiple_zones.html} has a much
more
> complex capability what is variously 'motion vectors' or 'lighting paths/
> automated path lighting' which uses multiple motion detectors to control
> multiple lights in a defined sequence and direction. This in turn places
> much higher demands on the sensors, control SW and HW, and lighting
system.

That's an area where X-10 breaks down.  With multiple occupants in motion,
there are bound to be collisions.  Such a situation requires a higher
bandwidth to avoid such problems.  That's probably where I would switch to
hardwiring.  Since I currently use TM751's spread throughout the house with
one near every controlled load and almost all a room away from each other,
it's worked out nicely.

What I need now is a "PIB" (Person(s) in Bed) detector that can reliably
tell me when either one, none or both of us is in bed.  A sleeping/awake
detector would be nice, too!  I want to automate phone muting, security
arming and a bunch of other things predicated on whether we've gone to bed
for the night.  As a corollary I'd need a travel detector that could also
tell me whether a single person in bed or no persons in bed was a normal
condition for that day.  My wife often works the swing shift, so there's no
really fixed bedtimes around here.

As you noted in an earlier message, home automation really has a long way to
go in terms of knowing detailed information about the home environment and
putting that information towards high level decision making.

> CyberHouse system reportedly worked fine with some combinations of
hardwired
> security system for the motion detectors and some hard-wired lighting
> systems.

As you noted, so many of these have interlink issues that come about from
RS232 or other comm delays.  If there's one area in HA where speed is of the
essence, it's in the automation of lights.  My wife really enjoys the
hands-free aspects of PIR switched lights for the laundry room and bathroom,
but ONLY when they work without delay or "futzing."

> To my dismay 6-7 years ago when I first tried to implement this, Napco's
> popular (among the DIY HA crowd) hard-wired security system was not 'fast
> enough' owing to limitations in the speed at which sensed events are
> reported over the RS-232 connection (and/or replicated on external
relays).

Yeah, that!

> The long-contemplated solution is to use a dedicated hard-wired controller
> such as ELK M1 or homebrew PC/PIC/Atmel. In any case, X10 lighting is
useful
> only under very limited circumstances because of bandwidth limitations
> intrinsic to the protocol - a limitation that is separate from signal
> strength and reliability issues that get most of the attention.

Oddly enough, there are two areas where X-10 is able to overcome bandwidth
limitations that still make it very, very useful.  Those are the "ALL LIGHTS
ON" and "ALL UNITS OFF" commands.  You can command 16 units simultaneously
with one keypress.  You can also stack commands which, at least on the maxi
controller and that allows all the commanded lights, even with different
unit codes, to go on and off simultaneously.

I would have preferred they had chosen "ALL LIGHTS OFF" on the Maxi
controllers - it never made much sense to have two not quite parallel
commands next to the very parallel "ON/OFF" and "BRIGHT/DIM" keys.  But I
agree, overall, that bandwidth is the only monster left in my X-10 setup
with the XTB in place, boosting the signal.  I've avoided long macros as a
result, but that's been more limiting than I like.

--
Bobby G.