Re: Do these exist: "Instant on" or very rapid start CFL???

["Robert Green" <robert_green1963@xxxxxxxxx>]

"Smarty" <nobody@xxxxxxxxxx> wrote in message
news:iss4aj$r72$1@xxxxxxxxxxxxxxxx
> On 6/9/2011 11:12 PM, Robert Green wrote:
> > "Art Todesco"<actodesco@xxxxxxxxx>  wrote in message
> > news:isrqp1$o36$2@xxxxxxxxxxxxxxxxxxxx
> > .
> > <stuff snipped>
> >
> >> You know, the funny thing is that I have had both good and bad with
> >> x10.  In my old house, in the southwestern suburbs of Chicago, it
> >> worked, but to a point.  I couldn't have too many modules, or the
> >> signals would get swamped.  I tried using the X10 amplifier/phase
> >> repeater and the results were disastrous.  So I had to go with a
> >> passive phase coupler.  It would continuously send out random x10
> >> commands which were apparently triggered off noise.  But, for the
> >> most part, if I "obeyed the rules"  it worked.  Here in the western
> >> mountains of NC, I use that amplifier/phase coupler, and it works
> >> perfectly.  I am basically using 1 house code and all 16 channels.
> >> I do have a motion detector on another house code.  I do
> >> occasionally see an anomaly, but they are few and far between.
> > I agree.  Some locations everything works fine, others, not so good.  I
had
> > two repeaters, a Leviton and an X-10 model and neither compared to the
XTB
> > repeater because the XTB boosts the signal to 25V whereas most X-10 gear
> > transmits with 5V or less.  Long cable runs, shoplights, UPS's and
switched
> > power supplies all eat into the X-10 signal in various ways.  The XTB
cuts
> > through them all.
> >
> > Jeff's XTB gear is superbly designed, thoroughly tested and flawlessly
> > assembled.  The kits he's built for me look machine soldered.
> >
> > http://jvde.us/xtb_index.htm
> >
> > --
> > Bobby G.
> >
> >
> Bobby,
> I can understand how signal processing and amplification can improve the
> proper  detection / triggering from X-10 controllers, but I can't
> understand how it could help reduce the false triggering from noise such
> as CFL-generated EMI.
>
> If the receivers in each X-10 module have frequent false triggering from
> noise, as mine most certainly do whenever the fluorescent lights are
> turned on, they remain susceptible to false alarms / triggers even if
> the controller signals are amplified.
>
> The receivers still suffer from a poorly designed X-10 signaling code
> design, and rely on millivolt-level amplitudes of the 125 kHz signal to
> threshold their detectors. In this regard, they should remain as
> vulnerable to noise as they were originally, despite the boost in signal
> strength for the controller signals they receive.

If you're unfortunate enough to have really noisy CFL's and other gear, they
probably need to be put behind a filter.  They come in a form that looks
like an appliance manual and can be had on sale from Ebay for $5 each (list
$20).

I've been looking at this problem with other posters for a long, long time
and I believe what you are seeing is not triggering from noise but a) either
a collision that corrupts a valid X-10 signal or the very annoying tendency
of the "local sense" circuitry to turn the module back on because it
interprets the flashing of the CFL bulb (as the trickle current builds up in
the CFL power supply caps) as someone jiggling the local switch in an
attempt to turn on the light.  Art T. referred to how to defeat this
problem, although the cure has its downsides, some serious.

To really understand the X-10 noise problem it takes something more than a
scope.  You need to be able to interpret X-10 commands frame by frame.
There are devices like the Monterey Powerline Analyzer,

http://www.smarthome.com/4814/X10-Powerline-Signal-Analyzer-XPSA/p.aspx

 to see that there are noise corrupted frames from loud noise sources, but
as much as I've tried to prove under laboratory conditions that an X-10 unit
can read "noise" as a properly formatted X-10 command and execute it are
very slim indeed.  The encoding, while primitive in terms of modern
electronics,  makes it very difficult to generate a valid code from noise.
I didn't always believe this, but smarter minds have convinced me over the
years that inadvertent turning on of equipment, which does happen, has a
source other than noise taking the shape of a valid X-10 command.

I've searched for the Holy Grail of the noise-created command it and have
even gone so far as to purchase a Lynx meter that can dissect an X-10 frame
down to the individual bits - the ones compliment level used to create an
X-10 bit that is used to create the X-10 frame.

http://www.smarthome.com/1153C/LynX-10-PLC-interface/p.aspx

I've been just as perplexed as you have been about where the F^CK all this
noise and these spurious turn ons (and offs) were coming from.
Unfortunately a scope can't easily show X-10 signals in the way that the
Lynx, Monterey or XTBM meters can.

http://thisautomatedhouse.com/XTBMReview.html

They dissect each cycle, read the bit, read the power behind each bit, read
the noise level at the time the command transmits, etc.

When I had collisions it was because I had multiple TM751 and RR501's on the
line and they would not always synch.  They could (and did) collide and the
resulting frame could be interpreted as a valid command.  But IIRC, both
Jeff and I did extensive testing with noise sources and they never generated
a single valid X-10 command.  I looked for weeks creating a special test
bench to "listen" to the noise that the Cellet charger (Bruce Robin of CHA
found the tiny Cellet, a legendary imp from X-10 hell - the charger, not
Bruce!) .  Isolated from the rest of the house by triple filtering, I let
the Cellet sing and sing its X10-like noise.  Nada.  Only when I put an X-10
transmitter on that test circuit did I see fragments of legitimate
commands - but still no true "phantom" commands.  When I put a second
transmitter in the circuit I began to see what looked like phantom codes,
created out of noise but were really two signals collided and being
corrupted with noise at the same time.

Noise can interfere with commands but it can't create them.  X-10 expert and
creator of several landmark X-10 devices and software, Dave Houston in CHA
explained it to me several times before I began to understand it.  I'll try
to Google up what he wrote back then about the X-10 encoding methods and the
creation of spurious yet valid commands from noise.

When the X-10 signal is decoded and boosted by the XTB-IIR it put onto both
phases at 25V through the use of enormous capacitors in the repeater that
you just can't fit into wall modulers.  With that sort of signal, EM
interference becomes mostly a non issue.  I'm betting that with a little
detective work, an XTB-IIR and *maybe* some plug-in filters for the really
bad noise sources, your system will work like it was designed to.

There are some insanely noisy devices that transmit noise right at 120KHz
and they have to be eliminated or filtered if they are plugged into an
outlet that's very close to the circuit panel.  I'm guessing (based on what
everyone else who's tried one reports) that you'll see almost all the
problems you've noted drop away when any stragglers go behind filters, which
you can get in bulk on Ebay for $5 each.  Those filters, on your worst
noisemakers, could significantly improve the performance of your other line
carrier devices, too.

X-10 is like any other home enhancement endeavor.  There are tricks, tools
and tips that can make the difference between a hack job and a pro
installation.  XTB's, filters and an X-10 signal meter are now what it takes
to be a player.  It's just like CAT5 and CCTV work needs good crimpers,
tools, testers, different meters, pliers, testers, etc. There's a minimum
ante to get into the big game.  X-10 is *mostly* plug-and-play but changes
to the nature of home electrical equipment have required some adaptation.
I am just incredibly glad I didn't have to ditch $1K plus of X-10 gear and
have to select a new protocol and rewrite all the code that enables security
lighting, motion detection, etc.  Gawd, the amount of work the XTB saved me
is just frikkin' enormous.  The reliability it has returned to the system
makes it now a 99.9 sort of proposition when it was almost 50/50 before the
XTB.  Would a command make it thru?  No one could say for sure.  Now, it
just works.  Every time.

Whenever I buy anything new that plugs into the line, I check it out on my
X-10 meters to see if it's a serious noise emitter.  Usual suspects are
UPS's, cheap plug-in chargers, laser printers and of course, CFL's.  If it's
outrageously bad I take it back because all the noise usually equals "bad or
cheap design."

Second tier offenders are dimmers, shoplites, LCD TV's and a few other
oddball items here and there.  But there has not been anything I couldn't
second source in a quieter model, from UPS's (avoid APS and Belkin) and CFLs
(The N:vision brand from Home Depot are cheap, come in several color temps
and have withstood the test of time.  The GE spirals did not.)

I've got an enormous X-10 installation and being able to extend its life for
a few hundred bucks has turned out to be a very good investment seeing how
slowly other HA protocols are evolving.  I have lots of lines of code in my
automation controller I'd be loathe to rewrite for some other language.  I
have X-10 modules and capacities that are yet to be duplicated by any other
protocol.  And most important, no other protocol comes close to X-10's price
point.

I can't speak for Jeff but I'll bet he'd take back a unit if it didn't work
out for you.  For anyone teetering on the brink of yanking their X-10
because of reliability problems, this is something they owe it to themselves
to try.

Sorry to sound like such a salesman but I've seen poster after poster in
Comp.Home.Automation get one and go "WOW!  This changed everything!"  If
you're X-10 is flaking out, you owe it to yourself to at least give it a
try.  Seeing is believing.   My wife, a saint for putting up with X-10's
devilish nature for 10 years made this remark: "It just works."  Anyone with
a flaky system knows how important that is to spouses who don't necessarily
share our love for gadgets.  Well, mine anyway!

--
Bobby G.