Re: X-10 RF home automation being overwhelmed by M2 OFF, C2 OFF and Gxx DIM signals

["Robert Green" <robert_green1963@xxxxxxxxx>]

"dlh" <dlh@xxxxxxxxxxxxxxx> wrote in message
news:64cc4445-2647-4426-a9ab-3237543d1e7f@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
> This gets too unwieldy - trying to post inline so I'll try to cover
> everything without any quoted text.

Sorry for the late responses.  I kept checking for new messages in this
group and for some reason, the flags weren't updating.  Since it's been so
long since you wrote this, I feel forced to quote so that my comments make
at least *some* sense.

The latest information I have is that for the 5 days preceding 9-11 the
interference was continuous.  The codes received were either M2 or C2 OFFs,
no Gxx DIMs were received.  Now it's gone.  Completely.  Nothing since the
day after 9-11.  Of course, the weather has turned cool so if it's a signal
to shut off air conditioners enrolled in the anti-brownout program,. it may
just not be transmitted anymore.  I still suspect an anti-IED transmission
of some kind, especially because of the nearly solid coverage for the 5 days
leading up to the anniversary.

Rather than play RF Sherlock Ohms, I punted and took a box of
mini-controllers that I got at closeout for $3 each and began to modify them
so that I could use them for housecodes 9-16.  Why they limited those units
to only the first 8 housecodes is a mystery to me.  I've basically replaced
all the RF transmitters like the "sticka" switches with the
mini-transceivers as an ugly temporary measure but they have EXTREMELY low
SAF.  One benefit is that switch response time is noticeable quicker not
having to be translated from RF to PLC.

> It's unlikely but possible that it's coming in via the powerlines.
> Your reading with Jeff Volp's device is _probably_ reporting the
> translated PLC command although I cannot think of a reason for the
> lower frequency. The transceivers tend to have very wide RF bandwidth
> and PLC receivers do as well, usually reacting to signals in the
> 75-200kHz range. There was something a few years back, which I dubbed
> the "Endless Dim Syndrome" where the valid 120kHz was being radiated
> by the in-wall wiring and inductively coupling to TM751s (mostly) via
> their antenna. It was fairly easy to induce by aligning the antenna
> with the wiring. However, only Dims & Brights were reported by the
> victims. The TM751 was apparently cheating, assuming that any further
> RF meant the remote was still sending and the poor design of the RF
> receiver allowed it to respond to the lower frequency. Recent TM751s
> use the same high-quality RF receiver used by the CM15A but I have no
> idea when that change occured.

At first, I thought I could test the "is it coming through the powerlines"
by using X-10 filter modules, but they only block 120kHz so that wouldn't
work.  I needed to make sure that the filter caught a much higher frequency,
in the 315mHz so I gave up on that.   My limited testing showed that the
interference was strong enough to inhibit proper operation of transceivers
set to housecodes other than the interfering ones, so I was not too sanguine
about doing serious detective work on the assumption that even if I found
the offender, convincing them to turn it off might be futile.  I'd also hate
to have HLS think I was trying to build some sort of IED immune to
interference and was fishing for information about how to do that. (-:
(FWIW, I recently read that in Iraq that terrorists have started to use IR
line of sight IED's to overcome the jamming.

> You can steal the RF receiver daughterboard from a TM751 but I'm not
> sure I recall the connections. The RF receiver daughterboard in the
> CM15A is much better but I'd hate to see you cannibalize a CM15A. The
> 315MHz receiver is the one, should you go that way - turn the tuning
> slug 1/8T CCW and you'll be close enough.

I couldn't bring myself to hack up a perfectly good CM15A.  The only upside
to this whole problem is that the Smarthome Control MaxiLinc all housecode
PLC controllers have an LED that flashes with X-10 traffic.  When the M2/C2
OFF demon strikes, they all start flashing like crazy (in addition to the
X-10 RF and PLC basically stopping working).  I have the WGL on an old RR501
set to an unused housecode.  When the storm starts, I just turn off the RF
completely by hitting the manual control button on the RR501.  If I try long
enough, I can even get an RF command to the unit remotely to shut off all RF
except that one RR501.  A kludge, but a workable one.  My attempts to detect
the "broadcast storm" indicator that Jeff's XTB-IIR emits when there's a
broadcast storm with HomeVision were of limited success because of HV's
inability to detect and react to the P Status Off request.

> You should be able to find a freeware/shareware oscilloscope program
> that uses the soundcard. It's easier than the methods in my article
> which was written long ago.

I'll look for one.  It sounds like a good tool to have in the toolchest.

> The decoded bits don't tell me what I want to know. It may be that the
> codes are actually longer and the X-10 receivers are just grabbing the
> first 32 bits - I never thought to test whether that was possible.
> But, if the codes are longer, we can possibly rule out an X-10 source
> (some X-10 security codes are longer). Of course, X-10's limited range
> probably rules out a nearby source. X-10 uses NEC's IR code which
> dates back 40+ years and has been copied with slight modifications by
> many others. I haven't encountered any others using it for RF but I've
> been occupied elsewhere (several surgeries and other hospitalizations)
> over the past 18 months or so. We might learn something from the raw
> codes. And, if it is a high-power source, it may just be overwhelming
> the receivers as those military tests a few years back did to the
> garage doors which normally only respond to rolling codes (different
> with each transmission).

I took a 100' extension cord and walked it around the house with the WGL and
a TW523 attached.  I couldn't find a location where the signal was too weak
to receive, so it's either riding the wire or it's being broadcast at a much
higher power than normal X-10 RF tranmissions from a location far enough
away so that it wouldn't appear directionally oriented (at least with my
primitive 100' cord setup - a directional antenna attached to the WGL was
one of the next things I was going to try.

> Pepco is unlikely to be helpful unless you reach the right person.
> Duke's regional engineer came to investigate an issue I had a couple
> of years ago and he was quite sharp but the ones I talked to before he
> came by were clueless.

I haven't even called yet.  I was trying to acquire more data and parse the
HV logs for dates and times when the problem occurred.  That got complicated
because there was no data for the long times where I had disconnected the
WGL and no data was being logged.  I would periodically check by turning the
WGL receiver back on, leaving it on if it was no longer getting "commands
from outer space."

> IIRC, some folks on the West Coast were having some X-10 issues with a
> specific smart electric meter. I don't recall details. I think there
> was a discussion (with Jeff Volp's participation) on the X-10
> Community Forums. Duke Energy installed Excellon meters here a couple
> of years back. I've seen no X-10 issues but have had two (different
> brand) inexpensive 700W microwaves lose their minds - beeping randomly
> and flashing or clearing the displays. I've suspected the new smart
> meters might have been outputting something that got by any filters
> the microwaves might have had. I now have a 1000W Haier Grill/
> Convection/Microwave combo that seems immune (so far).

I was going to buy that same unit!  What do you think of it?  The 25 year
old Litton isn't going to last forever.

Thanks for your input, as always, Dave.

--
Bobby G.