Re: Trapped X10 ghost signals

[nobody@xxxxxxxxxxxx (Dave Houston)]

Thierry,

There have been other, similar reports where switchmode power supplies
caused random turn ons of X-10 modules or switches.

With one exception, those reports, like yours, said no valid X-10 commands
were seen by powerline interfaces (e.g. CM11A). In the one case where the
individual claimed there were valid commands on the line, a Google search
will show that he was initially emphatic that there were no such commands on
the line and then later claimed otherwise. I suspect his later descriptions
were in error. Like your experience, only certain addresses were affected.
He said it only happened at night.

X-10 has documented that powerline spikes and brownouts can cause events
like this. Steven Bloom has documented a "fix" for some switches that were
being triggered by spikes.

http://groups-beta.google.com/group/comp.home.automation/msg/62d41d8a36fef5ea?oe=UTF-8&output=gplain

So far nobody has offered a plausible explanation of the events you describe
but the fact that only your X-10 made devices were affected is probably an
important clue. It may have a cause similar to the above.

It might be helpful to know what switching frequency the power supply uses
but I've never seen one that listed this in its specs.

"Thierry Parent" <TParent@xxxxxxxxxx> wrote:

>Just to contribute to the X10 home automation knowledge base, I discovered
>two weeks ago the reason of unexpected X10 modules lighted ON.
>
>Symptoms were :
>P16 ON triggered sometimes also P1 and/or one of the two modules coded on
>E10.
>P16 also triggered systematically E5.
>J3 ON triggered sometimes J1.
>
>Second case was quickly fixed. The reason was a wrong setup of smarthome
>2000STW module by defining the P16 as the "group scene" (secondary) address.
>
>"sometimes" means a periodicity of 1 time each three days, but at least one
>each week
>
>All the modules are based on X10 original design, but some are true European
>version (not converted), other US with a mix of appliance and lamp.
>
>Capturing the X10 commands with a CM11 or Powerlinc USB didn't provide any
>relevant information (no trace of issued commands relating to P1, E10 or J1
>command)
>
>By the fact that triggering occured only on one module of a shared address
>(E10) for two distinct module, I decided to exchange the module without
>improvement.
>
>I tried with a scope to tune the defect module (tuning self to improve 120
>kHz detection) without success.
>
>There was no real logic related to the module place. Some were far away form
>the distribution panel, other close and problem was not limited to one room
>but always to the same module with fixed address plugged at fixed places.
>
>Finally I discovered that the problem was a power switching supply providing
>the required 5V to my wireless router located very close to the breaker box
>and therefore my ACT X10 signal repeater.
>
>I have no rational explanation, just a fact. Remove the power supply and
>reliability is back. Let it on 24/24 and you will encountered false
>triggering in a week. I have already installed signal filter (especially on
>TV and computer plug) but, on the basis of my knowledge, filter are
>adressing loosing signal strength, not false triggering.
>
>It also seems that problem is related to the X10 original electronical
>design. I didn't encountered any problem with my smarthome or ACT modules
>(14 on a total of 37 globally installed).
>
>Hope it helps
>
>Thierry
>