Re: 3 broken WS12As?

[nobody@xxxxxxxxxxxx (Dave Houston)]

"AZ Woody" <Reply@xxxxxxxxxxxxxx> wrote:

>a signal sucker can do weird and inconsistent things.

I have to disagree.

Signal suckers (attenuators) are usually passive circuits that always do the
same thing. The only dynamic element in the mix is whether they are switched
on or not. Some appliances have ~0.1µF capacitors across the mains at all
times while others have their mains switch between the capacitor and the
line. 0.1µF acts as a near dead-short to 120kHz - it's the value recommended
for passive phase coupling. It's also a value frequently recommended for
decoupling DC power supplies.

In the US, the FCC limits the amount of "conducted" noise on the mains in
the AM radio band (approx. 450-1700kHz). The design engineers may also wish
to prevent any RF frequency noise from entering their device from the mains.
It's not surprising that they would choose a capacitor that happens to be
X-10 unfriendly.

I'm not an expert on the line filters used with PC power supplies but the
author of this page ( http://my.execpc.com/~endlr/line-filter.html ) appears
to have some expertise. If his "typical" line filter (about 40% down the
page) is truly typical they also use 0.1µF across the mains.

Switchmode power supplies should also have a more or less constant effect,
again depending on whether they are switched on or off. If they have
inadequate filtering, the frequency used internally (this frequency varies
from manufacturer to manufacturer) will be seen as 'noise' on the mains
which, depending on the frequency, may or may not pose problems for X-10. If
they are adequately filtered, the filtering may or may not attenuate X-10
signals, depending on capacitor values. Noisy CFLs would seem to fall in the
inadequately filtered class.

It might also be possible that two or more switchmode power supplies having
frequencies which are individually X10-benign create a beat frequency that
will create problems for X-10. This is the only scenario that I can envision
with a dynamic element (both must be on to create a problem) but there may
be others.

There are other factors that may contribute. For example, it may require two
or more signal suckers to be active before problems appear.

It's difficult to envision and difficult to trobleshoot without a signal
strength meter. True noise problems may require an oscilloscope and/or a
spectrum analyzer. Since most of us have none of the above, we speculate and
curse the undarkness in the middle of the night. ;)