SPGDCM@CMSA.BERKELEY.EDU (02/23/89)
MSG:FROM: SPGDCM --UCBCMSA TO: NETWORK --NETWORK 02/22/89 19:56:08
To: NETWORK --NETWORK Network Address
From: Doug Mosher
Subject: bizarre X10 problem
To: sci-electronics@ucbvax
I've had a houseful of X10 devices in my house for many years. I am familiar
with most of their quirks by now. For example, sometimes a signal does not
reach all the way between a desired control location and a desired device. And
the units do fail and need replacement. And it is helpful to put a .1
microfarad 600 v capacitor across the 220 to help signals bridge between the
two "halves".
Another reason that sometimes controls will not work, only discovered
recently: if any of your sending devices breaks and starts sending a
continuous signal, that will make other control actions less reliable, but
will not kill them all! For example, I had a computer controller fail and send
continuous signals, and lost the ability to turn off half my lights, but not
all of them, until I reset the errant controller.
But now I have an apparently insolvable problem that bugs me a lot. I have
used a wall switch module to control my overhead living room lights (400 W
incandescent track lites, 8-50W units, wall switch module rated 500W).
I control the lights from the module itself and from 3 other controllers plus
a homeminder. Everything has worked for years.
Recently I lost the ability to remotely control this circuit. The wall button
still works, but I can't turn the lights on or off or dim them from any of the
3 manual controllers nor the homeminder. (All functions are lost: on, off,
dim, bright, and all-on all-off).
I replaced the bridge capacitor, no help. I have tried 3 different wall
modules, of different vintages, and they all fail exactly the same way. Note
that the failure occurs even with all-different components, sender and
receiver.
About the only thing I can think of is that recently I got TV cable service
installed, and at the same time redid all the wiring in my hi-fi area
(rebundled all the power cords etc. to make them look nice). I am aware that
the X10 stuff is somewhat sensitive to what you do with your cords; perhaps
this is a problem with coiled cords being capacitive leaks or standing waves
or whatever, caused by the new cable or the rearranged power wires.
I'm sure curious what to do to fix this. Having controlled my living room
lights for a decade, and now all of a sudden not being able to no matter how
many devices I replace, is very inconvenient and very weird.
Obviously it would be nice to be able to see the signals at this and that
location. Years ago I bought a scope and found that the tiny little
line modulation signal, on top of the giant 60 cycle signal, is like
trying to study a flea on an elephant. Tiny, wiggly, and one looks like
another; can't see the real code being sent.
Any ideas?
( )
( Doug Mosher <SPGDCM@CMSA.Berkeley.edu> )
( ...!ucbvax!cmsa!spgdcm )
( 257 Evans, Univ. of California, Berkeley, CA, 415/642-5823 )
bizarre X10 problemjim@trsvax.UUCP (02/25/89)
You might use a 60Hz notch filter (or just a high-pass - cheaper) on the
'scope input when looking for the carrier-current signals.
btw: please don't use the earlier posted resistive 60Hz hum filters 8{)
James T. Wyatt UUCP:decvax!microsoft!trsvax!rwsys!jim KA5VJLroyf@killer.DALLAS.TX.US (Roy Frederick) (02/27/89)
A circuit for a twin-T network useful for viewing signals on the power line is described in the book "Handbook of Remote Control & Automation Techniques" by John E. Cunningham & Delton T. Horn (TAB Books # 1777). The circuit is a 60 cycle notch filter. One leg has DC continuity - so you must be sure that that leg is connected in the neutral line - else sparks will fly when you connect to your scope! It might be prudent to measure between that line and your scope ground before making the connection just to be sure. Well, here is the circuit given in the book: |-----------------------------| | | |----r2-------|---r2-----| | | | | Line in | | | o-----------| |--------o out | | | | | | | |----c1---|-------c1-----| | | | | |------------------------| r1 c2 | | o---------------------|--------|---------------o Neutral out Component values: c1 - 9 uf at 600v c2 - 18 uf at 600v r1 - 500 ohm pot set to 147 r2 - 500 ohm pot set to 295 Unfortunately, with these values the device is going to be bulky and expensive. The author justified these values because he had some 9 uf surplus caps available. It is my opinion that at the frequencies you want to see, much smaller capacitors could be used. Perhaps the following values would be ok: c1 - 0.05 uf at 600v or c1 - 1 uf at 600v c2 - 0.1 uf at 600v c2 - 2 uf at 600v r1 - 26500 ohm r1 - 1325 ohm r2 - 53100 ohm r2 - 2650 ohm Some adjustment of the pots will be necessary to null out the 60 hertz signal due the the inaccuracy of the values of the caps. Disclaimer: I have not tried this circuit. It looks like it should work, but there is danger any time you are dealing with the power line. Be careful. BTW, Jensen Tools sells a Line Noise Viewer that provides line isolation for the tiny sum of $450 or so. I don't know if the bandwidth is good enough to see X10 signals, though. Roy Frederick Dallas County Data Services 504 Records Bldg Dallas, TX 75202
henry@utzoo.uucp (Henry Spencer) (02/28/89)
In article <7334@killer.DALLAS.TX.US> royf@killer.DALLAS.TX.US (Roy Frederick) writes: >The circuit is a 60 cycle notch filter. One leg has DC >continuity - so you must be sure that that leg is connected >in the neutral line - else sparks will fly when you connect >to your scope! It might be prudent to measure between that line >and your scope ground before making the connection just to be >sure. It *would* be prudent. More prudent would be to avoid using circuits designed by the ignorant. Neutral is not ground. NEUTRAL IS NOT GROUND! ---> ***NEUTRAL IS NOT GROUND!!!*** <--- -- The Earth is our mother; | Henry Spencer at U of Toronto Zoology our nine months are up. | uunet!attcan!utzoo!henry henry@zoo.toronto.edu