>Article: 12390 of comp.home.automation
>From: ranecurl@engin.umich.edu (Rane Curl)
>Subject: Signal-Bridge "Tuneup"
>Date: 14 Dec 1995 17:00:56 GMT
>Organization: University of Michigan Engineering, Ann Arbor
>Sender: ranecurl@umich.edu
>
>I lived with an unreliable wall-switch light in my X-10
>system, when I learned from this newsgroup about the problem
>of passing the control signal from one phase to the other phase
>of a three wire service. The X-10 FAQ suggests bridging the
>two phases with a 0.1 mfd capacitor at the service panel. 
>However my house has an unused 240 V receptacle that could be
>used for a plug-in signal-bridge. In addition, it appeared to
>me that the 13.2 ohm impedance of a 0.1 mfd capacitor, at 120
>KHz, might hinder obtaining equal signal strength on the two
>phases. A larger capacitor could be used, but that is more
>expensive, physically larger, and increases the reactive power
>load on the circuit from the 2.2 VA with 0.1 mfd. I decided to
>tune out the capacitor's reactive impedance at 120 KHz by means
>of a series inductor.
>
>The complex impedance of a series R-L-C cricuit is given by
>
>		Z = [1/jCw][1 - CLw^2 +jCwR]
>
>where w = 2(pi)f and j is root(-1). If one chooses L = 1/(Cw^2),
>the reactive impedance of the capacitor is cancelled, and the
>circuit impedance is resistive with Z = R. One can also
>describe this as tuning the series circuit to resonance at the
>chosen frequency. 
>
>The inductance required to bring a 0.1 mfd capacitor to resonance
>at 120 KHz is 17.6 mH (microhenries). A 18 mH inductance is an
>"off the shelf" value, and therefore that was chosen. 
>
>The finished signal-bridge therefore consisted of a 0.1 mfd,
>630 V, capacitor in series with a 18 mH inductor - plus a
>1/4 amp fuse for "discretion" - all installed in a 240 V
>cord-plug housing, which can be plugged into the 250 V receptacle
>(it could also be installed in a box as part of an extension
>cord for the 240 V circuit). 
>
>I tested this "tuned signal-bridge" and found a resonant frequency
>of 120.7 KHz. This is slightly higher than that expected from
>0.1 mfd and 18 mH (118.6 KHz), because of the tolerances of the
>components. The non-reactive resistance of the bridge was measured
>to be R = 3.3 ohms, primarily from the resistance of the fuse.
>To attain the equivalent impedance with just a capacitor and
>the fuse would require C = 8.7 mfd. This circuit, however,
>does not increase the reactive power at 60 Hz above the 2.2 VA
>expected with 0.1 mfd. The circuit is also "wide band", with
>a stand-alone Q = 4, and a half-power bandwidth of ca. 30 KHz.
>
>The X-10 signal on the two power phases will be essentially
>in phase, so this does not address the problem of obtaining
>workable X-10 control across the two phases.
>
>Oh - incidentally - it works.
>
>Rane L. Curl	N8REG  
>
-- 
Joel Davidson
IBM R/S 6000 Division           The opinions expressed are mine and do
Austin, TX                      not represent the position of IBM.