I think I confused things a little with my comment about the neutral not being isolated. What I was thinking was that if I wire nut all the neutrals together, then the neutral on the protected side of the UPS is for sure not isolated in any way from the neutral on the input side of the UPS. If there is any MOS protection, for instance, in the neutral line, then I would "short it out" by tying all the neutrals together. I believe quality surge supressors do have protection in both hot and neutral lines. If the UPS has quality surge suppression (in addition to the battery backup), then I have just "neutralized" :-) the neutral line surge protection.
But... as you observe, there are two neutrals in the device I indicated as an example. So, thinking better this time, I guess I could make a separate splice for the neutrals on the output side of the UPS (rather than a single, big wire-nutted bunch), thus keeping the "separate identity" of the output side.
Maybe that would both allow me to get X10 signals into the protected side of the UPS, while still preserving any isolation and/or protection on the output side of the UPS.
Regarding your two points:
1) I don't fully understand. Wouldn't the bridge be adding the X10 signal on the output side of any filtering. That's what I'm trying to accomplish, anyway: getting the X10 signals on the other side (output side) of any destruction from SCR chopping, ferroresonant devices, etc. (UPS stuff). That would also be on the output side of any filtering, thus avoiding filtering attenuation. Or so my X10-challenged mind thinks. I welcome being set straight on this.
2) I agree that filtering efffectiveness would be reduced. Whatever (X10 signals and noise) is in the bridge passband in the source circuit would be added to the destination circuit (output side of the UPS). I don't see any way around that. But, it would be no worse off than all the other non UPS protected devices. Besides, my chief purpose in using the UPS is not surge suppression, but avoidance of power interruption.
I just want to find a way to provide X10 control (without having the X10 signal disappear into a black hole at the UPS) of an appliance module that will be plugged in to a UPS. (Actually, the appliance module might be physically plugged into an outlet in an outlet box fed from the UPS. The output side of the bridge would also be connected to the circuit in this same box. Thus, I "inject" the X10 signals "downstream" of the UPS.)
I would welcome all comments about how one controls UPS protected X10 devices.
