Phase couplers - multiple paths?

[PaulD]

To avoid hijacking another thread, I am bringing up a question on phase couplers in a new thread. In the other thread, I asked a question about the use of multiple couplers when you have multiple service panels. One answer suggested that was a bad idea because it would multiply signal crossings and lead to problems. At first I accepted that notion but the more I thought about it, it conflicted with the very concept of a phase coupler. Without a phase coupler, the natural path for signals from one phase to another thru the nearest transformer...that is path #1. If you add a phase coupler, you have now added path #2 thus the very use of a phase coupler creates multiple paths between phases. Obviously (???), with a phase coupler, multiple communication paths exist and the communication signals travel a different distance thus have different timing as they pass from one phase to another. Although I am an engineer, I am not up to speed on communication theory and this leads me to several questions. Anyone here with answers? Note...my application is using UPB.

Questions
1) Since duplicate signals will be hitting devices at different times, why don't devices get tangled up in executing their commands?
2) Is there a technical or performance reason you cannot make many different connection patch via multiple couples across your electrical network (example...one coupler per service panel when using multiple service panels)

 

[aehusted]

Guilty as charged.
Perhaps I should have qualified my suggestion better in that other post. While I do have a degree in physics, I do not have any training as an EE or an Electrician. So my advice was mainly theoretical. HA is only a hobby for me.
Thanks for challenging me. I would also really like to hear the answer to your question.

 

[fitzpatri8]

The 'hitting devices at different times' concern is a non-issue for passive couplers. Electrical signals travel at, what, the speed of light, so the difference between traveling 3 feet and 300 feet is well within the tolerance of the electronic processors involved.

The added issue with multiple passive couplers is that you are coupling whatever is on the line--good and bad, signal as well as noise.

Active couplers are a bigger challenge. Unless they are designed to multicast, the way Insteon gear is, you risk that two devices attempt to rebroadcast messages that are offset just a little bit, corrupting the messages instead of improving them. The thing about binary is, if two devices are trying to talk at the same time but the messages are offset by a bit, two devices sending 101010 can end up putting a message on your powerline that looks like 1111111.