There are no fixed rules, and I suspect vendors say otherwise so that their products are not blamed when used in situations with degraded performance.
The performance of powerline adaptors depends on (apart from the specification/performance of the devices) the attenuation and differential phase delay of the modulated carriers between devices, and the noise at those frequencies, on the path between the devices (mains wiring and connections).
All else being equal, the shorter the path, and the fewer joints in the cable (e.g. mains plugs, wall sockets, distribution boards, junction boxes, extension cables), the better speed you're likely to get.
In the UK it's common for different floors of a house to be on different fused circuits from the distribution board. I've been told by shop salesmen that powerline "just does not work" if the two devices are on seperate circuits, but it simply isn't true. There is still a path between them. Similarly, though it's recommended that the devices are plugged directly into the wall outlet, they will still (likely) work at the end of a long extension lead, it's just that the increased path length and extra junctions will degrade the signal somewhat, and thus potentially (but not necessarily) the data speed.
When talking about RF over copper, cable quality etc., a colleague of mine used to say "A piece of wet string will work if it's short enough". Well, maybe not, but you get the idea.
In practise, the only thing most folks can do when planning to use powerline is to suck it and see. Try it, find out if performance is good enough. Then experiment, keeping in mind that shorter path and fewer junctions are likely to be better. Also, keep it to two only if possible, as a third adaptor IME seriously degrades performance, as they have to share the available bandwidth, discriminate between each other, error correction working harder, etc.
Quote:
Originally Posted by SaintDanBert
Q1 Can someone tell me how to couple two circuits of my house mains for use of "powerline ethernet" (EOPL, IEEE 1901)?
It sounds to me like there is some sort of frequency pass-band filter that could enable the EOPL modulation to move from leg to leg without causing mains electrical troubles. Sadly, I've forgotten any knowledge I ever had about how to actually create something like this. I hope that there is some sort of existing product or known-to-work hack other there.
~~~ 0;-Dan
|
Quote:
Originally Posted by michaelk
|
Although USA domestic dwellings have a 120/240Vac supply, it's still single phase (AFAIK). Assuming OP is in a domestic dwelling, it's very unlikely he has a 3-phase supply, so at first look this device would not be suitable. Thinking more about it though, it may be the the case in USA (I don't know) that 120Vac outlets in different parts of the house are on different ends of the 240/120Vac transformer, in which case a coupler of some kind at the distribution board could help. Keep in mind that in USA you'd then be working with a 240Vac potential between the circuits, and with 3-phase in the UK, 440Vac. Advisable to use a qualified electrician IMO.
EDIT: I've read the OP in more detail (as I should have originally, apologies), and now think that something like the device michaelk linked to is highly likely in theory to help the OP get better point to point speeds between powerline adaptors on "Line 1" and "Line 2". That's assuming that the path loss through the coupler is lower than that through the transformer. A coupler designed for 3-phase would involve undesirable but unavoidable losses, and you'd be better with one designed specifically to couple between L1 and L2, with a single LC pass filter rather than three. I haven't seen one after brief search, but they may exist. It might be that the two paths cause phase cancellation, and you'd end up needing an RF choke or ferrite ring in the line from the transformer to get the benefit of the coupler. Again, short of a (moderately complex and so expensive) RF survey, I don't see a better option than to try it and see.