Hello all,
I have some coaxial cables in the flat I moved in and because there's coaxial cabling, I thought, bingo, I'll plug some G.hn adapters and have gigabit LAN. Unfortunately, the G.hn mesh is only stable up to three adapters, no matter where I plug them. I tried two different vendors, similar results. At 4 adapters, it starts behaving very strangely.
I then decided to properly reverse engineer (if there is such a thing as proper reverse engineering) the coaxial cabling without tearing down the walls, I stumbled upon very strange effects.
From the outside, it looks like a unicable setup: one cable goes from the basement to the first socket on floor 1, then the second, and so on up to floor 2 (appartment is on two stories and there's a basement). There are 5 coaxial sockets in the flat and all have two coaxial cables in them (I opened every single one) in and an out, except the last one, which has only an in. The splitter for the house is in the basement (I have access to it), I have disconnected the coaxial cabling of the flat from it. The cable modem is in the basement and connected to the house splitter. Then the G.hn device in the basement is connected to the coaxial cable that goes up to the flat.
Someone else lives in floor 0, which is why this coaxial cable is so important to me, it's my only way to have a local network up to the basement with gigabit speed: laying additional cables is not an option, the alternative is powerline, which I tested: it offers 20Mbps between basement and flat....
Now to the weird part: I plugged G.hn devices in 4 sockets (basement, 2 on floor 1 and one on floor 2), then removed the Axing 967-11X socket in Room 2 on floor 1, leaving IN and OUT dangling out the wall. This should normally cause everything in floor 2 to lose G.hn signal from the basement and vice versa, only that did not happen, they "only" went down from 1700Mbps to 200Mbps.
So now I suspect something is quite wrong in my coaxial cabling, which would explain why the G.hn devices are having such a hard time building a mesh bigger than 3 devices.
Induction on coaxial cables should be inexistent in my understanding. The 1 cm of naked copper dangling out the wall in Room 2 should not cause enough induction for this to happen.
I'm out of ideas, so I'm open for any wild theory and how to verify it. What could explain this drop in bandwidth instead of complete loss of signal when I disconnected the unicable in the middle?
I have some coaxial cables in the flat I moved in and because there's coaxial cabling, I thought, bingo, I'll plug some G.hn adapters and have gigabit LAN. Unfortunately, the G.hn mesh is only stable up to three adapters, no matter where I plug them. I tried two different vendors, similar results. At 4 adapters, it starts behaving very strangely.
I then decided to properly reverse engineer (if there is such a thing as proper reverse engineering) the coaxial cabling without tearing down the walls, I stumbled upon very strange effects.
From the outside, it looks like a unicable setup: one cable goes from the basement to the first socket on floor 1, then the second, and so on up to floor 2 (appartment is on two stories and there's a basement). There are 5 coaxial sockets in the flat and all have two coaxial cables in them (I opened every single one) in and an out, except the last one, which has only an in. The splitter for the house is in the basement (I have access to it), I have disconnected the coaxial cabling of the flat from it. The cable modem is in the basement and connected to the house splitter. Then the G.hn device in the basement is connected to the coaxial cable that goes up to the flat.
Someone else lives in floor 0, which is why this coaxial cable is so important to me, it's my only way to have a local network up to the basement with gigabit speed: laying additional cables is not an option, the alternative is powerline, which I tested: it offers 20Mbps between basement and flat....
Now to the weird part: I plugged G.hn devices in 4 sockets (basement, 2 on floor 1 and one on floor 2), then removed the Axing 967-11X socket in Room 2 on floor 1, leaving IN and OUT dangling out the wall. This should normally cause everything in floor 2 to lose G.hn signal from the basement and vice versa, only that did not happen, they "only" went down from 1700Mbps to 200Mbps.
So now I suspect something is quite wrong in my coaxial cabling, which would explain why the G.hn devices are having such a hard time building a mesh bigger than 3 devices.
Induction on coaxial cables should be inexistent in my understanding. The 1 cm of naked copper dangling out the wall in Room 2 should not cause enough induction for this to happen.
I'm out of ideas, so I'm open for any wild theory and how to verify it. What could explain this drop in bandwidth instead of complete loss of signal when I disconnected the unicable in the middle?
