Kitz Forum
Broadband Related => Telephony Wiring + Equipment => Topic started by: badfish99 on October 25, 2018, 02:39:39 PM
-
Hello,
I've got 2 master sockets, both 'star-wired' to a junction box outside the house. The wiring is very old and not twisted-pair. One master socket I'm using for phone and VDSL, the other is redundant.
I'm far from the cabinet and get 15Mbs down, 4.5 up, so I'm looking for any possible improvement.
The brown/green bell wire pair is connected at the unwanted socket but unconnected at the main socket: the BT installer left it like that when he upgraded the socket when I moved to VDSL. These wires are crimped together inside the junction box.
I tried disconnecting the brown wire at the redundant socket and my upstream fell from 4.5 Mbs to 2.5. I'm guessing that disconnected wires are acting as antennae and then coupling the interference to the non-twisted signal wires.
If the wire to the unwanted socket accidently broke near where it leaves fhe junction box (it is very exposed and just sticky-taped to the wall), would this be guaranteed to improve my uplink speed? Or is there any danger that things would get worse instead of better?
-
My thought is that the BT installer didn't do the job properly when he changed the socket - he should have disconnected the second socket. I'd suggest that your supplier should be insisting they come back and finish the job - at no charge to you.
-
I talked to my ISP about it: their position is that BT/Openreach will say "it's working, so any change will be chargeable".
-
If you will take a series of clear photographs -- starting at the external junction box, then proceeding to both sockets -- and post the images to an external site, for our viewing, it might be possible to guide you as to what could be done to rectify matters.
-
Photos will have to wait until I can turn off the internet and unscrew everything. In the meantime I've attached a bit of ascii art to show how everything is wired up. That won't be obvious anyway from a photo, as the junction box is a rat's nest inside (actually, a spider's nest).
I've just read some advice that surge suppressors in old master sockets will interfere with vdsl. I suppose my unwanted socket contains one of those.
-
Ideally the Redundant Master socket wants disconnecting from the External Junction Box. Any additional unfiltered wiring can greatly affect vdsl even if it's unused.
I recently removed a star configuration (2 masters, only 1 used) from my parents house (whilst also fitting a nte5c and mk4 faceplate + a slave telephone socket), and there vdsl speed jumped from 27/6.5 to 40/9 after a couple of days.
-
Photos will have to wait until I can turn off the internet and unscrew everything. In the meantime I've attached a bit of ascii art to show how everything is wired up. That won't be obvious anyway from a photo, as the junction box is a rat's nest inside (actually, a spider's nest).
I've just read some advice that surge suppressors in old master sockets will interfere with vdsl. I suppose my unwanted socket contains one of those.
A simple sketch on a piece of paper scanned would be much simpler to understand, can't make any sense of your ascii art. Not at all clear what goes where.
-
In the meantime I've attached a bit of ascii art to show how everything is wired up.
Thank you. I believe I understand the current configuration.
At the junction box, disconnect all four wires from the cable that leads to the old master socket. If you feel up to it, also remove the redundant cable and socket. As pointed out by EZ Tutty, the current wiring configuration is providing a classic bridging tap across the pair . . . to the detriment of the xDSL service.
-
Thank you everyone for your suggestions and for your patience with my ignorance.
Please, before I start irrevocably cutting wires that don't belong to me, indulge me with one more question. Right now, the brown 'bell' wire in my diagram is effectively tied to one of the signal wires via the 1.8uF capacitor in the redundant master socket. When I disconnected it at that end, my upstream rate fell sharply. I *guessed* that the disconnected 'floating' wire was somehow picking up RFI and then feeding it into the signal wires.
So: if I remove that socket and all its wiring, should the brown (and green) wires be connected inside the new master socket? If so, to what? If not, will the same phenomenon occur?
-
Right now, the brown 'bell' wire in my diagram is effectively tied to one of the signal wires via the 1.8uF capacitor in the redundant master socket.
It is actually connected to the mid-point of a series connected 1.8uF and 470Kohm shunt across the incoming pair.
When I disconnected it at that end, my upstream rate fell sharply. I *guessed* that the disconnected 'floating' wire was somehow picking up RFI and then feeding it into the signal wires.
That one wire can be the source of all sorts of evil to an xDSL circuit.
So: if I remove that socket and all its wiring, should the brown (and green) wires be connected inside the new master socket?
No. Both wires should be left unconnected at both ends of that length of cable. Don't wind them around the external sheath of the cable; wind them around two fingers, then flatten the resultant oval into a "sausage" shape and tuck it into the back of the junction box & NTE5 backing box.
-
I'm not sure about the picking up RFI theory, but perhaps disconnecting the bell wire shortened the length of the bridged tap, and the changed characteristics of the bridged tap resulted in worse speed.
Normally it would be a connected bell wire that would be picking up interference, and disconnecting the bell wire solves that issue.
-
It is worth considering whether there may have been other reasons for the change in speed, when the bell wire was disconnected.
Examples...
1) If the line had previously connected in daytime, and the ‘experiment’ took place in evening, then on resync there would have been an entirely expected increase in interference, that might have offset the benefit from cutting a bell wire, especially if it were quite short.
2) If the experiment involved more than a couple of iterations, causing the modem to reconnect a few times, DLM might have perceived it as ‘instability’ and penalised the line, again offseting any benefits.
Just food for thought. :)
-
OK, I cut off the wire as recommended, and it worked: about 50% increase in upstream speed, and a smaller increase in downstream. Thank you everyone.
-
Thank you for reporting back with details of your success. :)