Cat6 phone extension causing VDSL sync issues

[DragonQ]

Just had an Openreach engineer round and he found two issues that may have been causing our falling sync speeds and occasional evening of huge packet losses. One of those was the Cat6 cable I had running from the VDSL filtered faceplate (Mk3) into my patch panel. Despite the fact that it wasn't plugged into anything on the other end during the tests he ran, maximum sync speeds jumped from 65/15 to 80/22 Mb/s when he simply unplugged the cabling from the faceplate.

He recommended putting the modem next to the master socket and then using the existing extension cable as a WAN cable from there to the router instead, which is fine but the whole point of the cabling was to avoid having to put broadband equipment near the master socket. I don't think the wiring at the faceplate end was bad since he re-did it himself and it didn't help. The cable itself could be damaged somewhere, or the wiring into the patch panel could be dodgy I guess. It seems to me either of these things would affect its performance as a WAN cable too. Is there any way I can find out why this is happening without repeatedly disconnecting my modem (which'd ruin my speeds even more)?

Thanks in advance for any help.

 

[Mr.Shabz]

I'd keep the modem at the master socket end for a couple of weeks and see if you still get your original issues.

That way your modem will keep its sync with the WAN and you only loose LAN side traffic, thus keeping your 80/22 profile.

I'd also check all of your cat6 internal cabling with a proper tester, just to make sure there are no short pairs/damaged cables.

 

[ChrisD.]

Proper testers aren't cheap though unfortunately.

I have my modem at the master socket for this exact reason. Everything I tried to remote it to my server cupboard just dropped the connection speed.

 

[bledd]

Just had an Openreach engineer round and he found two issues that may have been causing our falling sync speeds and occasional evening of huge packet losses. One of those was the Cat6 cable I had running from the VDSL filtered faceplate (Mk3) into my patch panel. Despite the fact that it wasn't plugged into anything on the other end during the tests he ran, maximum sync speeds jumped from 65/15 to 80/22 Mb/s when he simply unplugged the cabling from the faceplate.

He recommended putting the modem next to the master socket and then using the existing extension cable as a WAN cable from there to the router instead, which is fine but the whole point of the cabling was to avoid having to put broadband equipment near the master socket. I don't think the wiring at the faceplate end was bad since he re-did it himself and it didn't help. The cable itself could be damaged somewhere, or the wiring into the patch panel could be dodgy I guess. It seems to me either of these things would affect its performance as a WAN cable too. Is there any way I can find out why this is happening without repeatedly disconnecting my modem (which'd ruin my speeds even more)?

Thanks in advance for any help.

This is what I always recommend, modem at the point of entry. Remove all extentions from your phone line (if possible)

 

[Elevon]

This is what I always recommend, modem at the point of entry. Remove all extentions from your phone line (if possible)

I agree, my SNR improved quite a bit(which improves stability) when I moved my router/modem to the master socket, now I will always leave it there, if I need very good wifi or wired upstairs I actually use powerline/wifi adaptors with no issues. I've my equipment wired via powerlines and network switch where I need it, also with very good wifi range even on 5GHz with my powerline/wifi adaptors.

 

[Caged]

What does the punching down look like on your patch panel? It's possible you've built a really effective antenna.

I assume you've wired things into the patch panel as an extension from the IDC terminals on the filter faceplate, and the tests were being done from the socket on the faceplate itself? I would expect more issues from a run of cable ending with a patch panel and nothing plugged in, than when there's a modem on the end of the cable.

Try terminating the run onto a tool-less keystone, might as well go shielded if you're buying new modules.

http://cpc.farnell.com/tuk/skfs/keystone-jack-6-6a-toolless/dp/CS30447

 

[Avalon]

A brand name cable tester such as a Fluke is not cheap, however the 'very similar ones' available from most of the large online market places are reasonably priced and for home usage would likely be adequate. The thing is the solution put forward by the BTO tech is my preferred option - you want the absolute minimum between the modem and the master socket.

 

[pcfarrar]

If the CAT6 run is quite short there is no way it should make that big a difference to the sync speed. I have a 30ft run of CAT5 cable from my master socket to the modem and the difference it makes to the SNR and sync speed is minimal.

 

[DragonQ]

Proper testers aren't cheap though unfortunately.

Yeah this is the problem, having the engineer around makes it easy but he can't stay here and help!

I have my modem at the master socket for this exact reason. Everything I tried to remote it to my server cupboard just dropped the connection speed.

It just seems so bizarre. I've had no issues with any other cabling I've done so I'm not sure how it could be a problem. Also a few metres of Cat6 should be nothing compared to the 350 m of crappy wiring to the cabinet, surely? I think he said the signal dropped from 17 dB to 16 dB when he removed the Cat6 too.

This is what I always recommend, modem at the point of entry. Remove all extentions from your phone line (if possible)

We have no extensions plugged in (aside from this Cat6 cable hanging off the unfiltered connections on the faceplate).

What does the punching down look like on your patch panel? It's possible you've built a really effective antenna.

I assume you've wired things into the patch panel as an extension from the IDC terminals on the filter faceplate, and the tests were being done from the socket on the faceplate itself? I would expect more issues from a run of cable ending with a patch panel and nothing plugged in, than when there's a modem on the end of the cable.

Try terminating the run onto a tool-less keystone, might as well go shielded if you're buying new modules.

http://cpc.farnell.com/tuk/skfs/keystone-jack-6-6a-toolless/dp/CS30447

I can't grab a photo since I'm away for a week but I did all of it myself and no other issues that I've found. Surely if the punching down was done badly this would have manifested itself in a far worse way?

A brand name cable tester such as a Fluke is not cheap, however the 'very similar ones' available from most of the large online market places are reasonably priced and for home usage would likely be adequate. The thing is the solution put forward by the BTO tech is my preferred option - you want the absolute minimum between the modem and the master socket.

I do have a basic network tester that shows all the cabling is wired up properly but it doesn't say anything about the signal quality, etc. so it doesn't seem useful in this situation. Having the modem at the master socket indeed is the optimal solution, it just looks daft on the wall in the middle of the living room.

If the CAT6 run is quite short there is no way it should make that big a difference to the sync speed. I have a 30ft run of CAT5 cable from my master socket to the modem and the difference it makes to the SNR and sync speed is minimal.

An issue I have is that I can't see stats when using the Openreach modem. When using my router's internal modem I can see the SnR to compare but not with the Openreach one. I know there are ways of "unlocking" it but all of the custom firmware images I've found online seem very outdated so I'm not sure if it's a good idea to try this.

 

[Caged]

I assume you've only punched one pair down to the rear of the phone socket?

 

[DragonQ]

I assume you've only punched one pair down to the rear of the phone socket?

Yeah, blue and blue/white into A and B.

 

[bremen1874]

Have you tried checking for mains powered devices causing interference?

I'd start but disconnecting anything that lives near the patch panel from mains (PSU out of the wall).

If that doesn't work keep unplugging things until you hopefully find the problem. I'd even go as far as flipping the breakers on unnecessary circuits.

 

[GeX]

Did he try running the test from the socket and also on the patch panel end? I can see why a floating / ungrounded cable might cause an issue, but if the test was being run from the end of it.. it should have been fine.

aside from this Cat6 cable hanging off the unfiltered connections on the faceplate

Why the unfiltered ones?

 

[Caged]

Because presumably the VDSL needs to work on the other end of it.

 

[GeX]

For some reason I'd thought that the filter filtered out the voice signal rather than the filtering out the VDSL signal for normal phones.

Not sure why.. I have a MK3 faceplate with an extension connected to the back of the faceplate for my VDSL. I'll blame it on a long day.

 

[DragonQ]

Did he try running the test from the socket and also on the patch panel end? I can see why a floating / ungrounded cable might cause an issue, but if the test was being run from the end of it.. it should have been fine.

No he only did tests to the front porch (where the BT cabling used to enter the property from the outside) and from the master socket. In the porch he got good results, then he got poorer results using the master socket with the Cat6 extension plugged in, then he got good results again by unplugging that Cat6 cable.

According to Plusnet, our sync speed is now 75 Mb/s after 3 days of solid connection, whereas before it was stuck at 66 Mb/s for weeks. When I get home I'll do some actual speed tests to see if it's any better in real usage. Still need to get an RJ45 connector if I want to keep this setup permanently though.

 

[Silly Goose On The Loose]

If you connect the modem to the socket whilst having a hardwired extension plugged in, you'll get a bridged tap.

 

[GeX]

bridged tap?

 

[DragonQ]

If you connect the modem to the socket whilst having a hardwired extension plugged in, you'll get a bridged tap.

Ah I see. So there might be nothing wrong with the extension wiring after all, just the method used for testing. It's a shame I can't test myself the difference between the master socket and the patch panel extension.

 

[DragonQ]

So after I asked Plusnet to reset my DLM (as download speeds were still ~60 Mb/s despite the modem syncing at 75 Mb/s), they identified another "possible fault" on the line. I haven't plugged the Cat6 cable back in yet. They asked me to remove all phone and broadband equipment from all sockets and remove the faceplate on the master socket for the day so that they can conduct tests. Anyone know what kind of tests they'll be doing that require removing the faceplate??

EDIT: They just told me they were detecting a "loop fault" before but not with my equipment and faceplate unplugged. No idea what this means, but it can only be caused by the two things that were plugged in when they did the original test:

- The modem
- The wiring going to the extension socket upstairs

EDIT: I have disconnected the unused extension socket and reconnected the Cat6 data extension cable. My router is now plugged directly into my patch panel as before and I'm now getting sync speeds of 76/20 Mb/s again. Speedtest.net shows 71/19 Mb/s which is pretty good and it might get higher, who knows. Plusnet also said there was a capacity problem at my exchange that is due to be sorted by the end of the month, but who knows how much effect that is actually having. Will keep an eye on it but right now everything looks good.