Kitz Forum
Broadband Related => ADSL Issues => Topic started by: renluop on November 24, 2019, 01:17:45 PM
-
Is worse than I am, but has asked me, if I can help! :-X
Basically he doesn't know why, but wonders if it could be his 4 yo Billion 8800 NL v2 router. It could of course be something else, and I am thinking of suggesting his running DSLstats. My forgettery is improving every day, so guys, what of the myriad of data should I ask him to collect to indicate if the problem is local or an ISP/BT one? :-[
-
1) Graphing the SNRM to see if there is any variance over the course of the day. This is probably the most important monitoring graph.
2) Monitoring Errors - specifically errored seconds.
3) QLN, Hlog and Bit loading graphs can show the presence of RFI/EMI, crosstalk and bridge taps... although you'd probably need to post the graphs for help with interpretation.
-
Thanks! I've sent advice on to him.
Before posting here i asked for a BT Test. Results FWW are attached. Magenta says exchange is 417 metres distant from his postcode.
-
Thanks! I've sent advice on to him.
Before posting here i asked for a BT Test. Results FWW are attached. Magenta says exchange is 417 metres distant from his postcode.
Is this wireless or wired?
-
I've emailed him which, and advised should be done wired. Reply awaited.
-
Verbatim reply received yesterday
Here are some stats, everything ok until around 7pm last night when Internet signal went down and only way to resolve was to reset router. Everything fine since then.
All done wireless.
He has sent me some stats, but I suspect there may be earlier ones from Event Log. I will ask him for those, should you need them.
Whether this comment will be relevant, I'm not sure. It's about the earliest 3-4 SNR and Speed graphs. They look odd on the upside.
[Moderator edited to replace [code][/code] tags with [quote][/quote] tags.]
-
Verbatim reply received yesterday Here are some stats, everything ok until around 7pm last night when Internet signal went down and only way to resolve was to reset router. Everything fine since then.
[u][b]All done wireless.[/b][/u]
He has sent me some stats, but I suspect there may be earlier ones from Event Log. I will ask him for those, should you need them.
Whether this comment will be relevant, I'm not sure. It's about the earliest 3-4 SNR and Speed graphs. They look odd on the upside.
Sounds like the wireless on the router is not coping, or set up with legacy 80.2.11b rates, is there any chance you could find out if the firmware on the Billion is the latest version, and how the wireless is setup?
-
Have any of our big kit paws managed to get their claws on the files I posted?
He phoned us last evening and something he mentioned gave me cause to ask him more about his set up. This what I received today.
FROM box on wall Line runs along wall and enters property via porch on left hand side then runs under door frame so it enters the house on the righthand side of porch in to main socket. From there it goes under carpet across hall to BT Extension in lounge.
Then from that extension a line goes back out to the hall and up wall to extension in main bedroom immediately above lounge..
Line runs out from this extension runs along bedroom wall and over door frames in to back bedroom..
A telephone extension lead is plugged into the socket in main bedroom with the wire going back though the wall to small bedroom with the router in, sited on front of house adjacent to main bedroom over stairwell
Wire then goes to further extension in to which router telephone line is plugged into.
Noise: Plugged analogue phone into main socket (not test socket), seems quite noisy (I'll try to get recording of it at some point), hissing. ssh type of noise.
Actually he hasn't used the desktop in the small bedroom for a couple of years: they rely on their lappys in the lounge, so I'm thinking, why the extension smorgasbord, and that noisy line? Also all the cables are legacy from the previous owner.
Guys, with all this added, what is your conclusion from this and the stats, please?
-
I've taken a look at all the data contained in the "Z files". To be honest, they do not tell me much. :(
Having read your description of the - <cough> - very messy internal wiring, I am surprised that the Hlog plot is as clean as it is. Even the QLN plot is not too bad. A quiet line test (17070) now needs to be performed from a wired telephone plugged into the test socket. If the hiss like noise is then absent, then that noise has its origin with the internal wiring.
Other than that I am at a loss as to what to recommend. Perhaps another member would be prepared to comment?
-
Hi
I’m sorry I have not looked at files but from your description of cable under carpet I would think this is where the issue is
B’cat is correct, perform the QLT from test socket and if no noise, then its internal
A silly question, but do all outlets have filters on them
Many thanks
John
-
I've taken a look at all the data contained in the "Z files". To be honest, they do not tell me much. :(
Having read your description of the - <cough> - very messy internal wiring, I am surprised that the Hlog plot is as clean as it is. Even the QLN plot is not too bad. A quiet line test (17070) now needs to be performed from a wired telephone plugged into the test socket. If the hiss like noise is then absent, then that noise has its origin with the internal wiring.
Other than that I am at a loss as to what to recommend. Perhaps another member would be prepared to comment?
Thanks
The stats being uninformative, was there anything we omitted that would have helped?
A small point. As the first test did use an analogue phone was used, which I presume would have been wired, how does a second test with no noise prove internal wiring problems? I'm asking to learn.:)
TBH, being the rocking horse I am, it's not easy communicating over 100 miles, as the chance of "send three and fourpence... ;)" is likely. I'm in SW Hants, he near Maidstone.
I have thought of telling him to consider what cabling he really needs, then re-run it all with new.
Yes, any one else, please join in!
-
Hi
When testing using the test socket, all internal wiring should have been disconnected fully in order to gain access to test socket
The only unknown would be if anyone had tapped into the main pair at back of master socket but they should not as they are not allowed to touch bt side
So if noise existent from test socket but existent from an extension then it is internal wiring
If noise present from master test socket, it needs reporting as a line fault (not a broadband fault) as the noise is on bt side
Many thanks
John
-
@d2d4j. Hardly a silly question! I thought I had asked him, but will sure ask him again.
-
When testing using the test socket, all internal wiring should have been disconnected fully in order to gain access to test socket
The only unknown would be if anyone had tapped into the main pair at back of master socket but they should not as they are not allowed to touch bt side
When the front cover is off the master to expose the test socket has he tried plugging a phone in to the other sockets to make sure they are all dead with no dial tone?
-
When the front cover is off the master to expose the test socket has he tried plugging a phone in to the other sockets to make sure they are all dead with no dial tone?
To confirm you mean exactly as you write, not that he should test the other sockets before accessing the test socket? Sorry to ask, but I want to be absolutely 150% definite in what I relay to him.
-
If he takes the front off the master socket to access the test socket he should find there are wires attached to the part he pulls out which go to the other sockets. These should then all be dead with no dial tone.
If as d2d4j suggested, someone has tapped in to the main pair at the back of the master socket, other sockets will could still work. This would cause issues and he'd need to get someone in to correct the wiring.
-
Once he has verified that all the sockets are correctly wired he should try just using the test socket only with the router and handset plugged in to a filter plugged in to the test socket. If that runs OK with no drop outs, the the best bet would be a filtered master socket with the router plugged in to that.
-
Thanks once more! Next I shall concoct an email with all the points raised by the A Team. ;D
Although nothing in particular was gained from the stats wad, the spikes in the SNR per tone graph, also the green line that drops down on the r/h edge of the SNR one, why and what do they mean?
My pseudo feline curiosity is welling up.
-
. . . the spikes in the SNR per tone graph, also the green line that drops down on the r/h edge of the SNR one, why and what do they mean?
The SNR plot is showing a number of frequencies where the signal is almost non-existent. Reading from the left to right, the very first one is at sub-carrier 32 which corresponds to the frequency where the US band and the DS band end and begin, respectively.
As for that green line at the right-hand end of the SNRMargin plot, it is where the US data finally drops low enough to actually appear. Basically the US is higher than the upper limit of that plot (10.5 dB).
As it is an ADSL2+ service, DSLstats needs to be configured to show up to 512 sub-carriers and no more. Currently there is a lot of wasted-space to the right where no data can be shown. Likewise the utility needs to be configured to accommodate a SNRMargin greater than 10.5 dB.
-
If I assume that the upstream SNRM now running at ~6dB is the norm, what would be the cause of it being so high before: part of a fault?
Sorry about the SNR per tone, I should've realised, but was muddling with per band graph on my copy of DSLstats.
-
If I assume that the upstream SNRM now running at ~6dB is the norm, what would be the cause of it being so high before: part of a fault?
It's really impossible to say. (Some intermittent fault could, possibly, be behind it.)