Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[limbo]

Hi there,

Retyped this all since I lost it the first time because I had an attachment over 900kB!

I'm new to this forum, but from the few posts I've read, it seems to me that there are a lot of experienced engineers and troubleshooters around!

So we have been with BT since 2004, and we first started with 1/2 Mbps. This was then upgraded to 1Mbps. However it seems to have dropped now (for a good few years) to 333Kpbs. Take a look at a test I ran earlier this evening (theses stats are fairly standard for our connection. It is odd that the upload speed is faster than the download speed!):



According to the BT line speed tester, it reckons we should get 1Mbps.

I've had a look at the Kitz Exchange Checker, and these are the figures it gives:

Distance:-    Direct:          2.59 km
     (appx)*    By Road:    3.22 km

Here is some data I pulled from the BT Homehub 2 earlier:

ADSL line status
Connection information
Line state   Connected
Connection time   0 days, 2:42:50
Downstream   575 Kbps
Upstream   462 Kbps

ADSL settings
VPI/VCI   0/38
Type   PPPoA
Modulation   ITU-T G.992.5
Latency type   Interleaved
Noise margin (Down/Up)   8.9 dB / 7.6 dB
Line attenuation (Down/Up)   55.5 dB / 29.8 dB
Output power (Down/Up)   17.0 dBm / 12.0 dBm
Loss of Framing (Local)   53
Loss of Signal (Local)   196
Loss of Power (Local)   0
FEC Errors (Down/Up)   515212 / 427
CRC Errors (Down/Up)   2079 / N/A
HEC Errors (Down/Up)   N/A / 124747
Error Seconds (Local)   298

Do these stats make sense with the distances involved?

I think there may be some potential issues with the wiring of the house:

When I went to look at the extension wiring in the loft, it looked like a rats nest! I did spot a REN booster; would that have any effect?

I installed a BT iPlate last year, but when it was fitted, the modem could not sync with the DSLAM. The BT Tech Support guys were stumped at this, and they ran some tests and informed me that there was a 'fault' on the line. An OR engineer was sent out (not sure which type), and he ran an attenuation test (not sure if that's the correct name!) from the test point in the house to the exchange, and got a figure in the 50s (sorry, it was a long time ago and I didn't have the foresight to write it down!) he said that the tech support guys had a figure in the 70s, which is why he was called out. He said that as far as he was concerned, there wasn't a fault and left it at that. I didn't really chase the issue up, until now. The BT tech support people also weren't happy to hear that I wasn't using the HH2.

We seem to have always had an intermittent problem that when we receive an incoming phone call (when the phone starts ringing), the modem loses sync with the DSLAM. It seems to come and go. For example it's happened twice today, once in the morning and once late afternoon. There have been a good fifteen incoming calls (admittedly a few were when I was trying to provoke it!). Sometimes it happens with virtually every call, and sometimes we don't see it for a good month or two.

There is also a fair amount of background noise on the line if I try a silent line test.

I have switched to my old Netgear DG834Gv1, and spent the evening logging the line. Below is a bit from this evening. The dip in the SNR coincides with when I picked up the phone to make a call.

Is there any logging which I can do overnight/for a longer period of time? I have noticed that the HH2 does occaisonally lose sync with no apparent reason.

I guess the next thing to try is to plug the router and a phone in to the test point and see if any of the problems are also experienced there. The problem with the disconnection is that they don't happen consistently.

When we learnt that we were in a FTTC enabled area, we jumped at the chance. Because the BT speed estimator gave us 11Mbps, it meant that we couldn't sign up for infinity services, and their Total Broadband FTTC Option 3 cost a bomb, so we're switching to Plusnet. If we were actually getting a reliable 1Mbps service then we may not, but this is so slow that streaming anything is a no no. Don't even think about someone trying to download something while someone else tries to browse the web!

They are coming at the beginning of next week to install this (yeah, I wish I had found this forum a bit sooner!). My concern is that there may be a fault with the line somewhere other than in the house (if it is in the house, I guess it doesn't matter too much any more), which then might also affect the FTTC service. If this is the case, I don't know whether it can be looked at when FTTC is set up- from what I have read, a lot of problems can slip through their 'standard set up tests' (such as LTOK?)?

From what I have read in some of the other topics, I could try to guess some faults, but I think a little knowledge in this area is a
dangerous thing! I was hoping that some of you experts might be able to provide more of an insight in to what is going on!

Phew, sorry for the long post (and congrats if you made it this far  ), I just wanted to get as much info as possible out there!

Thanks in advance,

Thomas

[attachment deleted by admin]

[waltergmw]

Hello limbo and welcome

Your line seems to be significantly under-performing. If the iPlate is still fitted I would remove it and remove the (ring) wire in terminal 3 of your master socket. Ensure your modem is connected directly to the master socket and try again.
You'll see from this link

http://www.kitz.co.uk/adsl/max_speed_calc.php

that Kitz's speed estimate is higher than you actually achieve.

It does appear to me that you have a high resistance fault on the line if you get a disturbance when you make a call.

Others may have more comments later

Kind regards,
Walter



[attachment deleted by admin]

[roseway]

I would second the above, but if it makes no difference then you should connect the router to the test socket and leave it there for long enough to establish whether you still get the same symptoms. It's the only way to be reasonably sure that your own wiring isn't the problem.

[burakkucat]

I agree with both Walter's and Eric's comments.

There definitely seems to be a fault in your copper pair. Whether it is your side of the NTE or the OR side is the first thing to determine. If you can clearly see where the OR D-side service cable reaches your home (underground or aerial) and are able to follow it to the NTE & then to all subsequent sockets, a clear set of photographs would be most beneficial to our understanding.

Without full information, remote diagnosis is fraught with errors and can result in an entirely inappropriate course of action being taken.

[limbo]

Hi everyone and thanks for your replies (and thanks Walter for the PM).

If the iPlate is still fitted I would remove it and remove the (ring) wire in terminal 3 of your master socket. Ensure your modem is connected directly to the master socket and try again.
You'll see from this link

When I fitted the iPlate, it prevented the modem from syncing with the DSLAM at all- hence it was a very short lived experiment!

[limbo]

I agree with both Walter's and Eric's comments.

There definitely seems to be a fault in your copper pair. Whether it is your side of the NTE or the OR side is the first thing to determine. If you can clearly see where the OR D-side service pair reach your home (underground or aerial) and are able to follow it to the NTE & then to all subsequent sockets, a clear set of photographs would be most beneficial to our understanding.

Without full information, remote diagnosis is fraught with errors and can result in an entirely inappropriate course of action being taken.

It's an underground cable, and I know where it comes in. I can take a picture of the incoming line and the master socket (with the cover taken off)- do you want pictures of all of the extension sockets? There are a surprisingly large number of them, and some of them got disconnected before I was around (I've not tried to figure out which ones are still connected) for ages.

[limbo]

I would second the above, but if it makes no difference then you should connect the router to the test socket and leave it there for long enough to establish whether you still get the same symptoms. It's the only way to be reasonably sure that your own wiring isn't the problem.

I am going to do this tomorrow when no one else is around! Which router is better for this kind of work? Do you think I should continue using the DG834Gv1, or use the HH2? I ask because I can't seem to get the attenuation figures from the DG834G (just SNR). This is probably me being a little silly, though.

I'll post data here as soon as it's available- presumably there should be an immediate difference in the SNR and/or attenuation?

[limbo]

I can't seem to get the attenuation figures from the DG834G (just SNR)

Sorry to clarify this, I can only see the SNR data in RouterStats- nothing on the line attenuation.

Thomas

[burakkucat]

It's an underground cable, and I know where it comes in. I can take a picture of the incoming line and the master socket (with the cover taken off)- do you want pictures of all of the extension sockets? There are a surprisingly large number of them, and some of them got disconnected before I was around (I've not tried to figure out which ones are still connected) for ages.

If you are quite sure that there is not a junction box anywhere before the NTE5/A, then a set of photographs of where (1) the service cable arises from the ground, (2) of its connection to the supply feed cable which enters your home and (3) the connection of the latter to the OR side of the NTE5/A should prove to be sufficient, for now.

If you also wish to include a photograph of the connection of the extension wiring to the customer's face plate at the NTE5/A, that might provide an indication of the standard of your extension wiring.

[burakkucat]

@Thomas -- If you would be willing to send me your postcode in a PM, I will look it up in a copy of the leaked BT Wholesale database file that I possess. Hopefully it will identify the PCP through which your line passes. Then all you will need to do is to walk around your neighbourhood and look for the correct numbered cabinet and its fibre companion.

[limbo]

Morning everyone,

I've taken some photos, and here are the results:

Outside the house. The cable which on the left is the one which goes to the main socket (and test point).


Presumably there is a junction box in there? Did BT install those two extension points? There are two points around which are smaller than a standard single telephone socket/single power socket/network point/light switch (you know what I mean!). They also have the same old style T logo that the test point has. However I have just gone and tested the one socket (which looks like that) which is still left- it is dead. Looking on the outside, I would guess that someone rewired this but didn't remove the cables?


A closeup of that wiring: I hadn't spotted that before: is this bad?


On the inside:


With a clseup of our internal wiring:


With everything else removed, the quiet line test was certainly...quieter! It was not silent, though. How quiet should it be? I would describe the sound now like a quieter version of the pressurisation noise you get when you're on a plane.

Thanks,

Thomas

[limbo]

I also (as you can probably see) switched to the test socket last night.

I took some data in the wee hours of the morning:


and a little more:


and some stuff from a little earlier on today:


I put my PC to sleep (forgot to turn data logging off). I think this is where the line drawn down the graph is. I came back to find that according to RouterStats, the SNR had dropped to 18dB, from about 30. I don't think that the first drop to 0dB coincided with any incoming/outgoing phone calls.

I got very confused about this, and so I (perhaps stupidly) went to power cycle the router. Since then the connection has been very stable, sitting at 31dB with only 2 dips by about 1dB in the past hour or so.

After running some speed tests, I can say that the speeds on this line have not changed.

Thanks,

Thomas

[Oranged]

Your Noise Margin is very erratic isn't it ?

On the Routerstat graphs you've posted so far it ranges from 7dB to 31dB 

[roseway]

I don't know what's inside that grey box on the outside wall, but the wiring to it looks very amateurish, and I can't believe that BTOR would have left it in such a state. Perhaps someone else can identify it for us. If it's just an external junction box, then I think you need to get BT to regularise the wiring from there to the NTE5, and if it was the work of an unofficial person then I guess you'll be charged for that job.

The internal wiring also looks a mess (perhaps the work of the same person?). There seem to be wires connected to all six terminals in the NTE5 faceplate. The whole lot needs to be stripped out and done properly. This is a job you can do yourself - all you have to do is connect terminals 2 and 5 in the faceplate to the same terminals in the extensions, using a single twisted pair for the purpose. Conventionally you would use the blue/white - white/blue pair for this.

Until the wiring is sorted out it's not going to be possible to diagnose the performance problems.

[burakkucat]

I've just reviewed your photographs, Thomas. Oh dear, not very good at all.

Unfortunately, as Eric has said, you will need to have OR to resolve that external "pig's ear". However as you will be having an engineer visiting to make a check and install the modem, it could all be resolved there and then. It really depends upon whether the visiting engineer is an ex-armed forces new-recruit or a time served CSE such as Mr Pag.

Stock up with chocolate biscuits, bacon (sandwiches, for the use of), tea and coffee. Then as soon as the engineer has arrived, show her/him the outside "folly" and say that in your opinion it does not look correct -- what does (s)he advise?