Print

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

[JohnJBurness]

I've seen all of the very helpful info, on Kitz's site, but it doesn't appear to answer a specific question that I'm interested in!

Is there anyway that I can estimate what my Stats should "typically" be, if I know the approximate overall length of the Cable between my home & the Exchange?

Based on BT supplied info, I know that my overall Cable Length is less than 500m (I do mean Cable Length, NOT straight line distance), however I'm currently getting approx 0.9dB Noise Margin on a fixed 2Meg connection.

Clearly I have a Cable Fault, but I want to ensure that BT properly fix the Fault & not just "marginally-tinker" with it! Hence I need to know what would be "reasonable" for that length of line!

Many thanks,
John

[roseway]

Hi and welcome.

What are your current router stats? The cable length doesn't really tell us anything.

[JohnJBurness]

Hi,

Thanks for the Welcome!

My current Router Stats are meaningless, to my question, because I have a Cable Fault! All the current Stats will confirm is that I definitely DO have a Fault (e.g. SNR of 0.9dB - taken at Master Skt with all other Cable/Equipment removed).

Assuming that my Cable was in a good condition, I would have assumed that I should be seeing something like Attenuation of 10dB (or less) & somewhere in the order of 25dB (or better) for SNR. It is these "Good Cable" conditions that I'm trying to establish! Are my assumptions reasonable?

Regards,
John

[mr_chris]

Hiya John

As regards "good cable" conditions, yes you are correct in your assumption, that for a cable length of 500m, the attenuation and SNR should be in the region that you stated, probably better.

I'm interested in what your full stats currently are (i.e. error counts, whether you're getting disconnected and how frequently, any graphs of SNR over time if you have them, etc). If you do indeed have a cable length of 500m, they may well give a clue as to what is possibly wrong with the line itself, and we might be able to advise on the best way to tackle the ISP about the fault.

Without posting what the stats are as they currently stand, it doesn't give us much to go on, except to say yes.. your assumptions are correct - but I guess the geek in me is interested in what's making your line perform so badly - I'm sure I speak for Eric too when I say that!

[JohnJBurness]

Hi, mr_chris,

The fault has already been reported to the ISP & a BT Openreach Broadband Engineer has already been out & confirmed that it is a fault on the external underground Cable. He also confirmed that there were NO spares available on either that Cable nor any other nearby Cables. This is how I managed to establish the Cable routing (direct to Exchange & NOT via any Roadside Green Box) & its approx length (less than 500m).

I would also add that this Openreach Engineer expressed extreme surprise that I managed to ever connect!

Originally BT gave me an assurance that they would be providing a complete new Cable & that this would be fully completed by close of Business last Friday.

However, BT have already back-peddled on this promise & I suspect that they are going to trying to attempt the minimum to establish a connection - regardless as to how poor it is. Hence I'm trying to obtain the realistic figures that I can use as a "yard-stick" in my increasingly acrimonious discussions with BT.

I keep getting long periods of disconnections (sometimes over a few days). Even when I am connected, my effective "usage-speed" (presumably after it is coping with all of the errors) is almost down to dial-up levels.


However, since I'm currently connected, my stats are:-

ADSL Link           Downstream          Upstream
Connection Speed        2272 kbps              288 kbps
Line Attenuation         10.0 db                      5.0 db
Noise Margin              0.9 db                9.0 db
The above is based on a fixed 2Meg connection.

Sorry, don't know how to get error-figures on my Router.

I have noticed some drifting up & down on these figures & I can only assume that, when not-connected, it is because these stats have temporarily got even worse!

Regards,
John

[jid]

Well, 500m is about right for the Attenuation reading, you obviously live very close to the exchange!

With an attenuation of that figure, you would be expecting the top notch speeds:-
ADSL Link           Downstream          Upstream
Connection Speed        8128 kbps              448 kbps

Seems there is a serious line issue somewhere?

[JohnJBurness]

Seems there is a serious line issue somewhere?[unquote]

Yes, this has already been confirmed by BT Openreach Broadband Engineer!

As well as the obvious tests in & around my House/Equipment, he checked the Cabling from the Draw-pit (outside of my house) towards the Exchange & also from the Exchange towards the Draw-Pit.

He reckoned that I was cross-connected to another BB User (even though both my local End & the Exchange Equipment were disconnected), but this cross-connect is NOT showing up when a Telephone Engineer arrives to test the Telephony aspects.

It is the fact that I DON'T have a telephone fault that is giving me grief!! BT, after first formally accepting the Cable Fault, are now trying to wriggle out of an expensive Cable Replacement exercise by arguing that there is "No-Fault-Found" when using ONLY Telephony tests.

Regards,
John

[roseway]

The reason I asked to see your stats was because they would tell us the attenuation. The line length on its own means nothing.

With an attenuation of 10 dB you should be able to get the full speed which is possible, i.e. 8 Mbps for ADSL Max, or up to 24 Mbps for ADSL2+ if it is available on your exchange.

[JohnJBurness]

I realise that both current Attenuation & SNR Stats are required to be able to determine what my current faulty line can obtain, but I can tell that anyway by the extremely sluggish response that I get when trying to update the simplest of WebPages!!

(NB The BT Openreach Broadband Engineer confirmed that if was detecting a massive error-level on my line (sorry, don't have these figures).

Bearing in mind that my Noise-Margin is currently at 0.9dB & Kitz's Web-Site states:-
Downstream SNR Margin

Should be at least 10 dB to get adsl - ideally above 12dB to get traditional fixed adsl.

The only surprising thing is that I manage to get any form of connection, currently!!

As previously mentioned, trying to determine what I could get on a faulty line was NOT the point of my question - I was trying to establish what I could get on a healthy line.

I realise that, by the Stds of people with a much longer line, that 10dB Attenuation is good. What I didn't know was whether 10dB was good, bad or indifferent for a healthy sub-500m line!

Many thanks & regards,
John

[Oranged]

ADSL Link           Downstream          Upstream
Connection Speed        2272 kbps              288 kbps
Line Attenuation         10.0 db                      5.0 db
Noise Margin              0.9 db                9.0 db
The above is based on a fixed 2Meg connection.

Sorry, don't know how to get error-figures on my Router.

What router are you using.....Netgear (?) if it's holding a 0.9dB connection ?

Also which ISP's LLU are you on ?

[JohnJBurness]

As you have guessed, it is a Netgear (DG834PN, to be exact)

Am not on LLU, it is a bog-standard fixed 2Meg connection via PlusNet.

Regards,
John

[roseway]

Have you tried connecting the router to the test socket (behind the master socket faceplate)? If you get the same result when doing this then it's almost certainly a line fault, and your ISP should be pushing BT on your behalf to get it fixed.

[Ezzer]

Lowest snr i've seen with a continous connection of fixed speed is 5.5snr, I find any lower then there are constant drops of sync. In fact typicaly 6 or less should cause repeated loss of sync so I'm a bit curious as to how 0.9db could come about.

500m of cable lenght then yes you should certainly get 7616kbps minimum on a rate adaptive/max dsl line. at that distance you wouldn't be going through a cab/green box

[kitz]

Its difficult to "reverse engineer" stats as theres many different variables that could come into it.
However if your cable run is less than 500m, then I would expect the SNRM to be high.  On a 2Mb fixed rate line  of around 10dB attenuation, somewhere around the 30dB mark or just short of.

For a cable length of 500m attenuation of 10dB is about right.  Bear in mind attenuation is only a guide and can vary depending upon which router you use. Ive found some routers (eg the Netgear AR7 chipsets) may report slightly higher than others (Voyager Broadcom chipsets).

One thing to bear in mind though is that on some shorter lines the output power can be reduced to eliminate the possibility of CrossTalk, which will in turn reduce your SNRM.
When taking line stats, you have to look at the whole picture of all the figures.

As you obviously realise a SNRM of 0.9dB is not good on a fixed rate line.. you will likely be seeing dropped connections and lots of errors which will cause slower than expected throughput.
Can you just confirm if you are seeing any of these symptoms - just to ensure its not a faulty router reading.

~ Out of interest have you tried a "quiet line test" (Dial 17070 and select option 2).  If you can hear noise on that test, it may indicate a line fault which could be reducing your SNRM.

~ Another question - Is there any reason why youre still on 2Mb fixed, rather than Maxdsl?


Finally you may be interested in these stats taken from a 600m cable length when on a fixed 2Mb line from Plusnet.  When I say 600m - thats the true reading from BT when they used to give the figure out, and not any estimation..  direct its a hell of a lot shorter.

SNR (dB): 30.5 27.0
Attn(dB): 7.0 7.0
Pwr(dBm): 15.0 11.9
Max(Kbps): 10624 1144
Rate (Kbps): 2272 288

The above stats were taken from a Voyager router, but as mentioned above the attenuation does vary on different routers eg Speedtouch would show 9dB attenuation. 
That same line performed well on adslmax getting the full 8Mb and now gets 24Mb on adsl2+

[JohnJBurness]

Replying to the various points in turn:-

1) Use of Master Test Socket
Yes, already did that before reporting the fault 3 weeks ago. Additionally the BT Openreach Broadband Engineer did the same (using his Test-Gear), furthermore he replicated the problems out in the street Cable-Pit. Hence NONE of my Equipment/Wiring is a factor in determining the Fault - the ONLY factor that my Equipment can provide is (as Kitz refers to) is the variability of the reported reading that I get (from my Router) with that possibly provided by another Router or BT's own Test-Gear. Clearly that "variability" is NOT going to account for the massive difference between my "reported" SNRM & that which I should reasonably expect!

2) Possible Faulty Router
Fault has been replicated, out in the street (i.e. EVERYTHING within my house disconnected, incl BT Socket), on a "wires-only" basis, by the BT Broadband Engineer.

3) Cable Fault is NOT generating any symptoms onto Telephony service (No noise detected via 17070). This is part of the problem!! Having initially sent out a Broadband Engineer who confirmed it was an external underground Cable, BT then keep sending out Telephone Engineers who keep find "No Telephone Fault Found" (as if I didn't know)! 

4) MaxDSL
When MaxDSL came out, I noticed far too many people having far too many problems, hence I requested PN not to transfer my connection (I regard reliability as being more important than the latest "go-faster" connection). However, as well as trying to get my current fault fixed, I'm now considering (now MaxDSL has "matured") going over to it.

Regards,
John