Speed Question

[Wobbly]

hi guys

IDNet newbie here, migrated over from Pipex on 10th December, came from a 2mb line so this max stuff is all new to me.

Yesterday my speeds were really off all day, I did a few speed tests, heres a couple:

Test1 comprises of Best Effort Test:  -provides background information.
    IP profile for your line is - 5500 kbps
    DSL connection rate: 448 kbps(UP-STREAM)  6816 kbps(DOWN-STREAM)
    Actual IP throughput achieved during the test was - 1213 kbps


Test1 comprises of Best Effort Test:  -provides background information.
    IP profile for your line is - 5500 kbps
    DSL connection rate: 448 kbps(UP-STREAM)  6464 kbps(DOWN-STREAM)
    Actual IP throughput achieved during the test was - 987 kbps

I run on an extension so I checked that, changed the splitters, read up on tweaking and changed my MTU, but it didnt make any difference. I dont know when this started, been away on biz during the week.

but this morning I do a check and

IP profile for your line is - 5500 kbps
    DSL connection rate: 448 kbps(UP-STREAM)  6560 kbps(DOWN-STREAM)
    Actual IP throughput achieved during the test was - 5152 kbps

Router stats (Netgear)

System Up Time 13:39:43
Port Status TxPkts RxPkts Collisions Tx B/s Rx B/s Up Time
WAN PPPoA 17063 23232 0 80 1159 05:52:02
LAN 10M/100M 196959 226400 0 3195 502 13:39:39
WLAN 11M/54M 36254 48614 0 771 81 13:39:30


ADSL Link Downstream Upstream
Connection Speed 6560 kbps 448 kbps
Line Attenuation 32 db 9.5 db
Noise Margin 7 db 16 db

The noise margin is higher this morning - usually down at 0 to 2 ?

Sooo, it cant be my end?, just wondered if anybody could guess at what the heck is going on - is the exchange still tweaking my line?

Cheers

[Gary]

Have a karma for a start,  I would expect to see variations in line speed due to bursts of noise etc which can cause issue as you probably have read in the faq as well as exchange work, even Christmas tree lights can cause it. Your noise margin looks better today and since you have been in a ten day training period the line speed and snr will vary as the system determines the most stable speed for your line, make sure you have no ring wire connected as well, thats all covered in the faq, but I would let your line settle down and see what happens, Rik will be along soon and will cast his wisdom over your stats

[Rik]

Hi Wobbly (love the name ), welcome to the forum.

I hate to be the bearer of bad news, but I think the problem is at your end - your line is clearly not performing as it should.

Your results this morning are spot on for your sync speed/profile, but it does look as if you may have a target margin of 9db, you can check this by seeing what the margin is immediately after re-booting the router. If this is the case, BT's line management software has increased the figure from the default 6db in an attempt to stabilise your line. Your sync speed is relatively low for your attenuation, so I think an increased target margin is likely. The fact that your three speed tests all have different sync speeds also indicates instability.

You say you are using an extension, is it flat or round? If flat, it's almost certainly picking up noise and adding to your problems. You may also have interleaving turned on, which would act as further restriction.

Enter this in the browser address bar:

http://192.168.0.1/setup.cgi?todo=debug

which should result in a screen which just says 'Debug enable'.
(That's assuming that you haven't changed the router's IP address from the default).

Then exit from the web interface and open a command-line window. Type:

telnet 192.168.0.1

You should get a BusyBox welcome message to confirm that your telnet connection is established.

Now type:

cat proc/avalanche/avsar_modem_stats

In those stats, look for an indication of fast path or interleaving - I can't remember exactly where (we forget so quickly) but it is there.

The variations in speed may be caused by contention at the exchange. IDNet guarantee no contention once you reach their network, but the BT part of the circuit is beyond their control. Even if the exchange is green, at peak times it's possible to see slow downs. However, it could also be caused by crosstalk or noise pickup. What you need to do is eliminate your internal wiring from the issue, and if the problem then persists, contact IDNet and get them to test the line and, if necessary, call out BT for you. Be aware, though, that BT will raise a charge of £160+ if they find that the fault is on your side of the master socket.

As Gary says, removing the ring wire is one of the first steps in this sort of problem, it can have almost miraculous results.

Let us know if we can help further.

[Gary]

The other thing is how long is your extension, the longer it is the more it degrades I think there is a limit to the run you can have possibly 30 meters but Im not positive on that. As Rik said flat wire will act like a giant aerial for interference, you really need cat5 shielded cable for long runs, but its best to avoid them if at all possible. Adsl Max is more sensitive to external issues than the old fixed 2mb line you had so you may have to do some internal adjustments. 

[Inactive]

Welcome Wobbly to the world of IDNet.



Have another welcome Karma.

[Wobbly]

Thanks chaps, very gratefull for you help

My connection is back down to sludge mode now, barely 1mb on speedtest.net

First the testing

I rebooted the router and this is what it said:

ADSL Link   Downstream   Upstream
Connection Speed   8128 kbps   448 kbps
Line Attenuation   32 db   9.5 db
Noise Margin   2 db   16 db

2) I did the debug and this is what I got:

usyBox v0.61.pre (2005.11.15-07:43+0000) Built-in shell (ash)
nter 'help' for a list of built-in commands.

cap proc/avalanche/avsar_modem_stats
ap: No such file or directory

cat proc/avalanche/avsar_modem_stats

R7 DSL Modem Statistics:
-------------------------------
DSL Modem Stats]
       US Connection Rate:     448     DS Connection Rate:     8128
       DS Line Attenuation:    32      DS Margin:              6
       US Line Attenuation:    19      US Margin:              16
       US Payload :            120768  DS Payload:             807312
       US Superframe Cnt :     15451   DS Superframe Cnt:      15451
       US Transmit Power :     0       DS Transmit Power:      0
       LOS errors:             0       SEF errors:             0
       Errored Seconds:        0       Severely Err Secs:      0
       Frame mode:             3       Max Frame mode:         0
       Trained Path:           0       US Peak Cell Rate:      1056
       Trained Mode:           3       Selected Mode:          1
       ATUC Vendor Code:       414C4342        ATUC Revision:  1
       Hybrid Selected:        3       Trellis:                1
       Showtime Count:         1       DS Max Attainable Bit Rate: 8896 kb
       BitSwap:                1       US Max Attainable Bit Rate:     n/a
       Annex:                  AnxA    psd_mask_qualifier: 0x0000
       ATUC ghsVid:  0f 00 41 4c 43 42 00 00
       T1413Vid: 00 00         T1413Rev: 00            VendorRev: 00
       ATUR ghsVid:  b5 00 54 53 54 43 00 00
       T1413Vid: 00 00 T1413Rev: 00    VendorRev: 00

       [Upstream (TX) Interleave path]
       CRC:    0       FEC:    0       NCD:    1
       LCD:    0       HEC:    0

       [Downstream (RX) Interleave path]
       CRC:    0       FEC:    0       NCD:    0
       LCD:    0       HEC:    0

       [Upstream (TX) Fast path]
       CRC:    1       FEC:    0       NCD:    0
       LCD:    0       HEC:    0

       [Downstream (RX) Fast path]
       CRC:    282     FEC:    0       NCD:    0
       LCD:    0       HEC:    0

ATM Stats]
       [Upstream/TX]
       Good Cell Cnt:  2516
       Idle Cell Cnt:  275018


       [Downstream/RX)]
       Good Cell Cnt:  16819
       Idle Cell Cnt:  5018383
       Bad Hec Cell Cnt:       60
       Overflow Dropped Cell Cnt:      0

SAR AAL5 Stats]
       Tx PDU's:       865
       Rx PDU's:       743
       Tx Total Bytes: 79688
       Rx Total Bytes: 699360
       Tx Total Error Counts:  0
       Rx Total Error Counts:  35


OAM Stats]
       Near End F5 Loop Back Count:    0
       Near End F4 Loop Back Count:    0
       Far End F5 Loop Back Count:     0
       Far End F4 Loop Back Count:     0
       SAR OAM Ping Response Drop Count=0


Next thing is to look at my socket re the ring thing

I am in an old house (100 years+), I suspect my external phone lines are a bit old, but I have had the intrenal socket replaced by BT whan we had a problem a few years back

The extension is about 10 metres I suppose - not sure what I am looking for to replace this extension with a better one and where to get it from?

the Mrs has put some Christmas lights up, I turned them off (she hasnt noticed yet) but no diifference

Here is a pic of my socket



What I dont understand is why it was flying at 8am and now its not

Anyway, maybe this will highlight something before I tackle the hardware

cheers

[Rik]

Hi Wobbly

The variations in your noise margin are a source of concern. Although Netgears do tend to vary, you're getting a 4db swing in a short space of time. The stats show that you do, indeed, have interleaving on, though you shouldn't need it for your line, so BT's software has seen enough instability to make the switch. It will be possible to get that turned off, but I don't think there's any point in trying until we solve the instability.

Clearly, you have a pre-NTE5 master, so you can't isolate your internal wiring to look for the source of the problems. How many BT-wired extensions do you have?

Any phone extension you use should be round (these cables use twisted pairs) not flat, but if your socket is only 10m away, could you move the router to the socket and run a long ethernet cable to the computer (ethernet cables can be up to 100m long in any segment, more importantly, on that side of the router, noise pickup is no longer an issue).

Finally, I would say that the variation in speed on a test site are often caused by load on that server. You might get a better picture of what's happening by downloading a large file, eg a Linux distro, and timing it.

[Wobbly]

thanks Rik

We have one BT wired extension upstairs (I presume its BT wired, it looks a proper job, it was here when we moved in 12 years ago). We have 2 cordless phones running off the upstairs extension and 3 on the downstairs, if that would interfere?

I have ordered a proper cable extension as per the spec, if that doesnt work I'll get a longer ethernet - moving the comp nearer atm is not an option wife wise with all the christmassy stuff thats going on

Thanks for the karma as well

[Rik]

The ring wire appears to be connected on the master, it's probably the orange/white wire in your photo, but I'm struggling to make out all the colours. You only need wires connected to terminals 2 & 5, and they should be a pair, eg blue/white and white/blue. Terminal 3 is used for the ring wire, and this should be disconnected at both sockets.

Are the phones DECT, and do you have one or two base units if they are?

[Wobbly]

Rik

better pic of downstairs socket



the downstairs phones are Dect, there is a base and 2 additional handsets

Upstairs,  there is one base station and one additional, I dont think these are dect, it doesnt say so

I turned these off just unplugged them) and no diffrence

I did another speedtest and it was about 350 download, less than upload!

My son is home for Christmas, connected by wifi, got him to turn off his laptop, no diffrence

I need karma, still cant understand why I was 5000 at 8am and now 350

[Rik]

Hi Wobbly

You do appear to have the ring wire connected on terminal 3. Ease this out with a pair of long-nose pliers, both at the master and at the extension socket. That will reduce noise pickup considerably. Take a look also at terminal 4, that may be the incoming exchange pair though - it's almost certainly on the wire at the back in your photo, the one with the brown/white wire.

The only reason that I can see for the varying sync speeds is noise. Have you looked in the Netgear's log for messages about loss of sync. My guess would be you are going to find a bunch of them.

It might be worth downloading Routerstats:

http://www.vwlowen.co.uk/internet/files.htm

and running that. The graph will give you a good idea of what's going on.

[Wobbly]

Rik

OK I have d/l the routerstats and will run that for some data

I did a reset on my router so lost the log, but you are right, I have had quite a few loss of synchs when I looked at it before

I'm loath to mess about with the back of the socket atm, I will have to think about that - I have a proper extension cable on order and may wait and see if that helps

Many thanks for your help, greatly appreciated, I have to go out now so will leave the router stats running

[Rik]

Bear in mind that the quality of the line is affected by the ring wire before you install the extension, ie adding a good extension to a noisy circuit will not improve things.

[Lance]

I would really recommend removing the ring wire. It serves no purpose these days, other than to cause problems on ADSL!

[Wobbly]

OK ring wire removed on both sockets

how long should it take to have an effect?

[Lance]

The effect will be immediate. Basically, if it is going to make a difference, you will either see a highed noise margin, or increased sync.

It doesn't always make a difference, but removing it eliminates it. I have seen several reports of massive sync speed gains from removing it in the past.

[Rik]

What Lance said, Wobbly, Any changes to the profile will take the normal 3-5 days though.

[Wobbly]

before ring removed



after

[Wobbly]

 

http://www.thinkbroadband.com/speedtest/results/id/119842442457821012071.html

[Rik]

You can see a re-sync there at 14:57, Wobbly, when the graph drops off the bottom of the page. Also, you have a target noise margin of 9db, so BT does regard you line as unstable, as has started to lift the target margin in an attempt to stabilise it.

The best thing I can suggest at the moment is to de-tune an AM radio, and see if you can identify the source of the noise by walking about with it.

There's not a lot of point in worrying about speed until the stability is resolved, tbh.

[Wobbly]

ok will give that a go

[Sebby]

There's something that causes surges of noise. You'd expect the noise margin to move, but only maybe 0.5-1.0dB every so often. Yours looks a bit too up and down. I think the AM radio suggestion is your best bet at the moment, as the removal of the ring wire hasn't made a great deal of difference, if any.

[Rik]

Wobbly is still using an extension lead which may be implicated, Sebby, but it's really now about trying to find that noise source rather than risk a £160+ bill from BT.

[Sebby]

So he does, Rik. Unless it's cat5, that's going to pick up interference like crazy. Definitely try a shorter cable, Wobbly, even if it's just for a moment to see if that helps.

[Wobbly]

I tried the radio thing, definitely a change in tone next to the downstairs socket, a bit of a change on the upstairs one, only other interfernce change was on the tv

noise margin seems to have decreased, thats bad isnt it?