Print

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

[terryc]

@kitz - many thanks for your explanations and time. As a matter of interest I have now got the Telnet working on router stats and it shows the following -  I don't know if that gives any further clues?

DownStream

Connection Rate:  4544
Line Attenuation: 45
Noise Margin:     10

Interleave CRC:   940
Interleave FEC:   158332
Interleave HEC:   0

Fast path CRC:    0
Fast path FEC:    0
Fast path HEC:    0
=========================
Upstream

Connection Rate:  448
Line Attenuation: 29
Noise Margin:     21

Interleave CRC:   8
Interleave FEC:   11560
Interleave HEC:   68

Fast path CRC:    0
Fast path FEC:    0
Fast path HEC:    0

=====================

Upstream (Tx):
  Good Cells: 126315
  Idle Cells: 180876151

Downstream (Rx):
  Good Cells: 683893
  Idle Cells: 1835186855
  Bad HEC Cells: 10943
  Overflow Dropped: 0

Bytes Transferred (PPP):
  Tx: 3115653 (2.9 Mb)
  Rx: 30409367 (29.0 Mb)

Interleaving appears to be on

The router uptime is just over 47 hours since I last re-connected following another BT "unplug and try" test.

Terry

[graevine1]

terryc go into routerstats and mark up the sample rate to every 5 seconds and also set it to record especially the noise and crc and FEC results set also the date to be automatically recording a new log each day so you have each day on record. Like I was receiving it looks as if you have some specific noise pulses. If you tie the graphing together you can see if all the peaks and troughs occur at the same time.

[kitz]

The prescence of CRC errors indicate that there are periods of time when you do get some interference on your line. 
FEC errors are errors corrected by interleaving.
CRC errors - well that more or less depends when they rack up - if you have a lot in a short space of time then its indicative of a short sharp noise spike..  or they could be slowly racking up over a period of time.
The figure is not unduly large and many lines cope quite fine with a similar amount of errors if they were racked up over time.  If there was a shorter sharper burst then you may have noticed your connection being a bit "slow" during that period.
The FEC errors show that interleaving is working well on your line, and its doing what its designed to do.


As mentioned in my previous posts.. continue to monitor your SNR with routerstats.
If that figure gets too low then you know theres something odd going on.

[kitz]

My response to Roseway, if what you are attempting to perport was true then some of the basic laws of Physics would have to be re-written. The example I must site would be a line of 5 kilometers using your analagy would give a line attenuation of 20db per kilometer over say 5 kilometers the result 100db of line loss. You therefore will be able to supply a broadband signal of usable level to all peoples in all corners of the UK with little difficulty.
My figure gives approx 50db for a five (5) kilometer length of line attenuation which is the reason why broadband will function up to approx 65db and six(6) kilometers. QED!

Please take this debate else-where so as not to detract from the OPs question.  This site is designed to help all users regardless of their ISP.

Ive started a new thread here. Loop Loss - dB's per km

[terryc]

I just thought I would post the latest.  On Monday afternoon after no call backs from BT's THD and no change to speeds I compiled a short but factual  (and polite) letter of events (in essence a compaint!) and emailed it directly to the CEO's office. Within a short time I had a call from BT Resolutions with a case number. After gathering more details the chap said not to expect to hear from him for at least three days (here we go again I thought!).

Anyway last night downstream increased to 7392 and this morning its at 7520 and interleaving has been turned off.  Now I doubt for one minute that these will be sustained on a 45db line as the SNR hovers around 3db but at least my Netgear seems to hold on at these levels. So if the target SNR does increase to sync drops back to 6.5 or so that will be fine.

I don't know what they've done as I haven't had a call back yet I suspect like me they are waiting to see how it holds up.  But it just goes to show with a little measured patience there are ways to get results. I've posted the current error rates below  - not sure how these look to the trained eye?

[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:    4       FEC:    145     NCD:    0
        LCD:    0       HEC:    0

        [Downstream (RX) Fast path]
        CRC:    367     FEC:    0       NCD:    0
        LCD:    0       HEC:    0

[roseway]

That's looking good. It's a shame that you had to go to the top to get a result, but it looks like a victory for politeness.

[mr_chris]

How long after reconnecting did you grab those stats? If that was overnight with no resyncs its looking pretty good, as eric said.

Incidentally, how much have you been using the connection in this current session too?

[terryc]

@Mr-Chris -  not long - there had been a couple of re-synchs overnight and I haven't been doing much online.  I'll see what happens with more usage over the next few hours  -  I suspect the CRC error will increase. Incicentally BT called this morning and left a message, didn't say what they had done but a are sending a brand new home hub and phone by way of compensation.  However, I think I may stick with the Netgear or does anyone know better? I guess I can take advantage of the Softphone facility which I never have to date.  I understand the HH is not particularly secure though?

I'm also moving to an iMAC from a PC/XP so I'm not sure if that should influence the choice?

Thanks All

Terry

[kitz]

Thanks for the update

As chris said need a bit more info really before we can judge the error situation.

However in view of what you have said I dont think I can make my mind up if they have reduced your target SNR to 3dB (which is seldom done)... or reset your line for training (in which case target snr would be 6dB and interleaving off).

The fact that you said your snr hovers around 3dB is the _only_ reason I suspected the former..   the best way to tell would be looking at your line stats immediately after a resync.

As you rightly say the current sync may be just a tad too high for the line length and it may well drop later on as the SNR decreases.  Theres a few more things I could say here... but it would probably be best just to observe what the line does itself over the next 24 hours first.

Re the HH - by all means you could give it a go if you like..  but unfortunately their reputation isnt that good on some of the more troublesome lines.  In fairness there did appear to be one particularly bad batch at one time.
Personally if I were you though Id stick to the netgear whilst your line is re-adjusting.