Print

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

[waltergmw]

Gentlefolk and especially ezzer,

The pictures below are screen dumps of a 2Wire diagnostics page using the v 6 firmware, but on a Zen connection rather than a BT one.

I do appreciate that at a down attenuation of 79.4 dB @ 300 kHz it is something of a minor miracle that we have any connection at all.
However when the modem is in a good temper, usually when it's reasonably dry we sometimes achieve an IP profile of 350 kbps.
When the modem is in a bad mood, as it is now, in very damp conditions (no snow around Ewhurst yet) there is a big "hole" in the bit loading diagram between 267.375 and 323.4375 kHz or bins 62 to 75, but there are a few more bins above this group which do still provide 2 bits each. My question is can this data be used to diagnose what type of noise is affecting the line and is there any chance of improving the bit loading?

Kind regards,
Walter

[attachment deleted by admin]

[kitz]

>> there is a big "hole" in the bit loading diagram between 267.375 and 323.4375 kHz or bins 62 to 75,

>> My question is can this data be used to diagnose what type of noise is affecting the line and is there any chance of improving the bit loading?

From the look of the graph it doesnt seem to be anything other than general noise v signal quality loss over distance.  On long lines its not unusual to see a line of 0 bits then 2 bits then 0 bits then 2 bits towards the end of its range.

With adsl1 a minimum of 2 bits per bin is needed for that frequency to be usable.  If 2 bits cant be loaded then you get nothing in that frequency range.  
Its highly likely that bins 60ish onwards are hovering either side of just enough SNR and just not enough SNR to be able to use those later tones.
You also need to bear in mind that some tones may keep 'something in reserve' for bit-swapping.

ADSL2+ can make use of just one bit per tone which may give a lil bit more speed as on adsl2+ as you many find that the gap (or some of the tones in it) could load just 1 bit.... but bearing in mind you are talking about a maximum of 48kbps for those tones at that sort of bit loading...  which could in effect be cancelled out by some other differences in adsl2+ and the fact that it doesnt really seem to work as stable on very long lines.

Rather than concentrating on 60 onwards youd stand more chance of seeing an improvement by looking at overall quality of the line.   The tones around 32 are a bit more unusual that what that line sees at 60+.  However low bit loading indicates low SNR at those tones.. and on a line that length  it is going to be very susceptible to even the slightest amount of noise. :/


---------

btw power output of 15.2 dBm looks a lil bit low.  
When was the last full reboot? Is it always around that level?  
It may benefit from a retrain & channel analysis thats performed during a fresh sync.  Make sure you choose your time carefully though!

[orainsear]

Unless it's been moved, the pilot tone is in that hole.  Could that somehow be having an impact?

[kitz]

Ignore what I said last night about tone 32 - sorry wasnt quite with it - and what I said was based on years of looking at DMTtool graphs which easily separate the upstream from the downstream.      

The gaps at tone 32 and around there is simply the guard band between upstream + downstream.  The rest of the post still stands.

Good point orainsear about tone 64 which will explain one of the tones in that region...  but the pilot tone shouldnt have any effect on neighbouring tones.

[orainsear]

Is the downstream pilot tone essential for synchronisation?

It would appear that there is something preventing the bit loading across those bins, so it might be reasonable to assume that the interference area includes the pilot tone frequency.  Unless the pilot tone can be automatically moved, wouldn't that suggest that the pilot tone could effectively be missing, and if so what impact does this have?

[waltergmw]

@ Kitz and orainsear,

Thank you both for this most erudite discussion.
It's worth noting that I am relying on the diagnostics 2Wire provide but I have no data as to how the firmware reacts without a downstream pilot tone, or even if it still uses that tone in some single bit mode.
Might it also be possible that the pilot tone bit-swops to one of the lower empty tones in these circumstances ?
I assume it's using a pilot tone to calculate latency and some of the other reported data?

Kind regards,
Walter

[orainsear]

Are there any other modems that could be tried on such a long line, perhaps another TI AR7 chipset model?  I'm not sure if you will get anything better than the 2Wire though.

It may also be worth trying some different filters.

[kitz]

>> Is the downstream pilot tone essential for synchronisation?

Yes - its used during the sync stage.

>> Unless the pilot tone can be automatically moved,

Not that Im aware of for adsl1
The frequency is defined in the G.992.1 standards.

>> wouldn't that suggest that the pilot tone could effectively be missing

Theorising that the pilot tone can run with just 1 bit?  It wont need any error correction allowances or overheads. Its a signal and afaik many routers dont even display the SNRM for that frequency. A min of 2 bits means that there must be at least 6dB of SNRM at that tone during the sync up stage or it its marked unusable.  Its therefore likely that the pilot tone can operate at less.
  
Error correction (RS encoding) has some overheads that has to be allowed for regardless if interleaving is on or not and therefore min 2 bits for data transmission for adsl1.
This probably has something to do with s=1/2 mode  and its the reason why despite adsl1 technology having bins to support up to 12Mbps, 8128 is the max sync speed.
S=1/2 mode is used in later dsl technologies such as adsl2 to get the full 12Mb.  

As an aside, some (adsl1) dslams can now operate in s=1/2 mode as long as your router supports it  which effectively makes RS overheads more efficient which is why you sometimes see an interleaved line sync'd at more than 7616 (I suspect these are the Fujitsu geostream dslams which are used in many maxdsl exchanges).

I dont know enough about the pilot tone to comment much further Im afraid.


------
What is entirely possible is that line over the later stages could have SNRM that goes something like 6, 6, 6, 6, 5.5, 5.5, 5.5, 5.5, 6, 6, 6dB  which is what would cause that pattern towards the end.

[kitz]

>> Are there any other modems that could be tried

Something like a ST585v6 via the CLI would show you the actual SNRM at those particular channels which may give a bit more insight.

[waltergmw]

Hi again Kitz and Orainseer,

I've tried different filters, including an ADSL Nation faceplate which failed in use. The only other modem which worked at a lower throughput was a DG834V2 running DG Team firmware. I've tried a 585V7 and a Linksys to no avail. Here's some pictures I took covering the modem operation on 16 & 17 Dec where you'll see the hole has shifted down between 59 to 72. I think the bit lines give the appearance of colour as they are remote screen pictures of my laptop, rather than containing real information.

I know to my cost that I can't re-boot the modem remotely and stay attached, but when the weather's a bit better I'll re-sync it at about mid day.

Kind regards,
Walter

[attachment deleted by admin]

[waltergmw]

I've just had another look and strangely the bit loading is a little better even though its very dark and wet. We have a TBB speed of 233 kbps which seems quite good with an IP profile at 250 kbps.

Kind regards,
Walter



[attachment deleted by admin]

[attachment deleted by admin]

[kitz]

Whilst I could be wrong... but with the 'hole' shifting about around the latter regions I think that still probably points to the SNRM hovering around the just enough and just not enough 6 or 5.5dB type of scenario.

Something Ive never looked into at great depth is how 'real' SNR is converted into how much SNRM and then to bit loading.  All I know is that there is several complicated algorithms involved at this stage and its not a straight forward theres 'x' SNR so there will be 'y' SNRM and therefore 'z' number of bits loaded in any one particular channel.

Even on very good lines if you split down the tones, you will see a small amount of this castle effect of up and down over a range of tones.  Its likely to be the effect of just enough and just not enough SNR to load an additional..  and also the 'reserve' for bit swapping how ever that is calculated.  You just dont notice it at much though because its not down at the enough to load 2 bits or not stage.


---------------

As mentioned bit loaded may not always directly correlate to the amount of real SNR.

For example Ive just plucked out an entirely random mid range of stats on my own line where you can see this happening

Real SNR
tone 240: 42 41 42 41 41 41 42 41 41 41
Bits Loaded
tone 240: 12 12 12 13 12 12 13 12 12 13


The slight castle effect actually goes on all the way through my own stats aside from those at the beginning which can load the full 15 bits...  then as my line starts to reach its  limits you see it becoming a bit more pronounced

Real SNR
tone 500: 33 33 33 33 32 30 30 29 29 26
Bits Loaded
tone 500:  9  8  9  8  9  7  7  8  7  6

[waltergmw]

Thanks very much Kitz,

Today I've tried reset broadband = reestablish broadband link.
3 attempts so far and each is actually worse than it was to begin with this morning with fewer bins used.
I can't reboot the system unless I'm on site.

@ Roseway - (unless you've found a way ?) I can't over-write the factory defaults.
As has been demonstrated by an unfamiliar end user, poking about with passwords etc. can easily find the ISP user-name returns to greenlight services again -

This is a bad noise day and the modem is obviously feeling grumpy so I'll just have to wait for a nice sunny day!

Kind regards,
Walter



[attachment deleted by admin]

[roseway]

>> @ Roseway - (unless you've found a way ?) I can't over-write the factory defaults.

No, I'm afraid not.

[orainsear]

Interesting that the DG834Gv2 managed to connect - that's a TI AR7 based ADSL router (I also notice that according to the 2Wire stats that the DSLAM is using TI hardware).  Can the target SNR margin be set using DGTeam with the TI chipset?

On lines of this length it's a shame that the upper part of the normal upstream frequencies cannot be used for downstream purposes.