Print

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

[jid]

Hi all,

As of Friday I have now fully moved to TalkTalk it seems.

Phone has had caller display for a while and Friday the connection resynced with ADSL2+

Posted some stats for you to take a look at, seems some form of DLM is running at the moment as well and there is a cap of 8191kbps i'd say.

Regards

Jamie

[attachment deleted by admin]

[GunJack]

Nice one !!

I still can't get any idea of timescales for my exchange's shift over to the TT kit....so still contemplating paying £40 for another line to be connected so I can go with sky, who've recently LLU'd my exchange, for ADSL2+ immediately..

must say, the bit-loading looks good over 1000kHz, does DMT auto-scale for that, or it it a manual setting ??

[jid]

I have noticed they Run DLM on the line though.

I've just gone as it seems from a 9dB SNR Target down to a 6dB. 11001kbps Downstream Sync!! Impressive!



I also had no idea of the timetable but the phone service seemed to go over first as the 1471 voice had changed, broadband didn't change for around a week to a fortnight.

[GunJack]

Well....they must have been having a spate of doing exchanges recently....

as of 05:30 this morning, I find that I've been switched over also
However, not got the increase I was hoping for...in fact, virtually no difference from adslmax

Am going to have to have a play round with it to see if I ca nget the extra 1-1.5 meg I was hoping for. One thing that concerns me, though, it the bitloading graph...it is very light top-end. My parents are on the same exchange but a longer line, and when I saw this characteristic on ther line, I thought it was just down to the longer line, now I'm not so sure. Is it possible that TT have restricted the higher tones on the DSLAM/MSAN ? Or possibly my router (they have the same) doesn't like 2+ even though it's a 2+ router ??

Any ideas anyone ??

p.s. I use a D-link DSL2640B, as it suited my line best. Just tried with a 834GT and the bitloading is the same, so unless the 6348 bcm chipset doesn't like 2+, or it's TT's implementation on the line

[attachment deleted by admin]

[waltergmw]

Hi Gunjack,

Remember "longer line" is BT speak for total line characteristics including higher frequency losses which could include higher aluminium alloy impedance etc. etc.
If I remember correctly Kitz does say that ADSL2+ can produce poorer performance on "longer lines".

Kind regards,
Walter

[GunJack]

Walter,

thanks for taking the time to reply. I understand what you're saying, however there are a combination of factors that led me to believe maybe there's something up with the actual dsl from the exchange....

1. D/S Attenuation has increased by 3dB - as expected

2. Line was slightly out-performing the kitz estimator on Max, so think it should be of reasonable quality

3. The bitloading graph looks too clean a cut-off instead of degrading to nothing, as if a low-pass filter has been put in-circuit....will have to check where the bit-loading falls off when I'm on G.DMT

4. Although my parents are off the same exchange, they are half a mile or so away and off a different pcp, and also different e-side (according to OR bod who fixed my water ingress fault a few weeks ago), yet their connection is showing pretty much exactly the same characteristic


I think it's the combination of 3 and 4 which are making me wonder....and I'm even tempted to call the "tech" support line to get them to check it (I can't use the TT members forum, coz I'm ex-tiscali  )

..or am I just being paranoid 
 

[GunJack]

p.s. having just checked, it looks like it IS only using adsl1/2 frequencies (up to 1.1MHz), not full 2+ (up to 2.2MHz), even though router and DMT are reporting 2+ 

[jeffbb]

Hi
can you post your full stats ? Might be a clue there

Regards Jeff

[GunJack]

jeff, as requested....

Mode: ADSL2+  
Line Coding: Trellis On  
Status: No defect  
  
  Downstream Upstream
Rate (Kbps): 5998  980  
  
SNR Margin (dB): 7.3  9.1  
Attenuation (dB): 50.0  23.7  
Output Power (dBm): 12.9  0.0  

  
Super Frames: 2137159  1903843  
Super Frame Errors: 81  4294967268  
RS Words: 115406616  0  
RS Correctable Errors: 15444  4294967288  
RS Uncorrectable Errors: 727  N/A  
  
HEC Errors: 60  2  
OCD Errors: 0  0  
LCD Errors: 0  0  
Total Cells: 484850450  402762  
Data Cells: 2323435  1850  
Bit Errors: 0  206  
  
Total ES: 56  4  
Total SES: 0  0  
Total UAS: 102  0  

looks reasonable to me, am I missing something ?? upstream power at zero is a reporting error i believe. max stats were

down    up
5696    896
8.5dB   11dB
47dB     24dB

[GunJack]

Further interesting observations.....

my snrm has spent most of the evening and night down around 5-5.5dB...not unusual, you may think. However, on max this would have led to error flooding and the line being unuseable....but I'm only looking at a few hundred CRCs since around 7:00am yesterday.

Don't mind that sort of improvement, and wondering now if a reduction in the interleaving may be appropriate to try?? Was so used to how the line behaved on max, this is a whole new ball-game  

Throughput upand down is almost exactly the same as on max

Oh, and what of the TT DLM I've re-synced god-knows how many times since the switch...anyone any ideas how detrimental this may be ?? Although, now it's been through an evening/nighttime period without any apparent issues, I'm inclined to leave well alone...current stats are:-

Mode: ADSL2+  
Line Coding: Trellis On  
Status: No defect  
  
  Downstream Upstream
Rate (Kbps): 6051  968  
  
SNR Margin (dB): 6.4  7.9  
Attenuation (dB): 50.0  23.7  
Output Power (dBm): 12.9  0.0  

  
Super Frames: 2909958  2605168  
Super Frame Errors: 174  4294967200  
RS Words: 159320212  0  
RS Correctable Errors: 38729  4294967090  
RS Uncorrectable Errors: 1778  N/A  
  
HEC Errors: 137  0  
OCD Errors: 0  0  
LCD Errors: 0  0  
Total Cells: 669347157  1207050  
Data Cells: 6805786  1092  
Bit Errors: 0  0  
  
Total ES: 172  0  
Total SES: 0  0  
Total UAS: 442  0  

As mentioned before, think the upstream is just a reporting/compatability issue, rather than actual errors/no power....

[jeffbb]

Hi
This seems OK The DS is as expected or even slightly better  . Unfortunately throughput is not only related to synch speed . As we are all aware the dreaded contention word can be a  limiting factor . So even  an increase in synch rate does not always translate into an improved throughput .While there were few people on ADSL2+, improvements in throughput could be seen even on previous lines (ADSL1) that had high contention  (because of the unused  higher frequencies availability), The more people that transfer to ADSL2+ the more contention there will be , unless more bandwidth is made available .

Regards Jeff

[GunJack]

Thanks jeff, appreciated  

One thing confused me though, your statement of ds being bang on or slightly better than expected. Using my max attenuation in the kitz estimator, I get a possible ds of 6400k, so I was thinking I was a bit down, especially considering as I said earlier that I was getting slightly better than the estimate for max... have I missed a factor out somewhere ?? e.g. interleaving depth more of a factor on 2+than on max?? (it's the same now, btw, as before - I have asked for it to be dropped down to see if it has an effect)

..like I said, whole new ball-game for this line, gonna take time to get used to it again  


...and there's still the TT DLM to suss out ...... 

[jeffbb]

Hi
Your current attenuation is 50db even allowing a couple of db* to allow for the higher frequencies in use (the calculator is based on adsl1 )gives an estimate of <6000kbps with the default 6db target.
You may have had a better than average synch rate on adsl1 but the higher frequencies may not benefit from the  low noise you experienced .
* I see you used a 3db allowance
The only point that seems odd is  Output Power (dBm): 12.9 that does seem low .

Interleave will NOT make a difference to your synch rate .It just affects your Latency
Regards Jeff

[GunJack]

http://www.speedtest.net/result/974771528.png

Jeff, thanks again for taking the time out to reply. I 've had a little help from the ex-tiscali forum peeps, and I'm now the proud owner of a 6dB profile as of 20:30 hrs this evening (was on 9dB), and it's magic  Curent stats are:-

6338k/4.5dB down
1020k/6.5dB up

and the above speedtest result. More like I was expecting. Thing is, no build-up of CRC errors - on adsl1, much below 6dB in the evening snrm dip, and I'd start to flood a bit. Now, it was as low as 4.2dB earlier on, so if this holds (as it looks like it will, the dip troughs out between 21:30 and 22:45ish), I'll be a happy bunny  

p.s. AND they've turned the DLM off for me !!!!!

p.p.s. there seems to be a compatability issue between the bcm 6348 chipset in my 2640B router and the DSLAM/MSAN, so some of the stats have errors in them..can't see me having 0dB power o/p downstream somehow (the router interface swaps upstream and downstream power for some reason...on DMT it's deffo downstream that's 0)

full new stats are:-

Mode: ADSL2+  
Line Coding: Trellis On  
Status: No defect  
  
  Downstream Upstream
Rate (Kbps): 6338  1020  
  
SNR Margin (dB): 4.3  6.8  
Attenuation (dB): 50.0  23.6  
Output Power (dBm): 12.5  0.0  

  
Super Frames: 924710  822951  
Super Frame Errors: 85  4294967285  
RS Words: 52708518  0  
RS Correctable Errors: 24894  0  
RS Uncorrectable Errors: 881  N/A  
  
HEC Errors: 70  0  
OCD Errors: 0  0  
LCD Errors: 0  0  
Total Cells: 221442433  1741362  
Data Cells: 9265426  7656  
Bit Errors: 0  0  
  
Total ES: 366  0  
Total SES: 1  0  
Total UAS: 474  0  


edit 08:10 Sat morning....overnight DMT trace attached....wow !!!

[attachment deleted by admin]

[jeffbb]

Hi

quote : I'm now the proud owner of a 6dB profile as of 20:30 hrs this evening (was on 9dB)

I thought you said you were already on 6db,as all your stats showed about 6db

Careful scrutiny of the bit loading graph (as far as I can) There does not seem to be any real difference in the bit loading . If anything the original one appears to be slightly better . see attachment

The difference in your synch rate

Rate (Kbps): 6051  968   and
Rate (Kbps): 6338  1020 

can just be that you resynched at a GOOD time .
quote  edit 08:10 Sat morning....overnight DMT trace attached....wow !!!

is that a reference to the SNR/tone fluctuation?

regards Jeff

[attachment deleted by admin]