Odd things about BE O2 conections

Any spare snr margin is not shown/hidden
TSNR values are the set to symmetrical values available only in 3db steps ie 3-6-9-12
But show in routers stats as 3.5 6.5 always an extra 0.5db
why are these stats the way they are with be /o2 and not the same with other isp's ,
Has it got something to do with the method used to connect (IPOA)that actually does not work as it was supposed to do ? BTW not trying to have a pop or anything remotely similar ,just curious really

In reply to a post by Anonymous:

Any spare snr margin is not shown/hidden
TSNR values are the set to symmetrical values available only in 3db steps ie 3-6-9-12
But show in routers stats as 3.5 6.5 always an extra 0.5db
why are these stats the way they are with be /o2 and not the same with other isp's ,
Has it got something to do with the method used to connect (IPOA)that actually does not work as it was supposed to do ? BTW not trying to have a pop or anything remotely similar ,just curious really

Any spare margin is used to up the sync speed.

I've always wondered this when I was on O2 but never got any explanation from the O2 gurus here.

In reply to a post by BatBoy:

Any spare margin is used to up the sync speed.

Like this one?

ISP Package............Orange 8M.........O2 8M

Downstream
Speedtest (bps).........7.1M................7.2M
Sync Speed (bps).......8061K..............8187K
Attenuation (dB)........27....................27
SNM (dB)..................22...................6.5
Power (dB)................22...................18

I don't think so!

In reply to a post by XRaySpeX:

In reply to a post by BatBoy:

Any spare margin is used to up the sync speed.

Like this one?

ISP Package............Orange 8M.........O2 8M

Downstream
Speedtest (bps).........7.1M................7.2M
Sync Speed (bps).......8061K..............8187K
Attenuation (dB)........27....................27
SNM (dB)..................22...................6.5
Power (dB)................22...................18

I don't think so!

yeah, it don't add up properly, even for those who can get a 24bit sync will only see what their target snrm no spare if they should have any, the upstream would be the more common out of the two, as a connection that i know used to give 8db spare margin with a sync of almost 1.3mbit before now gets just over 1.4mbit @ 6db no spare margin , attenuation is the same ???

In reply to a post by XRaySpeX:

In reply to a post by BatBoy:

Any spare margin is used to up the sync speed.

Like this one?

ISP Package............Orange 8M.........O2 8M

Downstream
Speedtest (bps).........7.1M................7.2M
Sync Speed (bps).......8061K..............8187K
Attenuation (dB)........27....................27
SNM (dB)..................22...................6.5
Power (dB)................22...................18

I don't think so!

See above

I've already seen! We are not talking about just 125K spare sync.

This is a line that does 16 Meg when it is uncapped from 8 Meg! As you should know from its attn.

Edited by XRaySpeX (Sun 20-Nov-11 21:03:28)

I can see from the power there is a line fault.

Are you talking about the two capped products? I.e.O2 Standard (legacy), and Be Value in both legacy and current versions?

Or something else, such as day-to-day running, where the margin varies as expected.

The 0.5dB immediately after sync I agrre is odd, bur seems to be a differnet, though perhaps related, issue.

What's TSNR? Or am I just having a brain failure?

In reply to a post by RobertoS:

Are you talking about the two capped products?

I know I posted examples above of two 8 Meg capped products, but the same findings held for the uncapped products, if you remember my Orange LLU vs. O2 experiment: http://forums.thinkbroadband.com/general/t/3887249-c...

TSNR = Target SNR Margin!

In reply to a post by XRaySpeX:

TSNR = Target SNR Margin!

TSNRM

In reply to a post by RobertoS:

Are you talking about the two capped products? I.e.O2 Standard (legacy), and Be Value in both legacy and current versions?

Or something else, such as day-to-day running, where the margin varies as expected.

The 0.5dB immediately after sync I agrre is odd, bur seems to be a differnet, though perhaps related, issue.

What's TSNR? Or am I just having a brain failure? T SNR = target SNR,
I was referring to both the legacy o2 packages and the be unlimited package,
but even if gdmt mode is forced there still is no spare margin visible , and that maxes out at around 12mbit which is a little high

In reply to a post by Anonymous:

In reply to a post by RobertoS:

Are you talking about the two capped products? I.e.O2 Standard (legacy), and Be Value in both legacy and current versions?

Or something else, such as day-to-day running, where the margin varies as expected.

The 0.5dB immediately after sync I agrre is odd, bur seems to be a differnet, though perhaps related, issue.

i'll try again, as my post , err well didn't show,???? TSNR =Target snr
the package is irreverent really, but should be more obvious if someone is on the capped @12mbit value product ,if their line can support a higher sync, but truth is only the tsnr is set to shows when is should show a higher (spare snr)

In reply to a post by Anonymous:

In reply to a post by Anonymous:

In reply to a post by RobertoS:

Are you talking about the two capped products? I.e.O2 Standard (legacy), and Be Value in both legacy and current versions?

Or something else, such as day-to-day running, where the margin varies as expected.

The 0.5dB immediately after sync I agrre is odd, bur seems to be a differnet, though perhaps related, issue.

i'll try again, as my post , err well didn't show,???? TSNR =Target snr
the package is irreverent really, but should be more obvious if someone is on the capped @12mbit value product ,if their line can support a higher sync, but truth is only the tsnr is set to shows when is should show a higher (spare snr)

Why?

In reply to a post by BatBoy:

In reply to a post by Anonymous:

In reply to a post by Anonymous:

... nested quotes trimmed ...

i'll try again, as my post , err well didn't show,???? TSNR =Target snr
the package is irreverent really, but should be more obvious if someone is on the capped @12mbit value product ,if their line can support a higher sync, but truth is only the tsnr is set to shows when is should show a higher (spare snr)

Why?

Because it has not been used for the sync and is "spare" and for our info too, so why not ?

When a line is syncing at the maximum speed on a capped product, and has an unnecessarily high SNR margin, the DSLAM can reduce downstream output power to minimise crosstalk interfering with other peoples lines. This is what appears to have happened with the stats that XRaySpeX has provided.

So why the diff. between the Orange & O2 behaviour? They were both LLU ADSL2+.

Because how you can configure a DSLAM is very varied

My BE connection is the same there both 3.5 SNR.

So basically, the answer to the OP is that it's that way because of how O2/Be configure their DSLAMs and other ISPs don't do it that way.

Yeah it seems like it

I'm sure it never used to be like this.

BE doesn't cap the value product by increasing noise margin hence the line stability capped is the same uncapped?

Edited by deleted (Tue 22-Nov-11 18:18:23)

In reply to a post by epyon:

I'm sure it never used to be like this.

It was in Sept. 2009, when the stats I posted here were applied.

Ohh i see

The router is also a factor as is different firmware on the same router

Very good point (which I'd long forgotten)!

The O2 stats I posted above were using the O2 Wireless Box 2.

The same O2 connection thro' other routers, within minutes of each other, gave similar syncs, attenuations and powers but much higher, more expected NMs: 17 dB with Orange/Siemens SE572 & 13 dB with D-Link DSL-2640R.

So it look like it's mainly down to the O2 router.

In reply to a post by nredwood:

The router is also a factor as is different firmware on the same router

How so? Is it the implementation of the cap or the reporting of the stats that is different? Does the bebox report stats correctly on a capped line as it would appear that reducing sync significantly by capping a line doesn't alter bit loading, snrm or error rates as one would expect?

My stats on O2 line capped to 8Mbps (line does approx 17Mbps uncapped). Netgear router.

ADSL Link Downstream Upstream
Connection Speed 8188 kbps 1087 kbps
Line Attenuation 20 db 4 db
Noise Margin 14 db 6 db

It certainly looks like the bebox (585v7) / dslam combination behaves oddly and doesn't optimise a capped line. I would say it is a bug. Time to get a Netgear.

In reply to a post by john2007:

My stats on O2 line capped to 8Mbps (line does approx 17Mbps uncapped). Netgear router.

ADSL Link Downstream Upstream
Connection Speed 8188 kbps 1087 kbps
Line Attenuation 20 db 4 db
Noise Margin 14 db 6 db

My netgear reports the Target sm even if i switch to the be value capped 12mbit mode when i know there would be spare snr showing with other isp's , strange that they would configure the dslam's in that way, i mean why bother doing that? i see no reason too,

In reply to a post by Anonymous:

My netgear reports the Target sm even if i switch to the be value capped 12mbit mode when i know there would be spare snr showing with other isp's , strange that they would configure the dslam's in that way, i mean why bother doing that? i see no reason too,

There is no reason, it is just poor engineering, throwing away overhead in noise margin for absolutely zero gain.

In reply to a post by Anonymous:

My netgear reports the Target sm even if i switch to the be value capped 12mbit mode when i know there would be spare snr showing with other isp's

What speed do you sync at anyway, and what other ISP offers an up to 12Mbps connection? (Other than perhaps a reseller of Be, in which case I would expect the Be behaviour in this respect).

Zero gain, have people looked at the power levels too?

Some CPE and DSLAM combos are good at backing off the power level, and if the noise environment requires it, then raising it again, before requiring a resync.

Keeping power levels lower can help avoid cross talk. Alas lots of CPE don't report it, or appear to have bugs in what they display.

It isn't power levels as the bit loading hasn't changed much.

I don't think it's a display bug either.

In reply to a post by LSDA:

It isn't power levels as the bit loading hasn't changed much.

I don't think it's a display bug either.

It's a line fault. That really is the SNRM and max sync at max power.

In reply to a post by BatBoy:

It's a line fault. That really is the SNRM and max sync at max power.

The line syncs at a higher speed uncapped than capped therefore your conclusion is false.

In reply to a post by RobertoS:

In reply to a post by Anonymous:

My netgear reports the Target sm even if i switch to the be value capped 12mbit mode when i know there would be spare snr showing with other isp's

What speed do you sync at anyway, and what other ISP offers an up to 12Mbps connection? (Other than perhaps a reseller of Be, in which case I would expect the Be behaviour in this respect).

mmm, using a reseller maybe , but via a wholesaler no, as they have there own network link to the be/o2 network pop's so can use pppoA my sync btw is 16mbps on pro 18.5mbps adsl2+ i also used to get a higher sync than i currently do, at a higher target snr & the attenuation is around the same @ +0.5 db just another observation the method that be use to connect or should do, (IPoA) although it can sync no traffic will flow(no internet) the difference in sync and throughput seems a little higher that my old pppoa connection

In reply to a post by LSDA:

In reply to a post by BatBoy:

It's a line fault. That really is the SNRM and max sync at max power.

The line syncs at a higher speed uncapped than capped therefore your conclusion is false.

You conclusion is based on minimal stats, therefore is wrong.

In reply to a post by BatBoy:

In reply to a post by LSDA:

In reply to a post by BatBoy:

It's a line fault. That really is the SNRM and max sync at max power.

The line syncs at a higher speed uncapped than capped therefore your conclusion is false.

You conclusion is based on minimal stats, therefore is wrong.

Your response is argumentative and illogical.

In reply to a post by LSDA:

In reply to a post by BatBoy:

In reply to a post by LSDA:

... nested quotes trimmed ...

The line syncs at a higher speed uncapped than capped therefore your conclusion is false.

You conclusion is based on minimal stats, therefore is wrong.

Your response is argumentative and illogical.

Your idea has already been shown to be false by another poster. Full stats, including tone graphs are required to indicate what is really going on.

In reply to a post by BatBoy:

Your idea has already been shown to be false by another poster. Full stats, including tone graphs are required to indicate what is really going on.

Who is this other poster?

I expect plenty of stats to be uploaded by posters in the coming weeks including my own. The tone graphs do reveal quite a bit.

http://forums.thinkbroadband.com/general/t/4067366-o...

In reply to a post by BatBoy:

http://forums.thinkbroadband.com/general/t/4067366-o...

It has been suggested that different routers behave differently, and I think that is a plausible explanation. However, looking more closely at John's stats, it looks like his line is noisy. The upstream sync is particularly poor at 4dB attenuation.

Edited by deleted (Sat 26-Nov-11 16:43:05)

As I said, line fault

In reply to a post by BatBoy:

As I said, line fault

Your response is illogical.

I'd prefer to call it inspired.

In reply to a post by BatBoy:

I'd prefer to call it inspired.

My stats on O2 line capped to 8Mbps (line does approx 17Mbps uncapped). Netgear router.

ADSL Link Downstream Upstream
Connection Speed 8188 kbps 1087 kbps
Line Attenuation 20 db 4 db
Noise Margin 14 db 6 db

That connection will almost certainly be interleaved, and depending on the inp setting this can also have an affect on sync speed boith upstream as well as downstream,
then there is the possibility that the router reports the attenuation incorrectly , but as said there could be a line fault but i think it is unlikely ,
Also when i forced a sync onto the capped 12mbit service, my upstream sync was also reduced by a few hundred kbps, so it could be down to the way be/o2 configured that profile

In reply to a post by Anonymous:

In reply to a post by BatBoy:

I'd prefer to call it inspired.

My stats on O2 line capped to 8Mbps (line does approx 17Mbps uncapped). Netgear router.

ADSL Link Downstream Upstream
Connection Speed 8188 kbps 1087 kbps
Line Attenuation 20 db 4 db
Noise Margin 14 db 6 db

That connection will almost certainly be interleaved, and depending on the inp setting this can also have an affect on sync speed boith upstream as well as downstream,
then there is the possibility that the router reports the attenuation incorrectly , but as said there could be a line fault but i think it is unlikely ,
Also when i forced a sync onto the capped 12mbit service, my upstream sync was also reduced by a few hundred kbps, so it could be down to the way be/o2 configured that profile

The point is that the SNRM was not adjusted as claimed, not that this connection has a line fault.

In reply to a post by BatBoy:

It's a line fault. That really is the SNRM and max sync at max power.

Your conclusion is also shown false by other posters here showing other routers reporting higher more normal SNRMs on the same O2 capped line, while the O2 router refuses to report anything but 6.5 dB..

Here an example of mine taken within minutes of each other:

O2 8 Meg LLU
==========
Router:......................D-Link..........O2
Downstream
----------
Speedtest (bps).........7.1M............7.2M
Sync Speed (bps).......8192K.........8187K
Attenuation (dB)........27...............27
SNM (dB)...................13................6.5
Power (dB)................17...............18

In reply to a post by XRaySpeX:

In reply to a post by BatBoy:

It's a line fault. That really is the SNRM and max sync at max power.

Your conclusion is also shown false by other posters here showing other routers reporting higher more normal SNRMs on the same O2 capped line, while the O2 router refuses to report anything but 6.5 dB..

Here an example of mine taken within minutes of each other:

O2 8 Meg LLU
==========
Router:......................D-Link..........O2
Downstream
----------
Speedtest (bps).........7.1M............7.2M
Sync Speed (bps).......8192K.........8187K
Attenuation (dB)........27...............27
SNM (dB)...................13................6.5
Power (dB)................17...............18

Full stats only, svp.

Grrrr, you really making me work:

Router:......................D-Link............O2
Upstream..
----------
Speed (bps)..............1129K...........1321K
Attenuation (dB).......16.................14
SNM (dB).................. 6.................. 7
Power (dB)...............11.................12

Are you still using your O2 connection then?

No, for 1 free month in Sept 2009.

That was 2 years ago. It's time to let go tbh.

It has been shown here that is still behaves this way.

You will clutch at straws. It has been proved empirically that the 6.5 NM is unique to an O2 router when used on an O2 capped connection.

• A diff router on an O2 capped connection reports a much higher NM.
• The O2 router on a diff capped connection reports a much higher NM.

In reply to a post by XRaySpeX:

It has been shown here that is still behaves this way.

You will clutch at straws. It has been proved empirically that the 6.5 NM is unique to an O2 router when used on an O2 capped connection.

• A diff router on an O2 capped connection reports a much higher NM.
• The O2 router on a diff capped connection reports a much higher NM.

Sorry, without full stats, what you are surmising is ridiculous - and two years old.

You've been supplied with full stats.

In reply to a post by XRaySpeX:

You've been supplied with full stats.

Where?

http://forums.thinkbroadband.com/general/t/4068395-r...
+
http://forums.thinkbroadband.com/general/t/4068421-r...

Those are not full stats.

Forget it.

Ohh its this thread again....

You have been banded before for being so stuck in the mud and unwilling to accept differences to your opinion.

Perhaps now is the time to simply accept there is more than one opinion on the issue

I have expressed my opinion, others have their own. In the absence of full stats it is impossible for me to contribute further. I consider the stats that are available insufficient.

And you keep repeating yourself, so time to walk away

In reply to a post by MrSaffron:

And you keep repeating yourself, so time to walk away

I already did

In reply to a post by BatBoy:

Forget it.

The following graphs provide some evidence of the bebox/dslam not optimising a capped line or some other bug.

Bebox (TG585v7 latest firmware) on a capped 12Mb line, attenuation 34dB

First graph: 6dB SNRM, sync 12Mb, power 19dBm

http://i44.tinypic.com/1ietmf.jpg

Second graph: 9dB SNRM, sync 12Mb, power 19dBm

http://i44.tinypic.com/15ezeq9.jpg

Increasing noise margin to 9dB reduces bits across all tones, however sync remains unchanged at 12Mbit, and therefore the line isn't optimised at 6dB . I believe this is what some people are seeing.

Edited by deleted (Sun 27-Nov-11 20:43:14)

Remember my Orange LLU vs. O2 experiment: http://forums.thinkbroadband.com/general/t/3887249-c...

Edited by XRaySpeX (Sat 16-Feb-13 21:35:38)

yeah can see the clear effect of extra signal power on orange.