Router vs. Telco DSL Quality Stats

[Jon.R.Kibler@gmail.com]

Hi,

We have been having an on-going battle with several local telcos over
DSL line quality. For some reason their testers always show a much
higher line quality than does the routers.

For example, I have taken the exact same DSL POTS cable, and plugged
it into a Cisco 827, 837, and 877 router and got essentially the same
line quality stats. When the local telco test the line (usually using
a SunSet MTT test set), they consistently see a good quality line,
where the routers see a marginal line -- one that keeps dropping. (And
this is not just a single line at a single location -- we have the
same problem at multiple locations, and at some locations, on multiple
lines at that location.)

For example, here is what the router reports:
ATU-R (DS) ATU-C (US)
Capacity Used: 98% 53%
Noise Margin: 5.0 dB 12.0 dB
Output Power: 17.0 dBm 8.0 dBm
Attenuation: 64.0 dB 31.5 dB
Interleave Fast
Interleave Fast
Speed (kbps): 1216 0
256 0

and the test set reports:

Capacity: 47% / 40%
SNR: 8.5dB / 15dB
Attenuation: 40dB / 28dB
kbps: 1472 / 256 (noise profile)

Why such a substantial disagreement between telco test sets and Cisco
routers? Especially when there is no difference between the wiring to
the device, up to and including the cable plugged into the device.

This is getting to be a real pain. We have flaky connections and
numerous drops, yet the telco says everything is fantastic. It just
doesn't make sense.

TIA for any insights into this problem.

Jon K.

[LouB]

Jon.R.Kibler@gmail.com wrote:
> Hi,
>
> We have been having an on-going battle with several local telcos over
> DSL line quality. For some reason their testers always show a much
> higher line quality than does the routers.
>
> For example, I have taken the exact same DSL POTS cable, and plugged
> it into a Cisco 827, 837, and 877 router and got essentially the same
> line quality stats. When the local telco test the line (usually using
> a SunSet MTT test set), they consistently see a good quality line,
> where the routers see a marginal line -- one that keeps dropping. (And
> this is not just a single line at a single location -- we have the
> same problem at multiple locations, and at some locations, on multiple
> lines at that location.)
>
> For example, here is what the router reports:
> ATU-R (DS) ATU-C (US)
> Capacity Used: 98% 53%
> Noise Margin: 5.0 dB 12.0 dB
> Output Power: 17.0 dBm 8.0 dBm
> Attenuation: 64.0 dB 31.5 dB
> Interleave Fast
> Interleave Fast
> Speed (kbps): 1216 0
> 256 0
>
> and the test set reports:
>
> Capacity: 47% / 40%
> SNR: 8.5dB / 15dB
> Attenuation: 40dB / 28dB
> kbps: 1472 / 256 (noise profile)
>
> Why such a substantial disagreement between telco test sets and Cisco
> routers? Especially when there is no difference between the wiring to
> the device, up to and including the cable plugged into the device.
>
> This is getting to be a real pain. We have flaky connections and
> numerous drops, yet the telco says everything is fantastic. It just
> doesn't make sense.
>
> TIA for any insights into this problem.
>
> Jon K.

The telcos are lying to save money. Escalate the complaint.

[Merv]

you should check that you have the latest DSL firmware loaded

see Cisco doc "IOS Software Release-to-DSL Firmware Version Mapping on
Cisco Access Routers"

http://www.cisco.com/en/US/prod/coll...d801d5c4c.html

[Bod43@hotmail.co.uk]

On 3 Apr, 10:12, Merv <merv.hr...@rogers.com> wrote:
> you should check that you have the latest DSL firmware loaded
>
> see Cisco doc "IOS Software Release-to-DSL Firmware Version Mapping on
> Cisco Access Routers"
>
> http://www.cisco.com/en/US/prod/coll...1/prod_bulleti...

There are free ftp downloads of firmware for the 8[57]7.
ftp://ftp.cisco.com/pub/access/800

I have found them to be useful in a number of cases but since the
IOS bundled firmware went to 3.x I have not bothered.

Certainly with the 857 and 877 I have seen numerous issues
that firmware changes have resolved.

This has been in the UK.

[Jon.R.Kibler@gmail.com]

Thanks to all. I will try some firmware upgrades this weekend.

I guess that leaves me with the question of how the firmware can
effect apparent line quality? It seems that line quality should be
independent of the chipset/firmware used. Can someone please explain
this?

Again, THANKS!

Jon Kibler

[Hans =?iso-8859-1?Q?J=F8rgen?= Jakobsen]

On Thu, 3 Apr 2008 08:13:13 -0700 (PDT), Jon.R.Kibler@gmail.com wrote:
> Thanks to all. I will try some firmware upgrades this weekend.
>
> I guess that leaves me with the question of how the firmware can
> effect apparent line quality? It seems that line quality should be
> independent of the chipset/firmware used. Can someone please explain
> this?

If your DSL vendor has a list of recomended modems follow that or
you are on your own!

Not all combinations chipset/firmware work equaly well.

When my local telco do a full test they are testing at all cable
lengths in 50meter(150feet) incriments. And they have found "funny"
bugs.
/hjj

[Who Me?]

<Jon.R.Kibler@gmail.com> wrote

> I guess that leaves me with the question of how the firmware can
> effect apparent line quality? It seems that line quality should be
> independent of the chipset/firmware used. Can someone please explain
> this?
>

If we forget for the moment that the term "line quality" is itself somewhat
ambiguous...........

The chipset/firmware may not be capable of making full use of a "good" line.
Or there may be "glitches" in the good line.
Or the equipment at your end might not be fully compatible with what the ISP
has at his end.
Or the equipment at your end might just be "bad".

Not too complicated really.

And to echo a previous post: If you insist on using equipment that is not on
your ISPs approved list, then you're on your own.......and it doesn't sound
like you are qualified to take up that challenge.

[Franc Zabkar]

On Wed, 2 Apr 2008 11:30:16 -0700 (PDT), Jon.R.Kibler@gmail.com put
finger to keyboard and composed:

>Hi,
>
>We have been having an on-going battle with several local telcos over
>DSL line quality. For some reason their testers always show a much
>higher line quality than does the routers.
>
>For example, I have taken the exact same DSL POTS cable, and plugged
>it into a Cisco 827, 837, and 877 router and got essentially the same
>line quality stats. When the local telco test the line (usually using
>a SunSet MTT test set), they consistently see a good quality line,
>where the routers see a marginal line -- one that keeps dropping. (And
>this is not just a single line at a single location -- we have the
>same problem at multiple locations, and at some locations, on multiple
>lines at that location.)
>
>For example, here is what the router reports:
> ATU-R (DS) ATU-C (US)
>Capacity Used: 98% 53%
>Noise Margin: 5.0 dB 12.0 dB
>Output Power: 17.0 dBm 8.0 dBm
>Attenuation: 64.0 dB 31.5 dB
> Interleave Fast
>Interleave Fast
>Speed (kbps): 1216 0
>256 0
>
>and the test set reports:
>
>Capacity: 47% / 40%
>SNR: 8.5dB / 15dB
>Attenuation: 40dB / 28dB
>kbps: 1472 / 256 (noise profile)
>
>Why such a substantial disagreement between telco test sets and Cisco
>routers? Especially when there is no difference between the wiring to
>the device, up to and including the cable plugged into the device.
>
>This is getting to be a real pain. We have flaky connections and
>numerous drops, yet the telco says everything is fantastic. It just
>doesn't make sense.
>
>TIA for any insights into this problem.
>
>Jon K.

Apart from the downstream attentuation, it seems that your two sets of
readings are within agreement, there being a consistent 3.5dB
difference which I suspect may reflect different refence points (?).
The big 64dB versus 40dB discrepancy is a worry, though. Could there
be a 24dB loss within your premises? (but wouldn't that affect the
output power also???)

FWIW, here is my ISP's explanation as to factors that may affect
performance:
http://www.internode.on.net/resident...ernet/extreme/

There is also a "how fast can it be?" graph that charts speed versus
attenuation.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.