dslreports logo
 
    All Forums Hot Topics Gallery
spc
Search similar:


uniqs
907

sm5w2
Premium Member
join:2004-10-13
St Thomas, ON

sm5w2

Premium Member

Question about Bell DSL line stats

A friend in Windsor with Bell DSL (sympatico HSE) was getting between 3 and 3.5 mbps download according to speedtest.net. His modem was 56xx or 65xx speedstream.

I brought down an alternate modem (TP-Link TD-W8961N) and was seeing these line-stats:

-----------
Upstream:
SNR Margin: Always 10 db
line attenuation: Always 31 db
data rate: always 800 kbps
max rate: always between 1196 and 1108 kbps

Downstream:
Data rate is always 4032 kbps

When incoming line is connected directly to modem (and disconnected from other lines in the house):
line attenuation: 43/43.5 db
noise margin: 7/9 db
max data rate: 4600 / 5000 kbps

When incoming line is connected directly to modem (and then connected through filter to other lines in the house):
line attenuation: 48/48.5 db
noise margin: 15.5/16 db
max data rate: 6784 / 7168 kbps
-----------------

So basically the modem is directly connected to the incoming line at the demarc, and the rest of the home's phone lines were alternately dis-connected completly, or re-connected through a filter.

When just the modem is connected, downstream noise margin falls to 7 or 9 db, line attenuation drops to 43 db, and max data rate is somewhere near or just below 5 mbps.

When the rest of the home's phone line distribution is re-connected at the demarc through a filter, noise margin goes up to about 16 db, line attenuation increases to 48 db, and max data rate goes up to about 7 mbps.

No change in the upstream numbers.

I find it strange that connecting the rest of the home's phone wiring to the demarc (through a standard DSL filter) would improve the noise margin (while also increasing line attenuation?) but still over-all improving the max possible data rate from 5 to 7 mbps.

The current profile (4032/800) seems locked for this line.

I know it was better in the past, so I think I'm going to open a ticket in the direct forum and see if they can fix this. I'm assuming I can leave the TP-link modem in place (it will auto-negotiate the best possible connection, probably do a better job than the ancient speedstream).

The aerial drop to the house was replaced recently, but I suspect they can connect it to a better circuit in the pad or what-ever that cylinder-thing is that hangs from the phone line.

But how do these line stats look as-is, and why should they improve when connecting the rest of the telco system to the phone line vs just having the modem connected directly by itself?
btech805
join:2013-08-01
Canada

btech805

Member

If it is connected to a pots filter that could explain the improvement in the attainable rates with the inside wiring connected. As for the speeds, the profile is obviously set and won't change unless you request it. I am assuming he is paying for 5 meg dsl?

sm5w2
Premium Member
join:2004-10-13
St Thomas, ON

sm5w2

Premium Member

> If it is connected to a pots filter that could explain the improvement in the attainable rates
> with the inside wiring connected.

The modem is NOT connected to a pots filter. It is directly connected to the incoming phone line at the demarc point.

When ALL OTHER LINES IN THE HOUSE are disconnected from the demarc, the best attainable sync rate is 5 mbps.

When the other lines are re-connected to a filter which is then connected to the demarc, the best data rate goes up to 7 and the downstream noise margin goes from 7/9 to 15/16.

If the best possible situation for a DSL line/modem is to have the modem *directly* connected to the demark, with nothing else connected to the demarc, and with no filter anywhere in this circuit, then why am I seeing a worse noise margin in that situation vs connecting the rest of the house wiring to the demarc through a filter?

> As for the speeds, the profile is obviously set and won't
> change unless you request it. I am assuming he is paying for 5
> meg dsl?

It's a legacy Sympatico/HSE account (no caps). Like mine, it started as (I think) a 3-meg connection back in 2001, and has slowly increased over time. My sync rate got bumped up to 7040/800 about 3 years ago, after being 6016/800 for a few years. So too should this account get bumped up to at least 6016 or higher (7040) if the line allows it (or if it can allow it by changing to another circuit at the pole).

someguy
@216.211.83.x

someguy

Anon

There could be a few reasons why this is happening. The speed is definitely governed by the DSLAM in thise case, and if I read your post correctly it appears the rate cap is 4mbps/800kbps.. A speedtest of ~3.5ish would be acceptable when you include overhead etc. The attainable rate is what the line is capable of, however Bell would have you capped below that in this case to limit issues trying to push higher bandwidth down a line that cannot handle it using the currenty ADSL technology out of that DSLAM (I doubt it's 2+ OR vdsl/2).. I find it really odd though that when you pull the rest of the house off the SNR/Rate get's worse- usually it's the opposite. Those SNR readings to me are borderline as well, The higher SNR margrins, the better..

To answer your original question "How do the stats look as is?" - Marginal but acceptable, however there are factors such as the loop length and equipment type/compatibility that throw a few wild cards in there. I would suggest giving Bell a call to check the line and confirm what speeds/readings should be attainable based on the line in question and DSLAM at the C.O.

minetoo
@98.142.246.x

minetoo to sm5w2

Anon

to sm5w2
sm5w2, I my wiring works the same way for my DSL line. When the house wiring is connected, the stats are better. In fact, my line has always given the worst numbers at the NID.

If I use a centralized filter or a POTS splitter, that also makes my stats worse. If I instead leave all house wiring connected and NOT isolated from the modem and instead filter every phone device individually, my stats are also better. Before anyone says "defective filter" or "defective POTS spitter", remember that the connection at my NID (with no other wiring connected, only modem hooked up) is always worse than inside my house.

The only problem is this: good luck in trying to get something like this fixed. Most technicians are going to bluntly tell you that this scenario is impossible and will then blame your modem, even if you have multiple modems that you've already tried. The fact is that this situation is not impossible. DSL sometimes acts strangely. When DSL stats are worse at the NID/demarcation, this points solely on an outside wiring problem on Bell's end.

I believe you have a bridge tap on your line. My theory is that when you have the rest of your house wiring connected (and not using a POTS splitter or some other type of isolation method), the extra length of the loop causes the bridge tap to be further away. The closer a bridge tap is to your home, the greater the effect it will have on your line conditions. In my case (and yours), adding loop distance works better.

Even though this goes against the way things are supposed to work for DSL, go with whatever works for you.

someguy
@216.211.85.x

someguy

Anon

Good call on the bridge tap scenario. And good luck getting Bell to fix that without a hassle from them..(sometimes they can be tricky to find - and can use up a good chunk of time/resources to do it).

Do you ever find that you get a hum on the line or intermittent static/popping when it rains or during the spring melt-off? If you do, report it to bell and ask that they at least check your lines up to the demarc. After a few reports of that where they can't fix it they may give the trouble to one of the cable techs to check, and they might remove that tap at the same time... Just a thought anyways.

Cheers.